SEO terms, explained.

Average position.

Enlarge

Your average position roughly indicates where your website appears in the search results, so achieving a lower number is a better result. It can move up or down because your pages rank better, or because the number of searches your website appears for increases or decreases.

Beginning to appear for more valuable search terms can pull the average down even though nothing has worsened. This is why the figure is read alongside impressions and clicks rather than on its own.

Impressions.

Enlarge

An impression is recorded each time one of your pages appears in the Google search results for something a person has searched. It counts the appearance rather than the visit.

Where impressions increase while clicks stay flat, your website is usually appearing for broader or less relevant searches than before. The two figures are therefore always read together.

Clicks.

Enlarge

A click is recorded each time somebody chooses your result in the Google search results and arrives on your website. It is a straightforward count, and historically it was the clearest indication of whether your search visibility was producing visitors. That is now changing, because Google increasingly answers a search directly at the top of the results page using information taken from websites like yours, so the person reads the answer without clicking anything at all.

A fall in clicks while impressions hold steady therefore increasingly means the answer is being given before anybody reaches you, rather than that your position has slipped. Clicks are read alongside impressions and your visibility in AI answers rather than on their own. They also indicate nothing about the quality of a visit, which is what your engagement and conversion figures measure.

Click-through rate.

Enlarge

Your click-through rate is the proportion of people who saw your website in the Google search results and then clicked it. A low rate alongside high visibility usually indicates that the wording of your listing is not persuading anybody, which is considerably more cost effective to correct than a ranking problem.

The rate also falls naturally as you begin appearing for broader searches, since fewer of those people were looking for you specifically. It is therefore read alongside the search terms you are appearing for.

Brand versus non brand searches.

Some people search your business name. They already know who you are, and you were always going to receive that visit. Everybody else searches for what you do, and that second group is where new customers come from.

The two are reported separately for that reason. Your overall figures can appear flat while non brand searches, which are where new customers come from, move significantly in one direction.

Sessions, users & visits.

Enlarge

A user is a person. A session is one continuous period of that person using your website.

Neither figure matches the click count from search, which only counts arrivals from Google. Sessions also include people who typed your address directly, followed a link from elsewhere, or returned later.

Engagement.

Enlarge

Engagement figures show whether somebody did anything once they arrived on your website. Somebody who waits four seconds and leaves is very different from somebody who reads two pages and submits the form on your website.

Where a page attracts reasonable traffic but nobody remains on it, the problem lies with the page rather than with the search work, and correcting the page is usually the more cost effective of the two.

Conversions.

A conversion is somebody making contact: completing your enquiry form, telephoning you, or making a booking. Conversion tracking is what counts those, so the work can be judged on enquiries rather than on visits.

Clicks confirm that people arrived on your website. They do not confirm that anybody made contact.

Where conversion tracking has not yet been configured, your report says so and reports on traffic only.

Attribution.

Enlarge

Attribution is how a system decides which source to credit when somebody enquires. Every enquiry is credited to one source, even where several played a part.

The rules for deciding which source receives the credit vary between accounts and change over time, so channel figures can move without anybody's behaviour having changed. Where a large proportion of credit shifts between channels in one month, we treat it as a measurement change until we have established otherwise.

Data gaps.

Enlarge

Google withholds a proportion of search terms for privacy reasons, so the list of searches your website appears for is accurate but not complete. Some tools estimate rather than count, and some figures only exist from the date that tracking was installed.

Your totals are still correct. What is affected is any figure we calculate by grouping search terms, such as the split between brand and non brand searches, because the search terms behind some of those clicks are not disclosed. Those figures are close rather than exact, and we say so wherever one appears in your report.

Month on month comparisons.

Enlarge

Months differ in length and in shape, and a public holiday moves demand from one week into another.

A fair comparison requires two complete and comparable months, which does not exist in the early stages of an engagement. In that period the comparison is made against your starting position instead, and your report states which comparison it is using.

Rank tracking.

Enlarge

Rank tracking means checking, in the same way each month, where your website appears for a fixed list of search terms.

Alongside the overall figures we track the search terms of commercial value to your business.

Search results vary by location and by device, so a tracked position is a consistent measure rather than what any one person happens to see.

How we decide what comes first.

Not every item in your scope of work is completed each month. We prioritise according to where each item will have the greatest impact at the time.

Anything actively costing you revenue is dealt with first. Anything preventing us measuring properly comes next, because until that is corrected we cannot tell you what is working. Growth work follows. Each month's plan is built from those priorities.

What your retainer hours buy.

Your retainer covers a number of delivery hours each month rather than a fixed number of tasks. We complete as much of the planned work as those hours allow, and anything unfinished rolls over into the following month.

Some of what your retainer covers is a standing commitment, such as a piece of content each month or a set number of pages optimised. Where something more urgent comes up, we may spend that time on it instead and move the standing item into the following month. We will tell you beforehand.

Your monthly plan.

Your plan for the coming month is written at the end of the previous month, alongside your report. We review the month's figures, audit the website again, and set the work for the coming month from what those show.

The plan covers the work we will carry out rather than the results it will produce.

If something urgent comes up during the month the plan changes, and your next report sets out what changed and why.

The report at the end of the month.

Your report covers what moved, why it moved, the work completed, and what happens next. Any notable movement is accompanied by a cause and a proposed fix. It ends with the work planned for the coming month.

The figures come from established measurement tools rather than estimates: the analytics installed on your website, and Google's own record of how your pages performed in the search results.

How we measure.

The figures in your report come from Google's own record of how your website performed in the search results, and from the analytics installed on your website. They are counts of real activity rather than estimates.

A small number of search terms are not disclosed by Google, so a few of the figures in your report are close rather than exact.

Where this affects a figure in your report, it is noted beneath that figure.

What we need from you, & when.

We research and write content, send it to you, and publish it once you are satisfied with it. Anything that changes what your business says about itself follows the same route.

Technical and structural work proceeds without needing your approval, because it does not change your wording or your claims.

Where we are waiting on an answer from you, it appears in the feedback section of your report. An unanswered question holds up the work behind it.

Why the first month looks different.

The first month is spent establishing rather than improving. We obtain access, establish what is true about your website, and record its starting position.

It also includes mapping every page to the searches it should be winning. That is a substantial piece of work and it uses that month's content allocation.

It is the least visible month of the engagement, and every month afterwards is built on what it produces.

Your starting baseline.

Before we make the first change we record your current position: search visibility, traffic, enquiries, page speed, business listings and reviews, each with the date range it covers.

Where a baseline cannot be captured, usually because tracking was never installed, that is recorded as well.

Blocked work.

Some work cannot begin until access has been granted, a question has been answered, or a developer has made a change. Those items appear in your plan and in your report, with a note identifying who we are waiting on. An item can be held up for several months, and it stays in the plan until it is resolved.

Requests that arrive mid-month.

Requests arise during the month. A page needs changing, a price is incorrect, something stops working.

These are accommodated, and the time comes from the same hours as the planned work.

Where a request is substantial enough to displace something already planned into the following month, we will tell you before we begin it.

What is not included.

Search work does not include paid advertising, managing your social media accounts, or building your website from scratch, unless those have been agreed separately. Where something outside our scope is clearly costing you revenue, we will tell you about it, even when the fix is not ours to carry out.

Access requests.

To report on your website accurately we need to see its own figures rather than estimates taken from outside. To carry out the work we also need to be able to make changes to the website itself.

Access is granted to a named user inside your own account. We do not ask for your password, you remain the owner of every account, and you are able to remove our access at any time.

Where access has not been granted, we work from what can be seen from outside the website, and your report sets out which work is prevented as a result.

Account ownership.

Anything we set up for you is created in your own name, with us added as a user.

Where an account already exists under a previous agency's login, we will ask you to request access to it from them.

Publishing rights.

Some of this work requires the ability to edit and publish pages on your website directly.

Where we have that access, changes are made once they are approved. Where we do not, the change is passed to whoever has it, and it appears in your plan as waiting on them.

We also establish how many people hold full administrator access to your website, because work we complete can be overwritten by somebody else editing the same page.

Existing accounts.

Before we set up any account, we check what already exists. That includes tracking codes still present on your pages, analytics and search accounts created by a previous agency, and business listings that were never claimed.

Where an existing account is found, we use it. Where it exists but cannot be reached, we record that and set up a replacement.

Keyword research.

Enlarge

Keywords, also called search terms, are the words somebody types into Google. "Emergency electrician Claremont" is a search term. So is "how much does rewiring cost".

Keyword research is the work of establishing which search terms are worth your website appearing for. We begin with what you sell and with what people actually type, which are rarely the same words. We then examine how many people search each phrase, how difficult it would be to compete for it, and whether the person searching is ready to buy or still gathering information.

A phrase with fewer searches and clear buying intent is therefore worth more than a popular one that attracts browsers.

Building your keyword map.

A keyword map is a record of every page on your website and the searches each one is meant to win.

We build the map at the start of the engagement and keep it current as pages are added or retired.

The map stops two of your pages competing for the same search, and it shows where there is demand your website does not yet cover. Your content plan is drawn from it.

Keyword groups.

A keyword group is a set of search terms that all mean the same thing, served by one page.

People ask for the same thing in a dozen different ways, so rather than building a page for each phrase, we group the phrases that mean the same thing and point a single page at the group.

This prevents your website fragmenting into near-identical pages, and it makes clear which of your pages should win each search.

Landing pages & supporting pages.

Within each keyword group, one page is the landing page, meaning the page intended to rank. Every other page on that subject is a supporting page, and it links through to the landing page.

We decide which page owns each group and ensure the supporting pages link to it.

Deciding this at the outset makes later decisions straightforward: any new article is a supporting page, and it links through to the page that sells.

Keyword cannibalisation.

Cannibalisation is what happens when two of your own pages compete for the same search, and both do worse as a result.

Any credit you have built up is also divided across the two rather than accumulating on one.

We identify where this is happening and correct it. The correction is to decide which page owns the subject, move anything useful out of the other, and then point the second address at the first.

Search volume.

Enlarge

Search volume is an estimate of how many people search a particular phrase each month. It is useful for comparing one phrase against another, not as a prediction of how much traffic you will receive.

We use these figures to compare one phrase against another when deciding what to target.

The figures are also rounded, and they are averaged across the whole year. A phrase searched heavily every winter and not at all in summer shows the same middling number in both seasons.

Keyword difficulty.

Enlarge

Keyword difficulty is a score estimating how well established the websites already ranking for a phrase are. It is a shortcut rather than a verdict, and it says nothing about whether their pages are any good.

We use the score to sort a long list quickly, then examine the results themselves for anything of commercial value.

Content gaps.

Where people are searching for something and no page on your website answers it, that is a gap, and it is usually the clearest justification for building something new.

We identify gaps by comparing what is searched in your market against what your existing pages are about. A gap is only worth filling where the people searching are potential customers.

Excluded search terms.

Some phrases attract entirely the wrong people: job applicants, students, buyers in countries you do not serve, or a business with a similar name.

We record these deliberately so that nobody pursues them, and so that next year's review does not rediscover them as an opportunity.

Competitor benchmarking.

Benchmarking means measuring your website against your competitors' on the factors that determine search results: how many websites link to them, how much content they hold, and which searches they appear for.

We use the businesses genuinely competing for your customers rather than the biggest names in your industry.

The useful output is the list of searches where a competitor appears and you do not.

Your competitor set.

The businesses you lose work to are not always the ones ranking above you, and some of those ranking above you are directories or national brands you will not outrank. We therefore ask you who you compete with and work from that. Comparing against the wrong set of businesses produces work that appears productive and changes nothing.

We review the set again during the year, because businesses enter and leave a market and the set that was right in month one is not always right in month eight.

Meta titles & meta descriptions.

Enlarge

A meta title and meta description are the headline and the short summary that appear in the Google search results. They are written for the results page rather than shown on the page itself.

For most people this is the only part of your website they see before deciding whether to click.

We write these for the commercially important pages, and replace the ones that are vague or missing.

Google sometimes rewrites these, but it uses a clear one far more often than a vague one.

Title length.

The search results display approximately sixty characters of a title before trimming the remainder.

A longer title is not penalised, but the end of it is invisible. Where a page is already ranking well we leave its title unchanged, because rewriting a title that is working is one of the few changes capable of losing ground for no benefit.

Heading hierarchy.

Enlarge

Enlarge

A page should carry one main heading, followed by subheadings beneath it in a sensible order. This is how people skim a page and how Google establishes what it is about.

The second page is the same idea done correctly. One main heading states what the page is for, each section beneath it sits a level down, and the individual services sit a level below that. A reader skimming it and a search engine reading it arrive at the same understanding.

We check the heading hierarchy on each page and correct it where it is wrong. Where the order has been broken, or where a design choice has turned ordinary text into a heading, Google reads the page as being about the wrong subject. Correcting it changes nothing you can see on the page.

H1 headings.

Enlarge

One heading on each page sits at the top of its structure, and it should state what the page is about in the words a customer would use.

Some pages carry no main heading at all, because the large text at the top is a design element rather than a genuine heading. Others carry five, because a template is applying heading styling to a list. Either way, Google is left without a clear statement of what the page covers.

We check every page for a main heading and add or correct it where necessary.

Your home page.

Your home page is usually the page other websites link to most often, which makes it the strongest page on your website. It is also where Google looks to establish what your business does.

Some of that strength passes to the pages your home page links to.

We work on the home page's headings, its copy and the pages it links to.

Your home page rarely wins specific searches on its own. Its value is in making the pages that do win them stronger.

Service pages.

Enlarge

A page can only rank for what it is about. Where three services share a single page, that page competes weakly for all three rather than strongly for one.

We identify where services are sharing a page and build the pages that are missing. Giving each service its own page, written in the words customers use, is usually the largest structural gain available on a website. It also means the person enquiring already knows which service they want.

Location pages.

Enlarge

Where people search using a place name, a page for each area you serve can win searches your main service page never will.

We build these where the search behaviour justifies them. Each page must say something genuinely different about that area, rather than being the same text with the town name swapped. People must also search that way, which they do for a plumber and rarely do for a bookkeeper.

Google sees duplicated location pages as spam and does not show them in the search results.

Thin pages.

A page consisting of a heading, a photograph and two sentences gives Google very little to assess, and gives a visitor very little reason to remain on it.

Thin pages on the route to an enquiry can cost you revenue directly rather than only costing you rankings. The correction is either to expand the page, or to combine it with a page that is already working.

FAQ sections.

Enlarge

The questions customers ask you on the telephone are the questions they type into Google first. Answering them plainly on the page helps you appear for those searches, and means fewer potential customers need to contact you simply because the answer was not readily available on your website.

We take the questions from you, or from the questions already arriving by telephone and email, and write them into the page.

AI assistants such as ChatGPT and Gemini also quote this type of content more readily than any other, because it is already set out in question and answer form.

Your contact page.

Your contact page is frequently the last page somebody views before making an enquiry, and it is frequently the thinnest page on the website.

It should carry your address, your opening hours, a map, a working enquiry form, and a sentence explaining what happens once somebody has submitted it.

It is also the page where an incorrect telephone number, or a form delivering to an inbox nobody monitors, causes the most damage, so we verify these rather than assume them.

Your about page.

Your about page is read more often than most business owners expect, usually by somebody deciding whether your business can be trusted with the work.

It is also where Google, and AI assistants such as ChatGPT and Gemini, look for the basic facts about your business: what you do, where you operate, how long you have been established, and who runs it.

Specific detail on this page therefore serves both purposes at once.

Your team page.

Naming the people who carry out the work, with their roles and photographs, is one of the clearest indications that a real business sits behind the website.

We set the page up and write it once you have supplied the names, roles and photographs.

It also supports the author credit on anything you publish, and gives AI assistants such as ChatGPT and Gemini a named person to credit as the source.

Legal pages.

Every website requires a privacy policy and terms of use, and they are usually the last pages to receive any attention.

They regularly carry generic text taken from a template that still names another company, or they have been hidden from Google for no reason.

We check what is published, inform you of anything carrying another company's name, and make the pages indexable where they should be.

Neither affects your rankings. Both are visible to anybody assessing whether your business is credible, and both are quick to correct.

Content depth.

Content depth refers to how much content is on a page. There is no ideal length, and a page does not perform better simply for being longer. The determining factor is coverage: whether the page answers the questions a customer arrived with.

A page is built in layers, because it has two audiences. The top is short and visual, and it is where somebody decides they are in the right place. The middle is what a serious buyer needs to know before they will make contact. Underneath sits the depth, the appended sections and questions, which is where Google and AI assistants find what to quote.

We expand the pages that are answering too little, working from the questions your customers actually ask. A page can be very long and still answer none of them.

Calls to action.

Enlarge

A call to action is the button or link telling somebody what to do next: request a quote, book a call, download a guide.

We review the calls to action across your pages, add them where a page has none, and change them where the wrong one is being offered. The action has to suit how far the reader has progressed. Somebody halfway through an article about choosing a geyser is not ready to book a site visit, but they will read a guide to geyser types. Offering the site visit at that point loses them.

Calls to action have no effect on your rankings and a considerable effect on the number of enquiries your website produces.

Clickable contact details.

A contact number for your business presented as ordinary text on your website is not clickable if a customer presses it on mobile. When presented as a link, a single tap leads to a call and potential revenue.

This is also the only way telephonic enquiries can be recorded as conversions, which means that a website without it is only reporting on a portion of the responses it receives from visitors.

Many websites lack both clickable phone numbers and email addresses. We check every page for this and correct it, usually in the shared footer or navigation bar so it applies to the entire website at once.

Hidden text.

Hidden text is wording that exists in a page's code but never becomes visible to a visitor. It is almost always accidental: an animation meant to fade text in has failed, or a tab's contents are never revealed.

Google reads what is in the code and judges the page on words a customer never sees, while the customer sees a page that looks half empty.

We compare what each page contains with what a visitor actually sees, and correct the cause. It is normally one setting rather than a rebuild.

Social sharing tags.

Social sharing tags decide what appears when somebody pastes a link to your website into WhatsApp, Facebook or LinkedIn: the headline, the short description and the image.

Where the image is missing, the link is shared as a bare line of text and is far less likely to be opened. Where the tags are missing altogether, the platform chooses something itself, which is usually your logo or whatever image sits first on the page.

We set these across your pages and give each of your commercial pages an image of the right proportions.

Placeholder content.

Placeholder content is text left behind from a website template that was never replaced with your own. It appears as filler paragraphs, a page still titled sample page, or a heading that still reads your headline here.

Google indexes this text in exactly the same way it indexes the rest of your website, which means a page of filler can appear in the search results under your business name. It does no lasting damage to your rankings, but it does damage the impression a potential customer forms of your business.

We identify any placeholder text during the audit and clear it, which usually takes a few minutes once it has been located.

Out of date claims.

Anything published on your website that is no longer accurate is a business problem before it is a search problem. That includes a branch you have closed, an award from a scheme that has since ended, or a service you no longer offer.

Search work brings more visitors to your pages, which means an inaccurate claim is read by more people than it was when nobody was finding the page. It can also produce enquiries for work you are not able to carry out.

We cannot tell from outside your business which claims are still true, so we will inform you of anything that reads as dated: a year that has passed, an accreditation with no renewal shown, a service mentioned on one page and nowhere else, or an address that disagrees with your Google Business Profile. You confirm what is wrong, and we correct or remove it.

Monthly content.

Content covers everything we write for your website. It includes landing and service pages, sections added beneath existing pages, frequently asked questions, location pages, articles, and rewrites of pages you already have.

The topic, the questions answered and the structure of the page come from research into what people are searching for and what Google and AI assistants need in order to rank and quote you.

A page cannot be found for something it never says. AI assistants such as ChatGPT and Gemini quote sentences, so where your answer is not written down, a competitor's is quoted instead. A fact that only exists in an image, a PDF or a video cannot be quoted at all.

Setting up authorship.

Authorship means crediting the content on your website to a named person at your business, with their role and a photograph, rather than publishing it anonymously.

Google states openly that it favours content with visible expertise behind it, and AI assistants such as ChatGPT and Gemini rely on it more heavily still when deciding which businesses to quote in an answer.

We create the author biography and add it to your blog template, which means it is set up once and then applies to everything you publish afterwards. All we need from you is the name, the role and a photograph.

Appended content.

Appended content is a section added beneath a page that already exists, rather than a new page or a change to the page as it stands.

It adds depth without touching the design at the top. A category page that is currently a grid of products, or a service page that is three sentences and a photograph, can gain the content it needs to compete without being rebuilt.

We write these sections and send them to you as a draft in the same way as any new content.

Your own information.

Your own information means the figures, examples and patterns that come from running your business. No competitor can publish the same thing, because no competitor has it.

We ask you for figures like these and write them into your pages. Google and AI assistants such as ChatGPT and Gemini favour pages carrying information they cannot find anywhere else. These pages tend to rank higher and are quoted more often than the general advice every competitor is publishing.

Content approval.

New content is sent to you by email as a draft, with the date that we intend to publish it on. Nothing new goes live until you have approved it.

What we need from you is a factual check: anything that is not true, and anything you would prefer not to promise your customers. The topics, the questions answered and the structure are decided from research before the draft reaches you.

Where a sentence is wrong, tell us the correct position and we will amend the draft and return it. Where a claim commits you to something you would not wish to honour, we will reword it.

Changes to a page you already have are written up in your monthly report rather than being sent for approval.

Content refreshes.

A content refresh means rewriting or extending a page that is already published, rather than writing something new.

It is frequently more productive than writing something new, because the page already has visitors and Google already trusts it. Refreshing it means correcting anything that has changed, such as prices, dates or a service you no longer offer. It also means answering the questions customers are asking now that the page does not cover, and linking the page to ones that did not exist when it was written.

We identify which pages are worth refreshing and dedicate your monthly content allocation to the improvement. Your report tells you which page was refreshed and what was changed.

Retiring content.

Some pages are worth removing from your website: pages written for customers you no longer serve, pages describing an offer that has ended, or pages so thin that they confuse Google about what your business does.

Removing a page is not the same as deleting it. We point the old address at the closest relevant page instead, which means anybody following an old link still arrives on a relevant page, and the value that page had accumulated is not lost.

We identify these during the audit and confirm each one with you before it is removed.

Regulated copy.

In some industries the law or a professional body limits what you are permitted to say in writing. In others, any claim about earnings, results or timescales carries risk regardless of the sector.

Where this applies to your business, we agree the wording rules with you once and write inside them from then on.

Photography.

Photographs of your own work, your premises and your people provide evidence of your work that stock photography cannot.

They also carry location information that supports your visibility in local searches, provided that information is present at the point that the photograph is uploaded rather than added afterwards.

We ask you for photographs of your own work and use them in place of stock images wherever possible. Where a business genuinely looks different from its competitors, its own photographs are usually the most under-used asset it owns.

Alt text.

Alt text is a short written description attached to an image on your website. Screen readers read it aloud for visually impaired visitors, and Google uses it to understand what the photograph shows.

We write these descriptions, working through the commercially important pages first. It is also a legal accessibility requirement rather than an optional improvement.

Stock photography.

Stock photography is acceptable as a placeholder but weak as evidence. The same images appear on your competitors' websites, and they show nothing that belongs to your business.

On a page intended to demonstrate that you are capable of the work, a stock photograph undermines the argument the rest of the page is making, which can cost you enquiries rather than rankings.

We will inform you where stock photography is weakening a page and replace it once you have supplied alternatives. We start with the pages that sell your services, because replacing the images on those few pages is worth considerably more than replacing every image on the website.

Internal links.

An internal link is a link from one page on your website to another page on the same website.

Links between your own pages indicate to Google which of those pages you consider important. A page that nothing links to is difficult for Google to find, and it is treated as a page of little importance however well it is written. The wording of the link is equally important, because Google reads it as a description of the page it points to. For example, a bicycle shop's article on choosing a helmet should link to the helmets page using the words "mountain bike helmets" rather than the words "click here".

We add links from your blog posts and related pages through to the pages you want to rank, using wording that describes the page being linked to.

Navigation.

Enlarge

Your navigation menu appears on every page of the website, which means whatever it contains receives a large share of the links pointing around your own site.

That makes it valuable, and it makes it easy to waste on links to empty categories, to pages that only redirect somewhere else, or to items nobody selects. It is also the shortest route a customer has from your home page to the pages that sell.

We review what is in the menu, remove the entries that are not earning their place, and ensure your most commercial pages are included.

Site structure.

Site structure refers to how the pages on your website sit in relation to one another, and how many clicks it takes to reach each one from your home page.

A page that can only be reached by clicking through three unrelated pages is treated by Google as a page of little importance, and customers rarely find it at all. Website structures deteriorate as pages are added one at a time over several years. The usual symptom is an important service page sitting three clicks away from the home page when it should sit one click away.

We map how your pages currently sit and move the commercially important ones closer to the home page.

Content silos.

A silo is a group of pages covering the same subject, deliberately linked to one another and up to a single main page on that subject.

We plan the groupings and add the links between them.

Orphan pages.

An orphan page is a page on your website that nothing else links to, which means it can only be reached by somebody who already knows its address.

These appear after a campaign has ended, after a menu has been redesigned, or when something has been published outside the normal process. The page is live and indexed, and it receives no visitors at all.

We identify them during the audit. Each one is then either linked in properly from the pages it belongs with, or retired.

Breadcrumbs.

A breadcrumb is a small line near the top of a page showing where that page sits within your website, with each step in the line clickable.

We add breadcrumbs where your website has categories and does not already display them.

URL structure.

A URL is the web address of a page, the line that appears at the top of the browser. A good one is short, readable, describes the page, and remains unchanged over time.

We will inform you of addresses that are working against you and change them, with redirects in place, where the benefit outweighs the disruption. Once an address has been published, changing it carries the consequences set out under redirects.

Sitemaps.

A sitemap is a file on your website listing the pages that are worth Google indexing. We submit it so that Google can read the list directly rather than relying on finding every page by following links.

It should contain the live pages you want customers to find, and nothing else. Problems arise where a sitemap still lists deleted pages, or lists pages your website elsewhere tells Google to ignore. Google is then receiving two contradictory instructions, and it resolves that by trusting the file less.

We check what your sitemap contains, correct it, and submit it. The same pass sets the page types that should never be in it, such as author, tag and date archives, to stay out of the search results, because each one is a near-empty page competing with the real ones.

Indexing.

Indexing is Google storing your page in the collection it searches. Finding a page and indexing it are two separate steps, and Google can read a page and still decide not to store it.

Only indexed pages are able to appear in the search results at all. A service page published in March that is still not indexed in June is not ranking poorly. It is not competing at all.

We check which of your pages are indexed and establish why the rest are not. Some are held back by a setting on the website or by a lack of links pointing at them. Others are not stored because Google has judged the page too thin or too similar to another, and in those cases improving the page is what gets it indexed.

Discovered but not indexed.

Google frequently finds a new page, records that it exists, and then leaves it for several weeks before deciding whether to store it.

The most common reason for a long delay is that nothing on your own website links to the page, so Google treats it as a page of little importance. A new service page that only appears in your sitemap can remain in this state for a month or longer, while the same page linked from your home page and from two related articles is usually stored far sooner.

We monitor which of your new pages are waiting and add the internal links that get them stored.

Crawling.

Crawling is the process by which Google finds pages, following links from one page to the next across the whole web.

Google reaches your website from links on other websites, from your own menu, and from the sitemap you submit. A page with no links pointing at it may never be found, and Google has to find a page before it is able to rank it.

We check that Google is able to reach every page on your website you want customers to find.

How search engines choose content.

Search engines hold a large collection of pages they have found and stored. When somebody searches, they select the pages that best answer that search and place them in order.

Google has to find your page, then decide the page is worth storing, and then judge it a better answer than your competitors'. A failure at any one of those three stages means your page does not appear, regardless of how good it is.

Most of the work we do is directed at one of those three stages.

Ranking factors.

Ranking factors are the elements Google weighs when deciding the order of the results.

Google considers how well your page matches the search, how clearly your website covers the subject, what other websites say about you, and whether the page works properly when it loads. The exact weighting applied is not published, and it changes over time.

We work on the factors we are able to influence: how well your pages match what is searched, how clearly your website covers the subject, how many other websites link to your website, and whether your pages load properly.

Search intent.

Enlarge

Search intent means what the person searching actually wants, which is not always obvious from the words they have typed.

We check the search results for each phrase we are targeting to establish what type of page Google is rewarding, and write to match it.

Search results features.

Enlarge

A search results page is rarely ten plain links any longer. It contains maps, images, questions that expand when selected, shopping panels, and answers Google has written itself.

All of this pushes the ordinary results further down the page, which means your website can hold the same position it held last year and still receive fewer visitors than it did. This is why your position and your clicks are read together rather than separately.

We record what appears alongside your results, because it changes what a given position is worth to your business.

Seasonality.

Most businesses have a busy period and a quiet one, and the number of searches follows the same shape.

Where a full year of history exists we compare against the same month in the previous year as well as against last month.

Algorithm updates.

Google adjusts the way it ranks pages continuously, and several times a year it makes a change large enough to move a large number of websites at once. These changes are announced, but they are not explained in enough detail to act on directly.

Where results move sharply across an entire industry within the same week, an update is usually the cause rather than anything that has happened on your website. Making changes in response, before the results have settled, tends to worsen the position.

We watch for these and tell you when one is the likely cause of a movement in your figures.

Manual actions & security warnings.

A manual action is a penalty applied to your website by a person at Google rather than by its software, usually for buying links or for pages written to manipulate the results. A security warning means Google has found malicious code on your website, which normally arrives through an out of date plugin.

Google reports both in one place. We check them at the start of the engagement and again each month, because either one makes every other piece of search work pointless until it is resolved.

Staging & development copies.

A staging copy is a second, working version of your website used to build and test changes, usually at an address such as staging.yourbusiness.co.za. It is meant to be invisible to search engines.

Where it is left visible, Google stores it, and every page on it duplicates a page on your live website. The copy competes with the real website for your own searches, and a customer can land on a half-finished version of a page.

We check whether any copy of your website is reachable and stored by Google, and have it hidden. It is most common in the weeks after a website has been built or redesigned.

One address for your website.

A website can usually be reached at four addresses: with and without www, and with and without the padlock. One of them should be the real one, and the other three should lead to it.

Where they do not, Google can treat them as separate websites, and everything your pages have earned is divided across up to four copies of each one instead of accumulating on a single page.

We check which of the four addresses respond and make sure they all lead to one. It is a one-off correction, and because everything else sits on top of it, it is checked before any other technical work.

Soft 404s & server errors.

A soft 404 is a page that tells the visitor it cannot be found while telling Google that everything is fine. Google keeps it in the search results and nobody sees an error to report.

A server error is different: the page fails to load at all. A handful over a month is ordinary. A pattern of them means Google is being turned away and will read your website less often.

We check for both, including pages that fail only at particular times of day, which is the version most likely to go unnoticed.

JavaScript rendering.

Some websites are built so that the page arrives empty and the words are added a moment later by code running in the visitor's browser. Google is usually able to read those pages, but not always, and not always in full.

We check what your pages contain before that code runs and what they contain afterwards. Where the two differ significantly, we tell you what has to change.

The robots file.

The robots file is a small file at the root of your website. It tells automated visitors, such as Google's crawler and the crawlers used by AI assistants like ChatGPT, which parts of the website to skip. Those are usually areas containing nothing useful to a searcher, such as administration screens, baskets and internal search results.

The file works equally well in the wrong direction, and a single incorrect line is capable of hiding your entire website from Google. A website that has disappeared from the search results overnight has frequently had a line added here by mistake.

It is one of the first things we check on any new account, because everything else depends on it being correct.

Noindex.

Noindex is an instruction telling Google that a page exists but should not appear in the search results.

Some pages need it. Somebody arriving directly on your thank-you page from Google has not submitted anything and will not understand what they are viewing. Problems arise where the instruction has been applied to a page that should be found. They also arise where it has been applied while your sitemap is still submitting that same page, which gives Google two opposite instructions at once.

We check which of your pages carry this instruction and correct the ones that should not.

Canonical tags.

A canonical tag is an instruction naming which version of a page is the real one, where the same content can be reached at more than one web address.

The tag causes problems when it points at the wrong page. A page whose tag names a different page as the real version will never appear in the search results, because the tag is telling Google not to show it. This usually happens where a website adds the tag automatically and somebody has also set one by hand. We check where each page's tag points and correct the ones that are wrong.

Duplicate content.

Duplicate content means the same content appearing at more than one web address on your website.

Where that happens, Google has to choose which address to show, and it sometimes chooses the wrong one.

The two versions also compete with one another rather than adding together. We find duplicates during the audit and name one of them as the real version, which consolidates everything they had earned separately onto a single page.

Redirects.

A redirect is an instruction that automatically sends anybody opening an old web address on to a new one. We put one in place whenever a page moves.

This means anybody following an old link still arrives, and the value the page had built up over the years moves across to the new address instead of being lost.

Google takes a few weeks to work through a change of this kind, so your position for those searches frequently dips before it recovers. That dip is expected, and your report will say when a change of address is the cause.

Redirect types.

There are two kinds of redirect. A permanent redirect tells Google the change is final and moves everything the old address had earned across to the new one. A temporary redirect tells Google to keep the old address on file.

We check which kind is in place on each redirect rather than only confirming that a redirect exists.

404 errors.

Enlarge

A 404 is the error message somebody sees when they open a web address that no longer exists. A small number are unavoidable and do no harm.

They require attention where another website still links to the address, where people are still arriving on it from the search results, or where several hundred remain from a migration that was never finished. For example, if a supplier's website links to a service page you removed two years ago, everybody following that link reaches an error message, and the value that link was passing to you is lost entirely.

We list the addresses returning errors and point the ones worth keeping to the correct page.

Broken links.

A broken link is a link pointing at a page that no longer exists. Where the link is on your own website it wastes the visit, and where it is on somebody else's it also wastes the credit that link was passing to you.

Broken links accumulate as pages are renamed or removed, and nothing alerts you when it happens.

We check for them across the website and repair them, which usually takes a few minutes for each one.

HTTPS & SSL certificates.

The padlock shown in the address bar indicates that the connection between your visitor and your website is encrypted.

Browsers actively warn people away from pages without it, particularly pages containing a form, which means a visitor may never reach your enquiry form at all. Google treats having it as a baseline expectation rather than an advantage.

We check this is in place across the whole website. Where a certificate has lapsed, or where only part of the website is covered by one, it is a live problem rather than a technical detail.

Crawl budget.

Crawl budget is the number of pages Google is prepared to read on your website within a given period. On a small website that limit is never reached.

We identify what is absorbing that attention and remove it from the crawl, so the pages that sell are read more often.

Hosting & server response.

Your server is the computer your website is stored on, and how quickly it responds is the first component of how quickly anything appears on a visitor's screen. It is entirely separate from how the page itself has been built.

We measure the server response separately from the page itself and tell you which one is the cause.

Plugins.

Most websites run on a collection of add-ons, each one adding features and each one adding weight to the page.

Problems arise from having two that perform the same function, from old ones left installed but deactivated, and from ones that load their code on every page for a feature used on a single page. This is where the majority of website speed problems originate.

We review what is installed and inform you of what should be removed, either by your developer or by us.

Making changes safely.

Anything that affects how your website is built is preceded by a backup, carried out outside your busy hours where it could interrupt a customer, and checked on the live website afterwards.

Where several people hold administrator access to your website, we check again the following month, because pages are sometimes overwritten by somebody else working on the site.

The backup is what allows a change to be reversed if it behaves unexpectedly.

Website redesigns & migrations.

A migration is any change to the foundations of your website: a redesign, a move to a different platform, or a change of domain name. It is the event most likely to lose a website its search visibility.

Where you are planning one, tell us before the work begins rather than afterwards. We record every existing address and how it performs, supply the list of redirects the new website needs, and check the new website against that record once it is live.

Language & region settings.

A setting on your website declares which language and which country each page is written for.

Where that setting is missing or incorrect, your website can be shown to searchers in the wrong country, or shown less often to searchers in its own. It is most significant for a business trading in one country from a web address that does not show which country it serves.

We check this setting and correct it where it is wrong, which is usually a single line.

Page speed.

Page speed is how quickly your website appears and becomes usable for a visitor. It is measured on real visits rather than on a test, because a page can score well in a test and remain slow on a phone using mobile data.

Google uses it as one input into rankings. Visitors also leave slow pages before they have seen anything at all, which can cost you an enquiry.

We measure your speed on real visits and work through the causes in order of how much each one will gain. Most of the available improvement comes from the size of your images and from code that loads before the page becomes usable.

Mobile usability.

Google judges your website on its mobile version rather than its desktop version, even for people searching on a laptop. A website that looks correct on a large screen and cramped on a phone is therefore being judged on the cramped version.

Most searches now happen on a phone, which means a customer's first impression of your business is usually formed on the version that has had the least attention.

We check whether buttons are large enough to tap accurately, whether text can be read without zooming in, and whether any part of the page spills off the side of the screen.

Core Web Vitals.

Core Web Vitals are the three measures Google uses to judge how a page performs for a real visitor: how long the main content takes to appear, how quickly the page responds when somebody taps something, and how much the layout moves while it is loading. Each measure has a pass mark.

All three are taken from real visits across a rolling period, which means an improvement takes several weeks to appear in the figures even once the work itself has been completed.

We report all three each month and work on whichever is failing.

Layout shift.

Layout shift is content moving down the screen as an image or advertisement arrives late in the loading process.

We identify what is arriving late and reserve space for it.

Image optimisation.

A photograph taken straight from a camera can be several megabytes in size and then be displayed on your website at a fraction of those dimensions.

On a phone using mobile data, that difference is the difference between the page appearing and the visitor abandoning it. Resizing and compressing images to the size they are displayed at is usually the single largest speed improvement available, and the page looks no different afterwards.

We resize and compress the images that are costing you the most.

Render blocking code.

Render blocking code is code that has to finish running before anything appears on a visitor's screen. Every add-on, tracker and embedded widget introduces some.

We identify what is loading unnecessarily and either remove it or delay it until after the page is usable.

Caching.

Caching means keeping a finished copy of a page ready to serve, rather than building the page from scratch on every visit. It makes a substantial difference to how quickly the page appears.

It introduces one complication. After a change has been made, visitors can still be served the previous copy, which is why a change can appear not to have worked when it has.

We clear the stored copies after every change we make, and check the result on the live website.

Structured data.

Structured data is a description of what a page contains, written in a format Google reads directly and placed behind the page rather than shown on it. It states that this is a product, that it costs a particular amount, that it carries these reviews, or that these are your opening hours.

We add this to the pages where it can earn you that additional space.

Organisation schema.

Organisation schema is a single description of your business itself, written in a format search engines read directly: legal name, logo, contact details, and the social profiles that genuinely belong to you. It appears on every page of your website.

It is what search engines, and AI assistants such as ChatGPT and Gemini, read in order to establish who your business is. Where it has been assembled from three separate sources, those sources disagree with one another, and one of them is usually out of date.

We write this once and hold it in a single place on your website.

Local business schema.

Local business schema describes the location itself for a business with premises or a service area: where you are, when you are open, the area you cover, and how to contact that branch.

Where there are several branches, each page should describe its own branch rather than every page describing head office, otherwise Google cannot tell your locations apart. It supports your Google Business Profile listing rather than replacing it.

We add this to the pages describing each of your locations.

Product schema.

Product schema describes each product in a format Google reads directly: what the item is, what it costs, whether it is in stock, and how customers have rated it.

It is what allows a price and an availability status to appear in the search result itself, before anybody has visited your website. It has to match what the page displays, because a price in the code that disagrees with the price on the page gives Google a reason to disregard all of it.

We add this across your product pages and check it against what each page shows.

Review & price schema.

Enlarge

Review and price schema describes your ratings and prices in a format Google reads directly, so that your search result can display a star rating or a price rather than only a link.

Google only permits this where the reviews are already visible on the page and are genuinely from customers, so we add the reviews to the page first and then describe them.

FAQ schema.

FAQ schema describes the questions and answers already published on your page in a format Google reads directly.

Google used to display those questions underneath the search result itself. Since 2023 it only does so for government and health websites, so for an ordinary business the questions no longer appear in the results.

We still add it. It gives Google and AI assistants such as ChatGPT and Gemini a clean, labelled set of questions and answers to read, which is the form they quote from most readily. We add it to the pages carrying visible questions and answers, and as with review schema, the questions have to be published on the page itself.

Article schema.

Article schema describes a published article in a format Google reads directly: the headline, who wrote it, when it was published and when it was last updated.

It is what tells Google who wrote the article, rather than leaving it to be worked out from the page. The dates carry more weight than they appear to, because a refreshed article still declaring a publication date from three years ago is judged by Google on that date.

We add this to your blog template, which means it applies to everything you publish afterwards.

Video content.

Enlarge

Where your website carries video, a description of it can be placed behind the page in the same format as your other structured data: what the video shows, how long it runs, and a still image to represent it.

We check whether your videos are described this way and add it where they are not. Where a video is hosted elsewhere and only embedded on your page, we make sure your page is the one being described.

Schema errors.

Structured data is code, and code containing a single misplaced character fails without producing any warning at all.

The page looks entirely normal, the description is sitting behind it, and nothing is able to read it. Testing tools report no errors, because there is nothing present for them to read, which means a broken description can sit on every page of a website for years.

We compare how many descriptions Google acknowledges against how many your pages are serving, which is the only way the failure becomes visible.

Analytics.

Analytics is tracking code installed on every page of your website that records visits, which pages were seen, and which actions were completed.

It only knows what it was configured to record, and only from the date that it was installed. Almost every measurement problem originates in one of those two facts rather than in the figures themselves being wrong.

We check what your analytics is configured to record and correct it where it is recording the wrong things.

Search performance data.

Separately from the analytics on your own website, Google keeps its own record of when your pages appeared in the search results, what was searched, and who clicked.

It is the source for everything in your report concerning search visibility, and it is the only place the actual search terms appear. It covers a rolling window rather than your full history, which is why a starting position is recorded early in the engagement.

We use it as the source for the search figures in your report, and we read it page by page rather than only as a site total, because that is what shows which searches each individual page is winning and which ones it is appearing for by accident.

Bing & other search engines.

Google is where the great majority of searches happen, and your report is built around it. Bing is significant for a second reason: Microsoft Copilot and parts of ChatGPT draw their answers from Bing's record of the web rather than Google's.

A website missing from Bing can therefore be missing from those answers even where it performs well in Google.

We register your website with Bing during setup, submit your sitemap to it, and check that its record of your website is complete. Your monthly report remains a Google report unless you ask for more.

Tag management.

A tracking container is a single piece of code installed once on your website, inside which every other piece of measurement code is then added. It means your website does not have to be edited each time something new needs to be measured.

It is efficient, but it can fail in a specific way when containers are duplicated or old ones are left in place after a website is rebuilt. The result is that the same visit is counted twice by two copies of the same thing.

We check what is installed, remove the duplicates, and manage what sits inside the container.

Key events.

Enlarge

A key event is an action you have told your analytics to treat as significant: an enquiry form completed, a telephone call started, a booking made.

Marking too many renders the figure meaningless. An account marking twelve different actions as important, including somebody scrolling halfway down a page, produces a number that tells you nothing about your business. Marking two versions of the same action counts every enquiry twice.

We reduce this list to what genuinely constitutes an enquiry, so the figure in your report is a count of people who made contact.

Form tracking.

Form tracking means recording the moment somebody submits your enquiry form, so that the enquiry can be counted as a conversion.

A common shortcut counts arrivals on the thank-you page instead. That also counts page refreshes, bookmarks, and anybody who reaches the page by accident from a search, which means a month with forty genuine enquiries can be reported as sixty.

We bind the tracking to the submission itself rather than to the thank-you page, so the figure you are given is a count of real enquiries.

Call tracking.

Call tracking records the moment somebody taps your telephone number on your website, so that telephone enquiries can be counted alongside form enquiries.

It only works where your number is presented as a proper link. It covers people using a phone, and misses anybody who reads the number and dials it manually, so the figure should be treated as a minimum rather than a total.

If most of your enquiries arrive by telephone and this feature is not set up, your reporting shows only the enquiries that came through your form.

We make your number a link and set up the tracking behind it.

Double counting.

Double counting means two systems recording the same action, or one system installed twice, so that every enquiry is counted more than once.

We compare the recorded count against what actually reached your inbox, which is also the only way to identify enquiries that were counted and never delivered.

Form delivery.

Form delivery means the enquiry reaching a person once somebody has submitted the form on your website.

A form can submit correctly, record a conversion, and send the message nowhere at all: an old address, a full inbox, or a spam filter. For example, if the address a form delivers to belonged to somebody who left the business two years ago, every enquiry since then has been recorded in your reporting and read by nobody. The website shows a success message, so neither you nor the customer has any reason to suspect it.

Where your website stores submissions we compare those against what reached the inbox, rather than sending test enquiries into a live business.

Your website's own search box.

Where your website has a search box, what visitors type into it can be recorded.

We check whether the search box is being recorded, switch the recording on where it is not, and read what it collects when we plan content.

Form spam protection.

An enquiry form with no protection on it collects automated submissions, sometimes hundreds in a month.

The cost is not only the nuisance. Each one is counted as an enquiry in your reporting, so the figure you are given stops meaning anything, and genuine enquiries are lost among them.

We check what protection your form has and what is reaching your inbox. Adding protection is usually a setting rather than a rebuild, and we never submit a test enquiry to a live business in order to find out.

Cookie consent.

A cookie is a small file a website stores on a visitor's device, and most tracking relies on them. Where the law requires it, visitors are asked for permission before anything non-essential is stored.

Configured properly, your tracking adjusts to what each person has allowed. Configured poorly, it either records nothing, so your website appears to have no visitors at all, or it records everything regardless, which is the outcome it was installed to prevent.

We check how consent is configured and what it is doing to your figures.

Google Business Profile.

Enlarge

Your Google Business Profile is the listing that appears when somebody searches your business name, and the pin that appears on the map for searches nearby.

For a business with premises or a service area, the listing frequently produces more telephone calls than the website does, and how complete it is directly affects how often Google chooses to show it.

We complete and maintain it, covering your categories, services, opening hours, photographs and description.

The map pack.

Enlarge

The map pack is the map and the three businesses Google displays above the ordinary results for local searches.

We work on the listing, the reviews and the local details that decide those three positions.

NAP consistency.

NAP stands for name, address and phone number. Those three details appear on your website, on your Google Business Profile, and across dozens of directories.

Google checks them against one another, and where they disagree it has less confidence in all of them. For example, a builder who moved premises three years ago may still have the old address on four directories, which means Google cannot establish which is correct and your visibility in local searches is weaker as a result.

We collect every version we can find and correct them one by one. It is slow work, and local visibility depends on it.

Business listings & citations.

A citation is any mention of your business name, address and telephone number on another website: directories, social profiles, and listings you may not remember creating.

These accumulate over years, and most businesses have entries they never created, added by directories that copy company records. The old ones carry whatever your details were at the time.

We audit what exists and correct it, including the old entries carrying a previous address or a former business name.

Service areas.

Where you travel to your customers rather than working from a shop front, your Google Business Profile allows you to list the areas you cover.

Listing everywhere is tempting but works against you. An electrician who genuinely works across the southern suburbs but lists the whole of the Western Cape does not gain the province, and becomes weaker in the suburbs where the work actually comes from.

We set these to match the area you genuinely cover.

Opening hours.

Enlarge

Your opening hours appear directly in the search results, which means incorrect hours send customers to a closed door and can result in complaints and bad reviews.

Special hours covering public holidays and seasonal changes have to be set in advance. Where they are not, your standard hours are displayed instead.

We keep these current, including the holiday hours, so they are set before the date rather than after it.

Profile photographs.

Listings carrying photographs are viewed more often than those without, and Google specifically asks for interior and exterior shots of your premises.

These are also the images a prospective customer sees before they ever reach your website, which for many local businesses makes them the most-viewed photographs the business owns.

We upload the photographs you supply and keep the set current.

Profile posts.

Enlarge

A profile post is a short update published to your Google Business Profile: an offer, something seasonal, or a service you have added.

They reach people who have already found your listing, and to somebody comparing two businesses they show that yours is active. They expire after a set period, which makes this an ongoing routine rather than a task that is ever completed.

We may write and publish these for you when you have an offer, a seasonal change or a new service to announce.

Reviews & replies.

Enlarge

The number of reviews you hold, how recent they are, and the rating itself all feed into your visibility in local searches. A steady flow of reviews is worth more than a sudden burst.

Replying is equally important, to positive and negative reviews alike, because everybody reading your listing can see the reply and forms a view of the business from it.

Replies are usually handled inside the business. If nobody is currently doing this, please let us know and we can draft them for your approval.

Review requests.

A review request is asking a customer to leave a review at the point that they are most likely to agree, which is immediately after the job is complete and they are satisfied.

The effective method is a direct link opening the review box for them, because anything that requires them to search for your listing loses most of them. Anything that screens out unhappy customers before asking breaches Google's rules and should be avoided entirely.

Where you would like help with this, we can set up the request link and a routine for using it. A small, steady flow of reviews is worth considerably more than one large push once a year.

Multiple locations.

Each branch of your business needs its own Google Business Profile listing and usually its own page on the website, and each should describe that branch rather than repeating head office.

Where all the listings point at the same page, the branches compete with one another rather than adding together, and Google struggles to establish which one to show for a given search.

We set up and maintain a listing and a page for each branch.

Backlinks.

Enlarge

A backlink is a link to your website from somebody else's. Google treats it as a recommendation, because another website has judged your business worth pointing its own visitors towards. It remains one of the strongest factors in deciding where your website ranks.

Not every link is a good one. A link from an unrelated or low quality website adds nothing, and a large number of them can work against your website rather than for it.

Which websites link to your website is more significant than how many do. For example, a single link from a well-known local venue recommending a caterer is worth more than three hundred directory entries, because Google can tell the difference between a recommendation and a listing.

We monitor what is linking to your website and report on it each month. We also search for valuable link placement opportunities and arrange them where your package includes this service.

Spam & artificial links.

Websites sometimes acquire large numbers of links from automated networks, either purchased years ago or aimed at them by somebody else without their knowledge.

Where the pattern is clear we formally disown them so that they are disregarded.

Negative SEO.

Negative SEO is a deliberate attempt by somebody else to damage your position in the search results, rather than to improve their own.

It is far less common than it is claimed to be, and a sharp drop in your figures almost always has an ordinary cause. We monitor your links, your reviews and where your content appears, and if we do find a deliberate attack we will inform you and disown the links responsible.

Disavow files.

A disavow file is a list of websites, submitted to Google, whose links to your website should be disregarded. It is the remedy where automated networks are pointing at your website.

We compile the file, submit it, and add to it periodically where the unwanted links continue to arrive. Submitting it once and leaving it in place is not sufficient while the flow continues.

The legitimate links buried in the same list have to be excluded, which is the part that takes the time.

Link building.

Link building is the work of securing links to your website from other websites.

Links are earned by being worth citing, by being listed where your industry is listed, and through ordinary business activity such as sponsoring, supplying and partnering. Some links are also bought, where the placement is relevant and the website is one your customers might actually read.

We pursue the routes your budget covers. Where link work does not form part of what you are paying for, your link profile grows slowly, and your report will present it that way.

Directory listings.

A directory listing is an entry for your business on an industry or local directory website. These are the affordable end of off-site work.

The useful ones are those a customer might genuinely consult, and those already appearing in the search results for your services. Many of them accept a submission at no cost.

We identify the directories worth listing your business in and submit it to them. Every listing carries your business details, so this work is done after your business details are settled rather than before.

Domain rating.

Enlarge

Domain rating is a score out of a hundred that SEO tools assign to a website, estimating the strength of the links pointing at it. It is a tool's estimate rather than a Google measure.

We report the score because it moves over time, but we do not treat moving it as the objective.

Social profiles.

Your social profiles rarely move your rankings directly, but they are worth keeping accurate.

They appear when somebody searches your business name, occupying space in the results that a competitor would otherwise hold. They are also one of the ways search engines confirm which accounts genuinely belong to your business.

We ensure they are listed correctly and connected to your website.

AI search.

Enlarge

When somebody asks an AI assistant such as ChatGPT or Gemini to recommend a business like yours, it answers in a few sentences and quotes a handful of sources.

Being quoted works differently from ranking. It depends on the facts about your business being stated plainly, agreeing with one another everywhere they appear, and being traceable to a named person or organisation. Where the web contradicts itself about your business, assistants tend to leave you out of the answer rather than guess which version is correct.

The same applies to how Google itself understands your business, as opposed to your website. Where your name, address, ownership and description agree across your website, your listing, your social profiles and anywhere else you appear, Google is able to treat you as a known business rather than as a collection of pages, which is what produces the panel of information beside your name in the search results.

We work on the things that make a business quotable: consistent facts, named authors, and information stated plainly. We also check that the basic facts about your business agree wherever they appear, and correct them where they do not.

AI Overviews.

Enlarge

An AI Overview is the answer Google writes itself and places at the top of the results page, above the ordinary results.

Being cited in one places your business in front of somebody who may never scroll any further down the page. At the same time, these answers reduce clicks for everybody, because the searcher has what they needed without visiting any website. A first position on a search that carries an AI Overview above it is therefore worth fewer visits than it used to be.

We monitor whether your business is being cited in these answers and report that alongside your positions.

AI crawler access.

AI crawlers are the software AI companies use to read the web, separate from the software Google uses for its ordinary search results. Each can be permitted or blocked independently.

Blocking them keeps your content out of their training and rules out ever being cited in one of their answers. Permitting them does the opposite. Which is right depends on the kind of business you run. A business that sells its written material, such as a publisher or a training provider, may prefer to block them. A business that wants enquiries usually has more to gain from being cited.

We check what your website currently allows and set it to whatever you decide, so it is a decision rather than a default nobody has examined.

AI assistant files.

This is a newer convention: a plain file on your website listing the pages that best explain what your business does, written for AI assistants rather than for search engines.

Not every assistant supports it yet. Where one has been generated automatically it usually lists internal template fragments rather than actual content, which is of less use to an assistant than having no file at all.

We write this by hand rather than leaving it to be generated.

Share of voice.

Enlarge

Share of voice is a measure of how often your business is named by AI assistants compared with your competitors, across a set of questions a customer might realistically ask.

It is the closest equivalent to a ranking figure that this kind of search currently has. The figures move considerably from month to month, so we report it as a direction rather than as a score.

We track it monthly and use it to decide where the AI work is focused.

Product pages.

A product page is the page on your shop for one individual item.

Each one needs a description written for that specific product rather than copied from the manufacturer, because several thousand other websites carry that identical text and Google has no reason to prefer yours. It also needs the details people search for: size, material, finish, and what the item fits.

We work through them in batches, starting with the products that already attract the most attention. On most shops these pages are the largest untouched opportunity, because there are so many of them and they are rarely written with search in mind.

Category pages.

A category page is the page grouping a type of product together, such as all your mountain bike helmets.

Category pages usually stand a better chance of ranking than individual product pages, because they match the way people search. Far more people search "mountain bike helmets" than search for any one helmet by name. They are also usually a grid of products carrying no written content at all, which gives Google nothing to judge the page on.

We write the copy for the commercially important categories, explaining what the category covers and how a customer should choose between the options.

Product titles.

A product title is the name a product is listed under on your website.

A product named with an internal code, or with a name only your business uses, will not be found by anybody who does not already know it. For example, a furniture retailer selling a chair listed as "Model 42B" is found only by customers who already know that code. Listed as "Model 42B Oak Dining Chair", it is found by them and by everybody searching for an oak dining chair.

We propose a naming format that keeps the names you have chosen and adds a plain description of what the item is, then apply it across the catalogue.

Discontinued products.

A discontinued product that has accumulated links and search history should not simply disappear from your website.

We decide the treatment with you. Depending on whether the item is returning, the page either stays published and says it is unavailable, points at the closest replacement, or points at its category.

Faceted navigation.

Faceted navigation is the set of filters on a shop allowing somebody to narrow a category by size, colour or price. Each combination selected frequently creates its own web address.

We decide which combinations are worth indexing and block the rest.

Category structure.

Shops accumulate categories over time: near-duplicates created by different members of staff, empty ones left behind when a range was discontinued, and products filed under no category at all.

A product sitting in no category can only be found by somebody searching for it by name, which means it is invisible to anybody browsing. It is usually the largest single piece of work on a shop, and the rest of the structure depends on it being done.

We work through them, merging the duplicates and filing the products that belong to no category.

Product feeds.

A product feed is a structured list of your products, submitted to Google Merchant Center so that they are able to appear in shopping results. Merchant Center is a free Google account, separate from your website and from your analytics, and it is where Google keeps its record of what you sell.

It has its own rules, and products are rejected for reasons that never appear anywhere on your website: image requirements, missing categories, or a price that disagrees with the page. Nothing notifies you when a product is rejected, which means a shop can have half its range excluded without anybody noticing.

We check the account each month and clear the rejections, and we make sure the feed matches what your product pages actually show.

Data protection.

Wherever you trade there are rules governing the collection and storage of people's information, including whatever your contact form collects.

In practice this means a visitor should be told what happens to the details they submit, and should not be tracked without being asked first. Whether your policy is legally sufficient is a matter between you and your legal advice.

We check that your form states what the information is used for and that consent is handled honestly, and we correct the parts that sit on the website.

Regulated claims.

Financial services, health, property and several other sectors limit what may be claimed in writing, and a claim about returns, results or timescales can be a problem in any sector at all.

Pages published years ago frequently carry a guarantee, a percentage or a timescale that would not be permitted today, and nobody revisits them until somebody complains.

We check what is already published against the rules that apply to you and inform you of anything that would breach them. We hold ourselves to the same standard, which is why we never promise you a particular position in the search results.

Exposed files & customer data.

Websites sometimes leave files reachable that were never meant to be public: database backups, development files, or exports containing customer details.

Anything reachable can be found, stored and read. Where it contains customer information it is a legal matter before it is a search one.

We check for this while crawling your website. Where we find something, we tell you directly rather than writing it into a report.

Accessibility.

Accessibility means your website being usable by people with a visual, motor or cognitive impairment. It covers text descriptions on images, sufficient colour contrast to read comfortably, labels on form fields, and a website that can be operated using a keyboard alone.

It overlaps heavily with what search engines require in order to understand a page, which is why it appears in search work. It is also increasingly a legal expectation rather than only good practice.

We deal with the parts that overlap with the search work and inform you of the rest so your developer can address it.