TellTale

The links you sent aren't the links that arrived.

TellTale reads the UTMs sitting in your analytics right now and tells you what's broken. Worst first.

A real example

One link. Four things wrong with it.

None of it looks wrong at a glance. That's the problem.

https://shop.example.com/spring-sale?utm_source=facebok1&utm_medium=cpc paid2&utm_campaign={{campaign.name}}3&utm_cntent4=carousel_3
  1. 1utm_sourceP1 High

    One letter out, so it splits off from facebook and becomes a second source. Your spend now reports across two rows that never add up.

  2. 2utm_mediumP1 High

    Not a medium any analytics tool recognises. The session falls out of paid search and lands in (other).

  3. 3utm_campaignP0 Critical

    The ad platform never filled the template in. Every click on this ad reports the placeholder instead of a campaign name.

  4. 4utm_contentP2 Medium

    Misspelled as utm_cntent, so creative-level reporting for this campaign comes back empty and nothing warns you.

A person typed three of these. The fourth was the ad platform's doing, long after the link left anyone's hands. A builder can catch the first three, assuming everybody uses it. It can't catch the fourth at all.

How it works

  1. 1Connect your analytics

    Point TellTale at the property you spend money against. It reads landing-page URLs and session counts. Nothing else.

    GA4 and BigQuery connections aren't finished yet. New accounts get sample data until they are, and we label it as sample data.

  2. 2We read what arrived

    Your landing URLs get grouped into the distinct parameter combinations actually in use, with a few real examples kept against each one. Individual sessions are never stored.

  3. 3You get a score and a queue

    One number per property, a breakdown of which channel is dragging it down, and a list of what to fix. Every item opens onto the evidence and a corrected URL you can copy.

What we look for

Four kinds of broken.

12 checks in total. 6 of them run from the day you sign up, without anyone configuring anything. We own the checks and version them, so a result from last month still makes sense after we tighten a rule. Severities and approved values are yours to edit.

Missing or untagged

Required parameters missing, recommended ones skipped, and paid clicks that turn up carrying no UTMs whatsoever.

Untagged paid traffic is money you can't attribute to anything. Nothing that only inspects your builder output will ever see it.

presence.requiredpresence.recommendedcoverage.paid-no-tagging

Values nobody agreed on

Sources and mediums outside your approved list, or off whatever naming convention your team settled on.

It ships with a working list, so a free account catches facebok on day one without anyone configuring it.

values.approvedvalues.pattern

Broken before it landed

Ad-platform templates that never filled in, plus the casing, spaces and stray characters that quietly split one campaign into four.

A link can leave your builder perfectly formed and still show up with {{campaign.name}} sitting in it.

format.unresolved-macroformat.lowercaseformat.no-spacesformat.no-special-chars

Data that never arrived

A property set up to throw query strings away, and vendor parameters that disagree with the UTMs next to them.

If your analytics strips query parameters you have no UTM data at all, and nothing else is going to tell you.

source.query-params-strippedattribution.vendor-matchattribution.disabled-vendor-params

Why a builder isn't enough

A UTM builder can't tell you any of this.

Builders lock links down as people create them. That's useful, and it's a different job from this one. It assumes the link you generated is the link that arrived, and that everyone who ever tagged a campaign went through the builder.

Neither stays true for long. Ad platforms rewrite parameters. Templates don't always resolve. An agency sends a link nobody on your team ever saw. The builder stopped watching at creation, so none of that gets back to it.

TellTale looks at the other end, in your analytics, against traffic you actually paid for. If you already run a builder, keep it. This covers the half it can't.

Questions

How is this different from a UTM builder?
A builder governs links as they're created. TellTale inspects the UTMs that reached your analytics, which is the only place you can see what the ad platform did to a link afterwards, or what got tagged by someone who never opened the builder. They do different jobs. If you already run a builder, keep it.
Do I need a BigQuery export?
No, but it's the better source. BigQuery gives you unsampled data and full landing-page URLs. The GA4 API is lower fidelity, and we say so in the product: attribution-mismatch checks only work on BigQuery, and GA4's own (other) bucket hides exactly the malformed long tail this is built to find.
What data do you access?
Landing-page URLs, source, medium, campaign, session counts and dates, for whichever properties you connect. We keep the distinct tag shapes and a handful of real example URLs against each, not individual sessions. Read-only throughout. TellTale never writes anything back to your analytics.
What does it cost?
Nothing during early access. There are three plans defined and no way to pay for one yet, so everybody runs free while we finish the connections. We'll publish pricing well before anyone gets asked for a card.
Is it actually ready?
Partly, and we'd rather say so than find out you're annoyed later. The checking engine, the scoring, issue tracking and the evidence behind each issue are all built and tested. The analytics connections aren't finished, so an account you open today shows sample data, clearly marked. If you want to be on it the week connections land, start one now.

See what your tagging actually looks like.

Start an account now, or wait until connections land. The tagging problem is there either way.

Start free