Should I optimize my conversion funnel or my retention first when both look weak?
The short answer
Retention almost always comes first, because a better funnel just pushes more people into a bucket that is leaking. If people who reach your core value do not come back, fixing that changes the entire unit economics and makes every acquisition and conversion improvement worth more later. The exception is a funnel so broken that not enough people ever reach value to judge retention, then unblock the funnel just enough to get a real cohort, and go back to retention.
Go deeper, your way
3 hand-picked resources, 3 link-checked. Pick how you want to dig in.
✍️ Essay
✓ Link checkedFreeIntermediate
Why we picked it
This is the clearest statement of the argument you need before you touch either lever: retention is not one metric among many, it is the base that every other metric stands on. Balfour's line says it plainly, every improvement you make to retention also improves virality, LTV, and payback period, so it is the foundation of all growth. If retention is genuinely broken, fixing the funnel just pours more people into a container that cannot hold them.
Retention compounds every other number: better retention lifts virality, lifetime value, and payback period all at once, which no single funnel fix can do.
When both look weak, weak retention is usually the deeper problem, because a leaky product caps the return on any acquisition or conversion you win.
Retention is an output of good activation, so the practical first move is often making users reach value fast, not chasing more top-of-funnel.
Why we picked it
This gives you the mental picture and the evidence in one place: a bucket full of holes barely fills no matter how fast you pour, which is exactly what a weak funnel plus weak retention feels like. It is backed by a benchmark study of 2,600 plus companies showing acquisition skill and retention skill do not correlate, so being good at one tells you nothing about the other. That is the honest diagnostic for deciding where your real leak is before you spend on either side.
Pouring more users in through acquisition does nothing if retention leaks them out, so measure the hole before you widen the tap.
Across 2,600 plus companies, how well you acquire and how well you keep users show no meaningful correlation, so treat them as separate problems.
Fast time to value in onboarding is the strongest predictor: 69 percent of top week-one activation performers were also top three-month retention performers.
Why we picked it
You cannot cut the right step until you know which step is the real cliff, and a funnel chart shows you exactly that. Amplitude lets you lay out your onboarding as a sequence of events and see the percentage who fall off at each one, so you fix the actual drop rather than the step that merely annoys you. The free tier is enough for an early-stage product to instrument this.
