The hero — "The AI Analytics Platform for your whole team" — addresses everyone, which means it addresses no one first. The data engineer evaluating a notebook replacement and the VP of Sales who wants to stop filing data tickets are both on this page, and the page treats them as one person with one problem. The FAQ section makes this explicit: it answers "Who uses Hex?" with a paragraph that covers both audiences in the same breath, then tries to serve each with a single "Get started for free" CTA. A data team evaluating a Jupyter + Looker replacement wants to see the notebook, the SQL cell, the Python environment, and the governance layer before they commit to a trial; a business user wants to know they can ask a question in Slack and get a real answer. The page gives both audiences a generic CTA and hopes they sort themselves out. Split the page into two explicit paths above the fold — one for data teams, one for business users — or pick the primary buyer and build the page for them, then let the secondary buyer find their path through navigation.

This is the one thing Tableau can't do, Looker can't do, and ChatGPT definitely can't do: a business user's question and an analyst's deep model produce the same number because they're drawing from the same governed context. That claim appears in the FAQ answer to "How is Hex different from traditional BI tools" and again in "How is Hex different than other AI Analytics platforms" — both places a buyer only reaches if they're already skeptical enough to scroll through an accordion. The hero says "trusted data insights" and "one integrated platform," which are phrases every BI vendor uses. Move the shared context claim — specifically, the idea that the analyst's logic and the business user's answer are the same answer — into the hero or the first section below it, in plain language, before the buyer has decided you're generic.

"Hex has helped us empower business analysts to explore and dig into data on their own" — Meghana Reddy — is a strong outcome claim, but the page doesn't tell the reader what Meghana's role is, what company she's at, or what she switched from. Dhaval Patel's quote — "days into hours, which for our business, is millions of dollars" — is the single most commercially compelling sentence on the page, and it's sitting in a carousel with no context about who Dhaval is or what problem he was solving when he found that. Named proof only converts when the reader can place themselves in the named person's situation. Add company, role, and the before-state to each testimonial, or pull the two or three strongest quotes out of the carousel and give them dedicated placement with full context — Dhaval's quote alone, with his title and company and the workflow it replaced, belongs in the first scroll.

This matters because a buyer who is actively evaluating Hex against Looker — which, given the ICP and the FAQ content, is a large portion of the page's traffic — arrives already holding Looker in their head. The page doesn't meet them there. "Finally — anyone in your business can get trusted data insights" doesn't tell a Looker user why they should switch; it tells them Hex exists. The FAQ answer to the Tableau/Looker question is actually sharp: "A complex model built by a data scientist can become an interactive app for business users with one click." That sentence is doing real positioning work. It belongs on the page before the FAQ, in a section explicitly framed around the switching decision, not hidden behind an accordion that only the already-convinced buyer opens.