Skip to content

Latest news

Eleutheros Ltd
City skyline and a lit bridge over a river under a dusk sky

Digital business trends

Ad attribution beyond the click: what the Instagram impression misses

Siècle Digital describes a customer who sees an Instagram ad, ignores it, then converts through Google Ads two days later. That gap between impressions and reported sales is where ad measurement budgets and channel credit get decided.

Siècle Digital published on 16 November an explainer on measuring what each advertising channel actually contributes, using a simple case: a customer sees an Instagram ad without clicking or buying, then searches for the product two days later, clicks a Google Ads ad and converts. The report's core question is why ad results do not map directly onto sales. That is a fair starting point for any team that splits budget across more than one channel.

The last click gets credit it did not earn

In the example, Google Ads receives the conversion. The Instagram impression, which may have created the initial interest, appears in no revenue column. Optimisation systems that learn from last-click signals will keep rewarding the closer channel and starving the opener, even when the opener does the persuasive work. For teams running both social and search, this can mean search appears efficient while social looks wasteful, and budgets shift accordingly.

The underlying mechanism is familiar to anyone who has managed paid media: platforms report on their own inventory, and cross-channel journeys are rarely reconstructed end to end. What the report does not specify is which measurement methods it examines in detail, nor how it weighs their relative accuracy.

Why the gap is hard to close

Several structural factors make the picture murky. Cookie restrictions and consent choices limit what can be tracked across sites. View-through windows are set by each platform, which has an interest in generous defaults. And people move between devices, so a phone impression and a laptop purchase may never be joined. General industry context: attribution modelling has long been an approximation, and no single method has settled the debate between click-based and modelled measurement.

What to check this week

Rather than chasing a perfect attribution model, a team can start by auditing assumptions.

  • Pull the conversion windows and view-through settings for each active channel and write them down side by side.
  • Compare reported conversions per channel against total sales for the same period; a sum that exceeds real orders signals overlap.
  • Run a short holdout or incrementality test on the channel whose contribution is most disputed, even at small scale.
  • Check whether the creative or audience that first touched converting users is funded at all, or only the closing channel.

The report raises the role of AI in this discussion, but does not detail which techniques it covers or what results they produce. Treat claims of solved attribution with caution: the honest position is that every method is a model, and the useful question is which decisions a given model is reliable enough to support. The article's value lies less in a verdict than in the reminder that reported performance and real contribution are two different numbers.

Also worth reading

All news

Aerial night view of Alexander Nevsky Cathedral in central Sofia

Contact

Tell us where the project stands.

A few lines are enough: what you are building, where it is today, and what is blocking it. We answer every message personally.

Contact us contact@eleutheros.eu