The Only 8 Marketing Metrics a SaaS Founder Needs
Skip the 40-widget dashboard. Eight metrics — from visitor-to-signup rate to channel-attributed MRR — that actually tell you whether marketing is working.
Early-stage SaaS dashboards fail in one of two directions: forty widgets nobody reads, or three vanity numbers that only go up. Both are ways of avoiding the question marketing exists to answer — are we acquiring customers we keep, at a cost we can afford, from channels we can scale? Eight metrics answer it. Everything else is garnish.
The acquisition four
1. Unique visitors — by channel, never in total
Total traffic is weather. Traffic per channel is information: organic growing 15% while paid shrinks tells you something total traffic hides completely. This requires honest channel data — disciplined UTMs and a tracker that is not losing 40% of visitors to a consent banner.
2. Visitor → signup rate
The website's job, expressed as one number. Healthy SaaS landing pages convert 2–5% of visitors to trials or signups. Below 1%, fix the page before buying more traffic — pouring water into a leaking bucket is a strategy only your ad vendor endorses.
3. Signup → activation rate
Define the moment a user first gets real value (first project created, first report run) and measure how many signups reach it. This is where most funnels actually die — and a one-line track('activated') call makes it visible per cohort and per channel.
4. Channel-attributed MRR
The king. Not 'which channel brings traffic' but 'which channel brings money'. Requires the full join — first-touch UTM, identify() at signup, revenue event from your billing webhook — described in Revenue Attribution for SaaS. Once it works, budget meetings get shorter: channels argue for themselves in euros.
The economics two
5. CAC payback, per channel
Months of a customer's revenue needed to repay their acquisition cost. Channel-level payback under 12 months is generally healthy for self-serve; over 24 means you are buying growth you may not survive. Blended CAC hides the one channel quietly bleeding you — always split.
6. Net revenue retention (cohort revenue over time)
Do customers from March spend more or less today than they did in March? Expansion and churn in one number, and the single best indicator of whether marketing is acquiring the right customers rather than just customers. Channels with great volume and terrible retention are expensive noise — visible only when revenue events carry their acquisition context.
The experience two
7. p75 page speed on money pages
Landing pages, pricing, signup. Slow money pages tax every metric above them — the conversion cost is measurable — and the regression always ships on a Friday. One number, watched weekly.
8. Funnel-step error rate
JavaScript errors on signup and checkout flows, counted as a rate per session. A broken button in one browser is mathematically indistinguishable from 'mysterious conversion drop' unless errors live in your analytics. Pair this with metric 3 and most 'why did activation dip?' mysteries solve themselves.
Why these eight work as a system
They chain: traffic (1) × conversion (2) × activation (3) → revenue (4), priced by (5), validated by (6), and protected by (7) and (8). A change in any one points directly at its neighbors — which is what a metric system is for. And they all hang off one technical spine: every signal on the same visitor record, from first pageview to paid invoice. That spine is precisely what Clycyo provides with one script tag, an identify() call, and a webhook — see it running on our own traffic at /open.
Review the eight weekly, in the same order, and delete every widget that does not change a decision. Your dashboard should fit on one screen and provoke one argument per week. That is what working analytics feels like.