ananas.team
§ 07·ARTICLE
TrackingThe ananas.team crew··

UTM Tagging for Fintech and Apps: One Standard That Will Save Your Analytics

The most underrated job in performance marketing. Not setting up CAPI, not server-side GTM — plain old discipline in your UTM tags. Because any analytics you build on top of it is meaningless if the sources coming in are a mess.

Sounds boring. But I have seen projects where the source report is 200 spellings of five channels, and there is physically no way to reconcile them.

Where the chaos comes from

Tags get added by different people at different times with nothing agreed in advance. One writes Facebook, another facebook, a third fb, a fourth FB_Ads. To GA4 those are four different sources. The same Facebook is smeared across four rows of the report, and you cannot honestly say how much it brought in.

Multiply that by every channel, every campaign and everyone who built a link over the past year, and you get a mush that cannot be analyzed. And no amount of server-side will save you, because the garbage came in at the entrance.

One naming convention — three rules

Everything in lowercase. No spaces — underscores or hyphens only, and consistently one or the other. And most importantly, a dictionary of fixed values.

The dictionary is a document where it is written down once and for all: source for Facebook is always facebook, medium for paid is always cpc, and so on for every channel. One source, one spelling. Forever. Anyone building a link checks against the dictionary instead of inventing something.

This is not bureaucracy for its own sake. It is the difference between a report you can make decisions from and a mush.

utm_content and utm_term — for testing

Plenty of people ignore these two, and they shouldn't. utm_content encodes the creative, utm_term the audience or the keyword. With them you can break performance down to the level of a specific ad: you see not “Facebook brought conversions” but “this creative on this audience did.”

Without content and term you are analyzing at campaign level, and that is blindness. The most important creative decisions get made at exactly that level of detail.

Tying it to the app

If you have an app, your web UTMs need to get along with your MMP parameters. Pass the utm values into OneLink (or its equivalent) so that web and app are counted in one coordinate system rather than in two parallel universes. Otherwise your web analytics say one thing, your app analytics another, and there is no stitching them together.

Template and automation

Don't let people assemble tags by hand — they will make mistakes, guaranteed. Build a generator: a simple spreadsheet where dropdowns (values straight from the dictionary) assemble a finished, tagged link. Nobody types a source by hand; they pick it from a list. Getting it wrong becomes impossible.

Plus validation before launch: a quick check that the tags match the dictionary, before the link goes out into a campaign.

One last take: UTM discipline is the cheapest and most ignored lever in analytics. It costs zero and requires nothing more than agreeing on a convention and keeping a dictionary. And it decides whether all your expensive CAPI and sGTM work will collect meaningful data or beautiful garbage. Start putting your house in order here, not with the fashionable server-side stuff.

§ 08·RELATED