Growth Analysis · Confidential

The pricing page is killing us — and our buyers are telling us exactly why

Source: survey-responses.csv (n=400) + 12-mo funnel Owner: Growth Confidence: High Status: Action recommended

The single biggest conversion problem is pricing opacity & uncertainty at the activation→paid step. People who try the product can't predict what they'll pay, can't tell which plan fits them, and hit the paywall before they've felt value — so they bounce on pricing and never come back.

This is not a top-of-funnel awareness problem and not a “product is bad” problem. The leak is concentrated at one place where intent is already high.

FUNNEL

Where the money actually leaks

Every step loses people, but only one stage combines a steep drop with a clear, repeated reason in the survey.

Visitors

41,800

→ 13.5% sign up

Signups

5,640

→ 38.7% activate

Activated

2,180

→ 31.6% pay

Paid

690

$ revenue

68.4% of activated users — ~1,490 engaged people every 12 months — fail to convert to paid. Separately, the pricing page bounces 62% of everyone who lands on it. These two facts are the same story: the place people decide on price is also the place they decide to leave.

The ONE problem, with evidence

Themes hand-coded from the 387 non-blank "what_almost_stopped_you" free-text answers.

"What almost stopped you?" — themed (n=387)

Pricing unclear / couldn't predict cost142 · 37%

Wrong plan / which tier do I need?71 · 18%

Hit paywall before seeing value58 · 15%

Setup / onboarding friction44 · 11%

Missing integration / feature39 · 10%

Trust / security / "need to ask my team"33 · 9%

Why it's the #1 problem, not just #1 count

The top three themes — unclear pricing (37%), wrong-plan confusion (18%), and paywall-before-value (15%) — are the same root cause: people can't connect what they'll get to what they'll pay. Together that's 271 of 387 answers (70%).

This is corroborated by behavior, not just opinion: the 62% pricing-page bounce and the 31.6% activated→paid rate point to the identical decision moment.

NPS confirms it's a packaging problem, not a product one: detractors (NPS ≤6) who also cited pricing skew toward "loved the product but…" language — high usage, blocked at checkout.

"I genuinely could not tell what it would cost my team. The pricing page lists tiers but everything useful is 'contact sales' — so I left." R-118 · Ops Lead · 51–200 employees · NPS 4 · usage: high

"Couldn't figure out if I needed Pro or Business. The feature differences read like a legal doc. I didn't want to guess and overpay." R-263 · PM · 11–50 employees · NPS 6 · top_request: "plan comparison that makes sense"

"It asked me for a credit card before I'd done anything real. I'd happily pay once I see it works — not before." R-051 · Founder · 1–10 employees · NPS 7 · usage: medium

"Per-seat got expensive fast and I had no way to model it for 30 people. Bounced to compare alternatives." R-329 · Eng Manager · 201–500 employees · NPS 5 · usage: high

Supporting findings

Three patterns that reinforce the diagnosis and shape the fix.

A Intent is high; the blocker is downstream. Among respondents who didn't convert, the most common top_request themes were "transparent / self-serve pricing" and "let me try before paying" — not new features. People reached the buying decision and were stopped at it. ~59% of non-converters who left a top_request asked for pricing clarity or a real trial.
B The pain scales with company_size. Pricing-uncertainty mentions rise from ~28% of 1–10-person companies to ~48% of 51–500-person companies, where per-seat math and approval chains compound. These are our highest-LTV accounts — and the ones we're confusing most. A pricing/cost estimator targets the most valuable segment.
C "found_us" rules out an acquisition fix. Converters and non-converters arrive through the same channels (search, referral, content) in similar proportions — so the problem isn't low-quality traffic. We're losing people we already won. Fixing pricing clarity compounds across every channel.

The experiment to fix it

One falsifiable test. If it doesn't move the metric, we kill it and move on.

Hypothesis

Activated users churn at the paywall because they can't predict their cost or pick a plan. If we replace the static pricing page with a transparent, self-serve pricing page + an interactive seat/usage cost estimator + a "no card needed" trial gate, then the activated→paid rate will rise because we remove the uncertainty that causes the bounce.

Exact change

On the pricing page (the 62%-bounce page), ship three changes as one variant: 1 · Inline cost estimator: user enters # of seats / usage → live monthly & annual total, no "contact sales" for SMB tiers.
2 · Plan recommender: 3-question "which plan fits you?" widget that highlights one tier and explains the diff in plain language.
3 · Remove the card-up-front gate for activated users; show a "you've used X, upgrade to keep going" in-product moment after value, not before. Run as a 50/50 A/B on all pricing-page traffic + activated users. Min 4 weeks or ~2× detectable-effect sample.

Primary metric

Activated → Paid conversion rate
Baseline 31.6%. Success = a statistically significant lift to ≥ 38% (≈ +20% relative) at 95% confidence. +6.4 pts on ~2,180 activated ≈ +140 paying customers / yr from the same traffic.

Guardrails

Pricing-page bounce (target: drop from 62%), and net new MRR / ARPA — to confirm we lifted conversion without training people into a cheaper plan. If ARPA falls enough to erase the conversion gain, the test fails.

Kill criterion

No significant lift in activated→paid after full sample, or a meaningful drop in ARPA/MRR. Either result falsifies the hypothesis — we revert and re-theme the next-largest cluster (onboarding friction, 11%).