How do you localize the stall?

Four questions, answerable from data you already have. Is new pipeline growing? — if pipeline itself is shrinking, the stall is upstream: positioning, demand, sales; product work won't move it this year. Do trials/new customers reach value? — healthy pipeline with weak activation is a product stall at the first mile. Do customers stay and expand? — flat NRR under healthy acquisition is a product stall at the value core. And does shipping correlate with metric movement? — steady releases with flat metrics is the signature of a broken learning loop: the team ships, nobody validates, the roadmap runs on opinion. My Growth Ceiling calculator runs exactly this localization with your numbers — ten minutes, free, and it tells you which of these four you're in.

What makes it a leadership stall rather than a product stall?

A product stall says what's broken; a leadership stall explains why it stays broken. The tells: the team knows activation is weak but the roadmap still ships features (evidence exists, doesn't drive decisions); the same fix gets proposed every quarter and dies in prioritization (no one owns the call); discovery stopped months ago (nobody's accountable for learning); and the founder agrees with all of this in every 1:1 while nothing changes (agreement without decision structure). If those sound familiar, adding another PM adds throughput to a system that can't steer — the six-signal checklist in the first-product-leader guide is your next read.

What does the fix sequence look like?

The wrong sequence — the common one — is to hire first: a growth PM, an agency, a VP. The working sequence: localize (the four questions above, one week), then decide ownership (who owns the stalled metric end-to-end — if the answer is "well, sort of the founder," that's the first fix), then run the loop on one metric — interviews, one thesis, one shippable bet, measured — before scaling the motion to the roadmap. This is precisely the arc of my 90-day engagements: diagnosis and interview engine first, segment/thesis decision second, funnel execution third. Not because the framework is clever — because the sequence is the only one where each step's output feeds the next.

What are stalls NOT caused by?

Worth naming, because these absorb quarters: not enough features (feature velocity is the most common false cure — stalled companies usually ship plenty), pricing alone (repricing a product with flat retention re-prices churn; fix stickiness first, then pricing — in that order), and effort (stalled teams are usually working hard; the constraint is decision quality, which more hours make worse, not better).

A stall is rarely a speed problem. It's a steering problem that looks like one.