# Message test — https://statsig.com/ After reading your page, 11 of 15 personas would take a meeting to learn more. - **Page tested:** https://statsig.com/ - **Audience tested against:** Heads of product, product managers and data or growth leaders at US software companies with 200+ employees choosing an experimentation and feature flag platform - **Personas:** 15 simulated - **Report:** https://grader.wynter.com/r/statsig-the-modern-product-development-platfor-euL7h5Y > These answers are generated by AI, scored on Wynter's B2B Message > Layers framework using behaviorally-diverse simulated personas. The > methodology is real and the critique is directional. What a simulated > persona cannot have is a live budget, a renewal coming up, or a boss > asking about this quarter. --- ## 01 · The scores Every persona answered all four questions. These are four independent proportions of the same panel, not stages of a funnel. | Layer | Question | Cleared the bar | Strength | Of those who passed | | --- | --- | --- | --- | --- | | 1. Clarity | Do they understand what you do? | 15/15 | 85% | 5 without hesitation, 10 with reservations | | 2. Relevance | Can they tell what it solves, and who it's for? | 13/15 | 72% | 1 without hesitation, 12 with reservations | | 3. Value | Do they actually want it? | 11/15 | 63% | all with reservations | | 4. Differentiation | Is there a reason to pick you over the alternatives? | 12/15 | 67% | all with reservations | **Brand alignment** (a side metric, not one of the four layers) — 13/15, 70% strength (all with reservations). Does the page read like the company you actually are? **Fix first: Value.** Earliest failing layer, walking the sequence in order — not simply the lowest score. --- ## 02 · What to change, layer by layer Ordered worst-first. Specific edits, not a restatement of the score. ### Value **Add a migration section after the product grid covering timeline, effort, and data import.** A buyer replacing five tools cannot tell what switching costs them in time or retraining. Add a short block naming typical migration timeline, who does the work, and what carries over from existing tools. *effort medium · impact high · tested against Answer the live objection* **Add pricing range or pricing-model line near the Brex 20% cost-savings claim.** The page claims 20% cost savings from consolidating tools but never says how Statsig is priced, so no one can do the math. State the pricing model and link to the pricing page beside that stat. *effort low · impact high · tested against Proof next to the claim* **Replace "shockingly affordable" in Feature Management with a concrete price or comparison.** "Shockingly affordable" is a claim with nothing behind it and reads as sales talk next to real numbers elsewhere. Give a per-event or free-tier figure instead. *effort low · impact medium · tested against Specifics beat superlatives* ### Differentiation **Add a baseline to each case study stat, such as the before figure for Brex's 50%.** "+50% time efficiency gain" has no starting point, so the number cannot be judged. Write what the team did before and after, with the timeframe. *effort medium · impact high · tested against Proof next to the claim* **Add source and measurement period under the 1+ Trillion and 2.5 Billion scale stats.** Readers discount "1+ Trillion events processed per day" because nothing says how it is measured or over what window. Add a one-line footnote naming the measurement period and source. *effort low · impact high · tested against Proof next to the claim* **Add a line under "5+ products in a single platform" naming the tools it replaces.** The consolidation story is the strongest thing on the page but stays generic. Name the categories of separate tools a team drops when they consolidate, and what that removes from their stack. *effort low · impact medium · tested against Give a reason to choose you* ### Relevance **Name the buyer in the hero subhead, such as product and data teams.** Readers work out who the page is for from logos and testimonials instead of being told. Put the role and team in the subhead under "Measure what ships." *effort low · impact high · tested against Name the audience* **Add a budget-owner line near the case studies covering vendor consolidation cost.** Everything speaks to practitioners running tests, with nothing for the person signing the contract. Add a sentence stating what consolidating separate experimentation, analytics, and replay contracts does to total spend. *effort medium · impact medium · tested against Name the audience* ### Clarity **Define "experimentation velocity" where it first appears in the Ancestry stat.** "30x experimentation velocity" and "velocity from 70 to 600+" assume the reader knows the unit. Say experiments run per quarter or per year in plain words. *effort low · impact medium · tested against Plain language* ### Brand alignment (side metric) **Cut "optimize at the speed of AI" from the hero subhead.** The phrase carries no meaning and could sit on any vendor's homepage. Replace it with the concrete job: test every release against real customer metrics before it ships wide. *effort low · impact medium · tested against Concrete over abstract* --- ## 03 · What is working ### The consolidation story — one platform replacing five tools — is the clearest thing on… Eight respondents read the unified platform claim (experimentation, flags, analytics, session replay in one place) off the hero and product grid and could repeat it back as the core value. > one platform, one source of truth for "did this feature actually move the number." > > — Data Leader, SaaS, 501-1000 > kill two or three vendor contracts — flagging, experimentation, and analytics — and get my engineers and data team working off one data model instead of reconciling numbers across LaunchDarkly, Optimizely, and Amplitude > > — Product Manager, SaaS, 201-500 > A/B testing infrastructure bundled with product analytics and session replay > > — Senior Product Manager, Technology, 501-1000 > the hero line "Measure what ships. Ship what matters." plus the subhead about testing "how each release - from features to AI-powered experiences - lands with customers" told me within a few seconds this is about shipping features and measuring their impact, and the product grid right below (Experimentation, Feature Management, Product Analytics, Session Replay) confirmed it. > > — Director of Product, Software, 1001-5000 > If it worked as promised, I'd kill 3-4 contracts — flagging, experimentation, analytics, maybe session replay — and get my data scientists back the time they currently spend stitching results together across tools. That's real money and real velocity if the Brex "50% reduction in time spent by data scientists" and Notion "30x experimentation velocity" numbers hold up under my own math > > — Data Leader, SaaS, 5000+ > the product list (flags, experimentation, analytics, session replay) was concrete. What got vague was the quantified payoff sitting next to it: "consolidate your stack," "optimize at the speed of AI," and "industry's most generous free tier" are all puffery with no number attached > > — Data Leader, SaaS, 5000+ > the hero line "Measure what ships. Ship what matters" plus "Test how each release — from features to AI-powered experiences — lands with customers" told me the problem (you ship stuff, you don't know if it worked) within the first screen. > > — Product Manager, SaaS, 1001-5000 > It was fast — the header "Measure what ships. Ship what matters." plus the subhead about testing releases "so you can protect trust and optimize at the speed of AI" told me the problem in one breath: you ship features blind and need to know what actually lands with customers. > > — Data Leader, SaaS, 5000+ ### Named case studies with numbers — Brex, Notion — are the proof respondents believed Four respondents cited the named customer case studies and their specific figures as credible evidence and the basis for a budget conversation. > The Brex case, "+50% time efficiency gain" and "20% cost savings from consolidating analytics and experimentation tools," is the kind of consequence that would actually move my budget conversation, because it speaks to headcount and velocity, not just features. > > — Director of Product, Software, 1001-5000 > The named customers (OpenAI, Notion, Brex, Ancestry) and specific before/after numbers like "70 to 600+ annual experiments" for Ancestry are what actually stuck with me, since that's a credible, checkable claim rather than marketing fluff. > > — Director of Product, Software, 1001-5000 > Brex's "50% reduction in time spent by data scientists" plus "20% cost savings from consolidating analytics and experimentation tools," and Notion's "30x experimentation velocity" with "600+ features released with Statsig flags." Those are specific, attributable, and map directly onto what I'd have to justify to my own CFO > > — Product Manager, SaaS, 201-500 > If it worked as promised, I'd kill 3-4 contracts — flagging, experimentation, analytics, maybe session replay — and get my data scientists back the time they currently spend stitching results together across tools. That's real money and real velocity if the Brex "50% reduction in time spent by data scientists" and Notion "30x experimentation velocity" numbers hold up under my own math > > — Data Leader, SaaS, 5000+ --- ## 04 · What the personas said ### The hero tagline and key terms don't land until the product tiles explain them Two respondents found the hero unclear on first read and said words like 'experiment' and 'velocity' are used without precise definitions. > "Measure what ships. Ship what matters." reads like a tagline built for rhythm, not clarity, and "test how each release... lands with customers" is soft verbs doing the work a noun should do. I had to get past that to the product tile labels before I had actual nouns — flags, experiments, analytics > > — Product Manager, SaaS, 201-500 > Mostly it was terms like "experiment," "velocity," and "AI-powered experiences" doing a lot of unexplained work — they're used as if everyone agrees on the definition, but an experiment counted in a dashboard could mean a rigorous A/B test or a trivial flag toggle, and the page never draws that line. > > — Director of Product, Software, 1001-5000 ### The page speaks to practitioners and skips what a budget owner needs to approve Four respondents said the messaging addresses technical users, not budget decision-makers, and omits the cost justification for consolidating vendors. > A line that names my actual budget owner problem — something like 'consolidate your experimentation spend and cut vendor count without losing analyst headcount' — plus a sourced TCO comparison against running Optimizely+LaunchDarkly+Amplitude separately > > — Head of Product, Software, 5000+ > The tone is written for a technical IC or a VP of Eng/Data who gets excited by "npx @statsig/wizard@latest" and SDK lists for fourteen languages, not for someone in my seat who needs a business case. > > — Head of Product, Software, 5000+ > I'd go in wanting to see the actual migration path and pricing at my scale, because "generous free tier" and case-study logos from OpenAI-scale companies don't tell me what this costs or how painful the cutover is for a 201-500 person company > > — Product Manager, SaaS, 201-500 ### Switching cost is the unanswered objection — migration and retraining are never addressed Four respondents wanted a documented migration case with verified numbers and pricing before they could judge whether promised time savings survive the cost of switching. > A documented, apples-to-apples migration case — current stack, swap-in timeline, and a verified before/after cost or headcount number from a company our size — not a pull-quote > > — Head of Product, Software, 5000+ > I'd go in wanting to see the actual migration path and pricing at my scale, because "generous free tier" and case-study logos from OpenAI-scale companies don't tell me what this costs or how painful the cutover is for a 201-500 person company > > — Product Manager, SaaS, 201-500 > The Brex case study claiming "+50% time efficiency gain" and "20% cost savings from consolidating analytics and experimentation tools" is the kind of number that would matter to me, but it's unsourced and context-free > > — Head of Product, Software, 501-1000 ### The big scale metrics are discounted because no methodology or source is given Six respondents flagged unsourced statistics and quantified payoffs — including the Brex numbers — as lacking baselines or measurement detail, saying this hurts credibility rather than building it. > "2.5 Billion unique monthly experiment subjects" and "99.99% infra uptime" with zero source or methodology are the kind of unsupported numbers that make me discount the rest of the page > > — Data Leader, SaaS, 501-1000 > The "50% time efficiency" and "1+ trillion events processed" numbers have no methodology behind them, so I'd discount those > > — Head of Product, Software, 5000+ > "1+ Trillion events processed per day," "2.5 Billion unique monthly experiment subjects," "<1ms post-init evaluation latency" — none of these say over what period, what counts as an "event," or how latency was measured > > — Head of Product, Software, 5000+ > the product list (flags, experimentation, analytics, session replay) was concrete. What got vague was the quantified payoff sitting next to it: "consolidate your stack," "optimize at the speed of AI," and "industry's most generous free tier" are all puffery with no number attached > > — Data Leader, SaaS, 5000+ > The Brex case study claiming "+50% time efficiency gain" and "20% cost savings from consolidating analytics and experimentation tools" is the kind of number that would matter to me, but it's unsourced and context-free > > — Head of Product, Software, 501-1000 ### No direct comparison against named competitors is offered One respondent noted the page never positions against LaunchDarkly, Eppo, or similar, leaving the competitive case to the customer quote alone. > the page doesn't do any direct comparison — the SoundCloud quote even namechecks "Optimizely, LaunchDarkly, Split, and Eppo" and just says they picked Statsig for "comprehensive end-to-end integration" without saying why those others fell short > > — Senior Product Manager, Technology, 501-1000 ### The target buyer is never stated — respondents had to infer it from logos and job titles Six respondents worked out the audience from customer logos, testimonials, and product language rather than being told, and noted the page never names who it is for. > The customer logos (OpenAI, Brex, Notion) and lines like "accelerate the speed at which we release new features" reinforce that it's for scaled product orgs, not solo devs or marketers > > — Data Leader, SaaS, 501-1000 > the "who it's for" is inferred, not stated — there's no "built for Heads of Product at mid-size SaaS companies" line, I had to piece it together from the logos (OpenAI, Notion, Brex) and testimonials > > — Head of Product, Software, 5000+ > logos (OpenAI, Brex, Notion) and testimonials from Heads of Data, Eng Managers, and PMs make it obvious who this is pitched at > > — Senior Product Manager, Technology, 501-1000 > The intended reader isn't stated in so many words like "for engineering and data teams," but it's obvious from the logos (OpenAI, Brex, Notion) and the language around "data teams," "engineers," and "product teams" throughout the testimonials — so I had to infer the audience rather than read it in a sentence > > — Director of Product, Software, 1001-5000 > the hero line "Measure what ships. Ship what matters" plus "Test how each release — from features to AI-powered experiences — lands with customers" told me the problem (you ship stuff, you don't know if it worked) within the first screen. > > — Product Manager, SaaS, 1001-5000 --- ## 05 · The hardest read An adversarial pass over the findings. Every claim below was checked against the panel's own answers; unsupported ones were dropped. - **The page's own proof points contradict each other — the same customer numbers are both the strongest evidence and the biggest credibility problem.** *(high)* Six respondents flagged unsourced statistics including the Brex numbers as lacking baselines, while four cited those same named case studies as the credible basis for a budget conversation. The proof is doing damage and work simultaneously. - **The page is unsellable internally: it convinces the person who cannot sign and ignores the person who can.** *(high)* Four respondents said the messaging addresses technical users and omits cost justification for consolidating vendors, and four more demanded migration, retraining and pricing detail before judging the value. Nothing on the page survives a budget review. - **The consolidation story is a liability, not an asset, because it raises the switching question the page then refuses to answer.** *(high)* Eight respondents read 'one platform replacing five tools' as the core value, but four demanded a documented migration case before they could judge whether the promised time savings survive the cost of switching. The headline claim generates its own… - **Readers are doing the positioning work the page should do, and they are doing it from the footnotes.** *(medium)* Six respondents inferred the target buyer from logos, testimonials and job titles because the page never names who it is for, and two said the hero tagline only makes sense after the product tiles explain it. Comprehension arrives late, if at all. - **With no competitor named and no methodology shown, every differentiation claim rests on a single customer quote.** *(medium)* One respondent noted the page never positions against LaunchDarkly or Eppo, and six discounted the scale metrics for lacking sources or baselines. Both differentiation routes are closed at once. - **Vocabulary the page treats as self-evident is not — 'experiment' and 'velocity' are asked to carry the value proposition without definition.** *(low)* Two respondents found the hero unclear on first read and said these terms are used without precise definitions, requiring the product tiles to retroactively explain them. The hero is spending attention it has not earned. --- ## 06 · Who answered | # | Role | Industry | Company size | | --- | --- | --- | --- | | 1 | Data Leader | SaaS | 501-1000 | | 2 | Growth Lead | Technology | 1001-5000 | | 3 | Head of Product | Software | 5000+ | | 4 | Product Manager | SaaS | 201-500 | | 5 | Senior Product Manager | Technology | 501-1000 | | 6 | Director of Product | Software | 1001-5000 | | 7 | Data Leader | SaaS | 5000+ | | 8 | Growth Lead | Technology | 201-500 | | 9 | Head of Product | Software | 501-1000 | | 10 | Product Manager | SaaS | 1001-5000 | | 11 | Senior Product Manager | Technology | 5000+ | | 12 | Director of Product | Software | 201-500 | | 13 | Data Leader | SaaS | 501-1000 | | 14 | Growth Lead | Technology | 1001-5000 | | 15 | Head of Product | Software | 5000+ | --- ## 07 · Before you act on this The methodology is real, and the critique is directional. What a simulated persona cannot have is a live budget, a renewal coming up, or a boss asking about this quarter. **Validate anything you're betting on with real ICPs who are actually in-market.** Being wrong is more expensive than you think. Finding out is cheaper than you'd guess. Wynter runs message testing with verified B2B professionals — trusted by HubSpot, RingCentral, Shopify, Cognism, Paddle, Veeam, Rippling and Miro. This report is kept for 60 days from 2026-10-28, then deleted along with the personas and their answers.