How we measure whether content creates business value

We do not treat traffic as the final result

An article doubles its monthly visits.

We follow the complete path from visibility to business action, then use the evidence to improve the content.

The increase may look successful in a report.

However, the new visitors may leave without reading another page. They may never reach a relevant service. They may submit enquiries that do not match what the company provides.

The traffic increased. The business result did not.

We therefore measure content as a connected chain:

1. Did the right people discover the page?
2. Did they choose to visit it?
3. Did the content help them understand the problem?
4. Did they continue to a relevant next step?
5. Did that activity contribute to a useful business outcome?
6. Was the resulting lead, enquiry, or action relevant?

A single metric cannot answer all six questions.

1. We agree on the intended business outcome

Before measuring a page, we define what useful performance should look like.

The expected outcome depends on the page and the reader's stage.

For one article, success may mean helping a new audience discover the company. For another, it may mean moving readers into a service page. A comparison article may support evaluation, while an implementation guide may help qualified prospects assess whether the company fits their requirements.

We define the intended outcomes before reviewing the data.

These may include:

• qualified traffic from relevant searches;
• visits to connected service or product pages;
• downloads or registrations;
• consultation or project enquiries;
• trial or demo requests;
• assisted conversions;
• improved lead quality;
• or stronger visibility for important technical questions.

We do not assign the same conversion expectation to every page.

An early-stage educational article and a commercial service page serve different purposes. We measure each one against the role it is intended to play.

2. We establish a reliable baseline

We record performance before making major content changes.

Without a baseline, we cannot separate the effect of our work from seasonal demand, paid campaigns, website migrations, tracking changes, product launches, or unrelated market activity.

We use a consistent comparison period and document anything that could affect the results.

At minimum, we record:

• impressions, clicks, and queries by page;
• average search position where it adds useful context;
• organic landing-page sessions;
• engagement with relevant pages;
• movement into service or product pages;
• agreed key actions and enquiries;
• pages that assist those actions;
• and lead quality where the commercial team can assess it.

Google's guide to using Search Console and Analytics together distinguishes the two systems clearly.

Search Console shows how a page performs in Google Search. Analytics shows what visitors do after entering the website.

We use both because visibility without on-site behaviour gives us an incomplete picture.

3. We measure relevant visibility

The first question is whether the page appears for searches connected to a real company capability.

We review:

• the queries generating impressions;
• the pages appearing for those queries;
• the countries and devices involved;
• changes in visibility over time;
• and whether the search intent matches the intended audience.

A rising impression count is not automatically positive.

A technical article may begin appearing for broad educational searches that have little connection to the company's market. That visibility can increase the graph without creating commercially relevant traffic.

We therefore ask:

• Are the queries connected to problems the company solves?
• Do they reflect the intended buyer or user?
• Is the page appearing for the correct subject?
• Are irrelevant interpretations beginning to dominate?
• Does the content need clearer positioning?

We value relevant visibility more than visibility alone.

4. We measure whether searchers select the page

Once a page appears in search, we review whether people choose it.

We examine clicks and click-through rate alongside:

• the search query;
• the page;
• the device;
• the country;
• the approximate search position;
• and the type of result shown.

We do not rely on one site-wide click-through rate. It combines too many different pages, queries, devices, and intentions.

A low rate may indicate that:

• the title does not match the question;
• the description is vague;
• the page appears for the wrong search intent;
• competing results provide a clearer promise;
• or the answer is being satisfied directly within the search results.

A higher rate is useful only when the selected page attracts the intended audience.

We improve titles and descriptions to clarify the answer, not to create exaggerated promises that the page cannot fulfil.

5. We evaluate whether the content is useful

A visit is not proof that the content helped the reader.

We assess whether visitors engage with the page and continue through the website in a way that fits the content's purpose.

Depending on the page, we may review:

• engaged sessions;
• time spent with the content;
• movement to related articles;
• visits to service or product pages;
• downloads;
• form starts;
• form completions;
• return visits;
• and relevant events defined for the website.

GA4 defines an engaged session as a session that lasts longer than ten seconds, contains a key event, or includes at least two page or screen views.

We treat that as an analytics definition, not as proof of persuasion.

A visitor can remain on a page for longer than ten seconds and still leave without understanding the answer.

We therefore interpret engagement beside the page's intended role, navigation path, and business actions.

6. We track movement toward a relevant next step

Useful content should help the reader make progress.

That progress may involve:

• understanding the problem more clearly;
• comparing available approaches;
• reviewing a relevant service;
• reading supporting evidence;
• examining a case study;
• requesting an assessment;
• or starting a project conversation.

We check whether the page provides a logical next step and whether readers follow it.

When readers leave without moving forward, we investigate the reason.

The problem may be:

• an unclear opening;
• weak evidence;
• poor internal linking;
• a call to action that appears too early;
• a commercial step that does not match the reader's question;
• or an article that attracts the wrong audience.

We improve the transition between information and action rather than adding a generic call to action to every page.

7. We measure business actions and their quality

We track whether content contributes to agreed outcomes such as:

• project enquiries;
• consultation bookings;
• demo requests;
• sign-ups;
• trials;
• downloads;
• or another meaningful commercial action.

The number of actions is only part of the result.

We also need to know whether those actions are relevant.

One enquiry from a company with a genuine technical problem may be more valuable than ten submissions from people outside the target audience.

Where possible, we work with sales or delivery teams to review:

• whether the enquiry matches the service;
• whether the organisation fits the intended market;
• whether the prospect understands the offer;
• whether the opportunity progresses;
• and which questions or pages influenced the conversation.

This feedback helps us evaluate content quality in a way that analytics alone cannot.

8. We do not force one-page attribution

Technical buying journeys rarely follow one direct path.

A reader may:

1. discover an educational article;
2. leave without taking an immediate action;
3. return through branded search;
4. read a service page or case study;
5. share the information internally;
6. and contact the company several days later.

The final enquiry may be credited to the last visit, even though the original article introduced the company and shaped the decision.

We therefore review:

• organic landing pages;
• assisted paths;
• return visits;
• movement between content and commercial pages;
• and available attribution reports.

We also recommend adding a simple question to sales or enquiry reviews:

Which page, topic, or question influenced this enquiry?

This does not create perfect attribution.

It adds useful human evidence to the behavioural data.

9. We measure AI visibility separately

Visibility in AI-generated answers is not the same as ranking in traditional search.

A brand mention is not automatically a website visit. A citation is not automatically a qualified lead. An accurate answer can support awareness without producing a measurable click.

We therefore measure AI visibility as a separate layer.

We maintain a defined set of important buyer questions and test them across relevant systems at consistent intervals.

We record:

• whether the company is mentioned;
• whether a company page is cited;
• which competing sources appear;
• whether the description is accurate;
• whether the cited page matches the question;
• and how the result changes over time.

We do not claim revenue impact from an AI mention unless we can identify a traceable path to the website or a commercial outcome.

AI visibility gives us evidence about discoverability and source authority. It does not replace traffic, conversion, or lead-quality measurement.

10. We turn every report into a decision

A useful content report should not end with a dashboard.

It should identify what we need to do next.

Depending on the evidence, we may decide to:

• strengthen a weak answer;
• improve the title or opening;
• add first-hand evidence;
• connect an isolated article;
• improve the next step;
• consolidate overlapping pages;
• update outdated information;
• redirect an irrelevant page;
• or stop investing in a topic that attracts the wrong audience.

We connect each recommendation to an observed weakness in the performance chain.

For example:

• Relevant impressions but few clicks may indicate weak search positioning.
• Clicks without useful engagement may indicate a mismatch between the search promise and the page.
• Strong reading without further movement may indicate poor internal linking or an unsuitable next action.
• Enquiries without commercial fit may indicate that the content attracts the wrong audience.
• High visibility without business influence may indicate a topic that is informative but commercially disconnected.

This approach helps us improve the system rather than reacting to isolated numbers.

Our measurement process is continuous

We measure content through a repeated cycle:

1. We define the intended audience and business outcome.
2. We establish the performance baseline.
3. We measure relevant search and AI visibility.
4. We review selection and on-site behaviour.
5. We track movement toward useful actions.
6. We assess conversions and lead quality.
7. We identify the weakest stage.
8. We improve the content and measure again.

This process shows us whether content is merely receiving attention or creating meaningful business value.

Return to why useful technical expertise remains invisible in search, or review the opportunities created when content and expertise align.