Content agencies measurement mistakes buyers should check before signing

What to take away

Reporting reach as if it were engagement

Impressions and sessions measure delivery, not effect. A reader who leaves in four seconds counts the same as one who reads to the end.

Ask the agency to separate distribution metrics from outcome metrics. If they cannot name which figure answers which client question, the reporting is decorative. Ask for the raw data export behind each chart rather than a screenshot of a dashboard.

The wider content agencies measurement and reporting guide 2027 sets out how to structure that split across a year.

Ignoring how search engines actually crawl

Some agencies promise indexing without describing the mechanism. Google publishes its crawler documentation, and Googlebot is only one of several user agents that may fetch a page.

Robots directives and JavaScript rendering both affect what gets indexed, and neither appears in a traffic report.

A useful test is whether the agency can explain which user agent fetches which asset. The Google crawler user agent overview is the reference to work from.

Treating accessibility as optional

The W3C publishes accessibility standards that shape inclusive content and are referenced in UK public sector duties. Content that fails them is harder to measure, because assistive technology users behave differently in analytics.

Ask for a sample page audited against those standards, not a policy statement. Build accessibility checks into the measurement plan rather than a separate audit. The W3C accessibility standards overview explains the conformance levels a brief can reference.

Using tracking that consent rules do not allow

Analytics and marketing pixels that rely on cookies need a lawful basis. The ICO publishes guidance on electronic marketing, consent and cookies under PECR. A dashboard built on data collected without valid consent is not evidence you can rely on.

Ask which tools fire before consent and what happens to that data. A privacy notice on its own does not make a marketing pixel lawful. The ICO guidance on direct marketing and privacy is the starting point.

Building dashboards nobody opens

A dashboard that restates platform numbers adds no decisions. Good reporting answers a question the client already has, then shows the next action. Three views usually cover acquisition, engagement and outcomes, plus a short note on what changed.

The content agencies reporting dashboard in England covers how to choose views people actually use.

Comparing the mistakes

Each row shows the symptom, the evidence to request and the fix. None of these need a new platform to resolve.

What to check before signing

Ask for one worked example of a past client's reporting pack, redacted if needed, and trace each figure to its source.

What to check in month three

Review whether the dashboard has changed any decision. If not, the measurement plan needs rewriting rather than more data.

Common questions

Do these mistakes apply outside England?

The ICO and PECR framework covers the whole UK, so the consent and cookie points apply in Scotland, Wales and Northern Ireland too. Public sector accessibility duties differ by nation, so check devolved rules.

Should I pay more for an agency with better measurement?

Not automatically. Compare what each supplier will report and how they will evidence it. A higher fee is only justified where reporting changes decisions.

How many metrics should a monthly report carry?

Few enough that someone reads them. For example, a team paying £400 a month for reporting might track three outcome measures and one distribution measure.

Can we fix these mistakes without changing agency?

Often yes. Ask for the evidence in the table and set a review date. If the agency cannot supply traceable sources after that, the brief is the problem.