Attribution for media and content apps

The product is a habit, so measure the habit

A content app has no checkout on day one. A reader who installs, reads one story and never returns is worth almost nothing, whatever the install cost. A reader who opens the app most mornings is worth a great deal, through ad impressions, a subscription or both. That makes retention the first thing to read for every campaign, ahead of any revenue figure.

Look at it by source. Two campaigns with the same cost per install can differ widely in how many of their readers are still active after a week or a month. The one that keeps readers is the one to fund.

Three engagement events worth tracking

App opens are counted automatically. Add a small number of events that describe consumption, and keep their definitions strict.

Content started

Fired when an article, video or episode opens. Include a category such as "sport" or "podcast", so you can see what each campaign's audience came for.

Content completed

Fired when the reader reaches the end, or a set share of a video or audio item. This separates people who skim headlines from people who consume.

A commitment action

Following a topic, saving an item, turning on notifications or creating an account. These are stronger signs of a returning reader than a single view.

High-volume apps should think about event counts before adding more. Some attribution plans bill by event volume. Airbridge Core, for example, counts each collected event as a data point.^[1] On such a plan a reader who views thirty items a day produces thirty events. The data point entry explains the unit. Tracking completions and commitments, and skipping every scroll and tap, keeps the bill and the reports manageable.

Where revenue fits

How you attach money depends on the model.

• Ad-funded: revenue per reader is small and spread over many sessions. You need ad revenue passed into the attribution tool per user, or at least an average value per session, to compare campaigns.
• Subscription: treat the paywall like any subscription product. The guide to attribution for subscription apps covers trials, renewals and cancellations.
• Both: add the two streams per reader. A free reader who never subscribes can still pay back an install through ads.

Whichever applies, revenue builds slowly, so read it per install-week group over months and compare it with spend as ROAS at matching ages.

Shared links are an acquisition channel

Part of a content app's audience does not come from ads. People arrive from a link a friend sent, a social post, a newsletter or a search result. If those links are plain web URLs, the resulting installs are all labelled organic and you learn nothing about which stories or partners bring readers.

Put a tracking link behind every share button and every newsletter story. The link should open the same item in the app, and for a new reader it should still open that item after the install, which is what a deferred deep link does. Airbridge Core includes deferred deep links and a Retention report in the plan.^[1] Push notifications and email digests that bring existing readers back are measured as re-engagement, a topic covered in re-engagement attribution.

What to do this month

Define "content completed" in one sentence and have it implemented. Add tracking links to in-app sharing. Then build one report: retention at one week and one month by campaign, with completed items per reader beside it. Run it for a month before changing budgets, so each campaign's readers have had time to show whether they stay.