Flowsk vs GA4 for attribution
GA4 is free, powerful and answers a different question. It measures sessions and models what it lost; Flowsk records people and shows you the events. Here is where each one wins.
GA4 = sessions, sampled, modelled · Flowsk = people, complete, inspectable
Keep GA4 for behavioural analytics — funnels, content performance, audience shape. Use Flowsk when you need to know which campaign paid for a specific customer and be able to prove it. They are not substitutes; only one of them can answer the second question.
GA4 is free and excellent at what it does. It is simply not an attribution ledger, and was never designed to be.
GA4 is the default. It is free, it is everywhere, and roughly nobody chose it deliberately — it arrived when Universal Analytics was switched off.
The honest comparison is not “which is better.” It is “which question are you asking.”
The head-to-head
| GA4 | Flowsk Signals | |
|---|---|---|
| Unit of analysis | Session / event | Person |
| Who owns the data | You | |
| Identity durability | JS cookie — 7 days in Safari | Server-set — as long as you set it |
| Anonymous → email → purchase | Not stitched | Stitched, retroactively |
| Conversion counting | Modelled where unobserved | Server-confirmed, de-duplicated |
| Sampling / thresholding | Yes, on large queries | No |
| Inspect one conversion’s path | No | Yes, event by event |
| Client + server de-dup | Via GTM, you build it | Built in, on a shared key |
| Price | Free | $29 / mo |
| Learning curve | Steep | One report |
Where GA4 genuinely wins
Behavioural depth. Funnels, path exploration, content grouping, audience overlap, scroll depth. GA4 has a decade of product investment behind these and Flowsk has none of them. If you want to know which blog posts precede a signup, GA4 answers better.
Price and ubiquity. Free, and your next hire already knows it.
The Google ecosystem. Audience export to Google Ads, BigQuery streaming, Search Console integration. If your paid programme is Google-first, that loop is worth real money.
We ship the flowsk.com site with our own product and keep behavioural analytics separately. That is the recommended configuration, not a compromise.
Where GA4 cannot help you
It measures sessions, not people. A customer who researches anonymously for two weeks and then buys is, to GA4, a series of unrelated visits. There is no built-in mechanism that retroactively binds those anonymous sessions to the email that eventually arrives.
Its identity dies in seven days on Safari. The _ga cookie is written by JavaScript. Safari caps it. Every reporting window longer than a week is, for that population, decorative. Server-side GTM does not fix this: it improves delivery of the events, but the client id is still minted client-side.
It models what it lost, and does not tell you which rows are modelled. Conversions it could not observe are estimated and reported in the same column as ones it did. That is a defensible product decision and an impossible basis for reconciling against revenue.
You cannot audit a single conversion. There is no view in GA4 that shows the actual event sequence for the person who placed order 10482. When your CFO asks why Meta says 180 and the bank says 122, GA4 gives you a third number and no way to check any of them.
The practical split
- GA4: how is traffic behaving, which content works, which pages leak, what does the audience look like.
- Flowsk: which campaign acquired this customer, what did they do before buying, is this conversion confirmed by our own server, what is our real CAC.
Two tools, two questions, $29 between them. Anyone telling you to rip out GA4 is selling you something.
What actually changes when you add the second one
The observable difference in most accounts is not “more conversions” — it is the same conversions, credited differently. Prospecting stops looking 30–50% worse than it is. Email stops looking like a miracle. Direct traffic stops mysteriously scaling with ad spend.
And when someone challenges a number, you open the journey and read the timestamps out loud, which is a conversation that lasts ninety seconds instead of a quarter.
Frequently asked questions
Can GA4 do first-party attribution?
Not durably. GA4's client id is written by JavaScript into a first-party cookie, which Safari caps at 7 days. Server-side GTM helps with signal delivery but the identity still originates client-side, so the cap still applies.
What are GA4's modelled conversions?
Statistical estimates for conversions Google believes occurred but could not observe — declined consent, cross-device, blocked tags. They appear alongside observed conversions and cannot be separated out in most reports.
Is GA4 sampled?
Standard properties apply thresholds and sampling on large or complex queries, and apply data thresholds that suppress rows with small user counts. For high-level trends this is fine. For reconciling revenue it is not.
Should I remove GA4 if I use Flowsk?
No. Keep it. It is free, your team knows it, and it answers behavioural questions well. Just stop using it as the arbiter of which channel earned a sale.
What about GA4's data-driven attribution?
It is a reasonable model, computed on partial data, presented as fractional credit. Useful for directional budget shifts, impossible to reconcile against a bank statement.
Check what GA4 can actually hold on your site
The cookie inspector shows whether anything on your site sets a durable server-set identity, or whether every id is on Safari's 7-day clock.
Stop guessing which ad made the sale.
Flowsk Signals stitches the anonymous click to the email to the purchase — first-party, server-side, de-duplicated. One snippet, $29/mo, and every conversion comes with a receipt you can inspect.