Blog

A practical guide to attribution in a privacy-first world

A grounded take on attribution in a privacy-first world, written from real production work.

Attribution used to be a deterministic problem, which it is no longer. The interesting question is what to do about that. Below are the questions we hear most often, with the answers we give in real conversations.

Is server-side tagging worth it for a small site?

Below roughly 50,000 monthly events, the answer is "probably not, unless you have a specific reason." Compliance pressure or a brand that depends on one or two large advertising partners can flip that calculation quickly.

Will it speed up my pages?

Modestly. The wins on Core Web Vitals are real but not dramatic. The bigger improvement is in data quality, not in milliseconds.

Do I still need a client-side container?

Yes, almost always. The client-side container is what captures the events. The server-side container is where you decide what to do with them.

How does this affect Consent Mode v2?

It works well together, provided you propagate the consent state to your server container as part of the request payload. We have a separate post on that.

What about Meta and TikTok CAPI?

Both are exactly the kind of workload sGTM was built for. Send them server-side and you will see a meaningful lift in matched conversions.

What does this cost to run?

On SprTags, plans start at $5 a month. If you self-host on GCP, expect more, mostly because someone on your team will be paid to keep it healthy.

In 2026, the teams making good calls on spend are the ones who stopped looking for one true attribution model and started looking at three converging ones.