Most analytics setups answer the cheap question and dodge the expensive one. They'll show you traffic sources, page views, and maybe signups. But that's not the same as knowing which channel actually produces paying customers.
I was reviewing some campaign notes on a Sunday morning with coffee, and the pattern was painfully obvious. Traffic lives in GA4 or Plausible. Payments live in Stripe. SEO data is in Search Console. Campaign context is in a spreadsheet with vague notes like "posted in niche subreddit" or "newsletter swap."
None of those tools are bad on their own. The real cost is in the gaps between them.
A traffic-first setup starts with sessions, referrers, UTMs, and signup events. That works when the question is "did this launch get eyeballs?" It fails when the question is "should we do this again next month?"
A revenue-first setup starts from the paid event and works backward. Which customer paid? What was their first touch? Last touch before purchase? Which landing page? Was there a UTM? Did they come from Reddit, SEO, paid ads, or something untagged? Did they visit once and buy, or come back three times over 21 days?
That changes how you judge channels. A Reddit post with 80 visits can beat a paid ad campaign with 900 visits if the Reddit traffic drives two purchases and the ad drives zero. An SEO product page with modest traffic can outperform a viral blog post if it converts to sales. Paid ads shouldn't survive just because CPC looks good if the path from click to purchase is weak.
I've seen too many Shopify stores optimising for traffic spikes while their cash flow flatlines. The simple method that works before adding more tools: for the next 30 days, tag every intentional acquisition experiment with clean UTMs, keep one notes column for context, and review only paying customers every Friday. Not all signups. Start with orders. Then trace each one back to first source, last source, landing page, campaign, and time to conversion.
If you can't connect those fields, that's the real analytics gap. Not another dashboard chart.
Attribution will never be perfect. B2C buying cycles are messy too. People switch devices, cookies get blocked, they find you through Reddit then Google you later then pay after an abandoned cart email. Small samples lie badly, especially under ten paying customers per channel.
But imperfect revenue attribution is still better than optimising for pageviews, upvotes, clicks, or signup spikes. The directionally useful question is not "which channel brought the most people?" It's "which channel brought people who eventually became worth something?"
Curious how others here are tying acquisition experiments to actual revenue. Are you doing this in GA4, Stripe exports, a warehouse, spreadsheets, or something else? Honest question.
