Measuring Whether Content Marketing Is Working

Which signals indicate a content programme is compounding, how long to wait before judging, and the metrics that mislead.

Content programmes are usually abandoned at the point where they would have started working, because the measurement window was set by a quarterly reporting cycle rather than by how the mechanism actually behaves. Knowing what to look at, and when, is what allows a company to persist through the period where the numbers are unimpressive and the trajectory is correct.

The first thing to understand is the shape of the curve. A new page is crawled within days, evaluated over weeks, and positioned over months, and the compounding effect of topical depth and internal linking arrives later still. A programme judged at ninety days is being judged during its slowest phase. Six to twelve months is the honest window for a content library, which is uncomfortable and true.

The leading indicator that arrives first is impressions rather than traffic. A page that is appearing in results but not being clicked is being found and not chosen, which is a title and description problem rather than a content problem, and it is cheap to fix. A page with no impressions at all has a different problem: it is not being found for anything, which usually means the topic was wrong or the page is too thin.

The second indicator is the breadth of queries a page appears for rather than the volume of any one. A page ranking for a single phrase is fragile; the same page appearing for thirty related phrasings is demonstrating that the subject has been covered properly. Ekström et al. (2024) demonstrated that the way users phrase a query materially changes the results they see, which is why breadth of matching is a better health signal than position on one term.

The third is what happens after arrival. A visitor who reads one page and leaves has been served an answer, which is a legitimate outcome. A visitor who moves to a second and third page is in the process the library was built for. Rising pages per session across a content library is one of the clearest signs that the structure is working rather than the pages merely existing.

Mladenović et al. (2023), examining determinants of online search visibility, found that visibility depends on the combination of on page factors and content relevance rather than on any single lever, which has a measurement implication: a programme should be judged as a set rather than page by page. Individual pages will vary considerably, and the meaningful question is whether the library as a whole is gaining ground.

The metrics that mislead are worth naming. Total traffic is inflated by irrelevant visitors and by branded search that would have arrived anyway. Average position across all queries is close to meaningless because it mixes valuable and worthless terms. Time on page is unreliable for pages that answer a question quickly, which are precisely the pages doing their job.

The commercial signals sit outside the analytics and are the ones that should decide whether the programme continues. Enquiry volume, and more importantly enquiry quality, meaning whether prospects arrive already understanding the service. Whether the sales team notices better informed conversations. Whether the company is being found for the specific things it wants to be known for. These are answerable by asking people rather than by opening a dashboard.

For video specifically, the platform behaviour adds a further layer. Kim et al. (2025), studying playback interactions and engagement in mobile video viewing, found that behaviour during playback carries information that aggregate counts obscure. A video page whose visitors consistently watch most of the film is doing something a page with more traffic and no retention is not.

The practical reporting arrangement is a short quarterly review looking at four things: impressions and query breadth across the library, pages per session, enquiry quality as described by the sales team, and which specific pages are producing enquiries. Anything else is detail. A programme that improves on those four is working, and one that does not after twelve months has a content problem rather than a patience problem, at which point the honest response is to change the approach rather than continue.

References

Ekström, A. G., Madison, G., Olsson, E. J., & Tsapos, M. (2024). The search query filter bubble: Effect of user ideology on political leaning of search results through query selection. Information, Communication & Society, 27(5), 878–894. https://doi.org/10.1080/1369118X.2023.2230242

Mladenović, D., Rajapakse, A., Kožulјević, N., & Shukla, Y. (2023). Search engine optimization (SEO) for digital marketers: Exploring determinants of online search visibility for blood bank service. Online Information Review, 47(4), 661–679. https://doi.org/10.1108/OIR-05-2022-0276

Kim, E., Oh, S., & Park, S. (2025). An empirical study of user playback interactions and engagement in mobile video viewing. IEEE Access, 13, 78272–78289. https://doi.org/10.1109/ACCESS.2025.3566402