Synthesize survey + analytics into a report

You're a growth lead at a B2B SaaS. You ran a signup survey and pulled your funnel analytics, and you need to find the single biggest reason people don't convert — and what to do about it.

Your raw data is the attached file survey-responses.csv (400 responses: respondent_id, plan, signed_up, nps, role, company_size, what_almost_stopped_you [free text], top_request [free text], usage, found_us). Read and theme the open-text answers yourself — don't just count the structured columns.

Funnel (last 12 months): 41,800 visitors → 5,640 signups (13.5%) → 2,180 activated (38.7% of signups) → 690 paid (31.6% of activated). Biggest single-page drop-off: the pricing page (62% bounce).

Produce a single self-contained HTML one-page report that states: (1) the ONE biggest conversion problem, with evidence from the data (counts + representative quotes); (2) 2-3 supporting findings; (3) ONE falsifiable experiment to fix it — a clear hypothesis, the exact change to make, and the success metric. Style it like a sharp analyst brief. Vanilla JS/CSS, no libraries. Return only the HTML.