Clarity
Do they understand what you do?
15 could name what kind of product this is, unprompted.
https://www.shrinkmystack.com/15 AI-simulated buyers
Your message needs work: they know what it is and who it's for, but not why it's worth their time or why to pick you.
Do they understand what you do?
15 could name what kind of product this is, unprompted.
Can they tell what it solves, and who it's for?
15 could quickly tell what problem it solves and who it is for.
Do they actually want it?
8 would take a meeting to learn more.
Is there a reason to pick you over the alternatives?
3 could name a reason to pick you over a similar option.
Your page describes: SaaS spend management. They said:
12 couldn't name one; 3 got it right.
Four separate measures, not stages: all 15 personas answered all four questions. Each square is one persona.
Define 'agent' where it first appears. Not one of the four layers, and it does not affect the scores above or the order to fix them in.
These are 15 simulated buyers. Want 15 real ones?
Test with humansThe first is on your weakest layer, the second on the next, the third on the layer the most buyers had a problem with. Each says what to change on the page and why, with one simulated answer behind it.
Why: ServiceNow is mentioned once and never positioned, and readers whose contract and renewal data already sits in procurement suites have to guess at the overlap. Add a short section stating what this does that a procurement or ITSM record does not: reads the real terms out of the contract document, sizes capability-level overlap between tools, and labels every figure by evidence rather than accepting declared numbers. State the boundary in terms of what it adds, not what competitors lack.
2 of 15 raised this
“the page never names what ServiceNow's contract/renewal module can't do, so I had to infer the differentiation from scattered lines like 'agent,' 'living record,' and 'overlap sized in dollars' rather than being told it directly”
Why: Nothing on the page carries a customer name, reference or before/after figure, so 'tool overlap sized in dollars' and 'unused seats and dead tools surface on their own' rest entirely on assertion. Put a single specific proof point directly next to the overlap claim in 'Where do we overlap?' — a named company or, failing that, an anonymised but concrete result with sector, stack size and dollars of overlap found. Proof placed at the claim does more than a logo strip further down.
6 of 15 raised this
“half the value — Okta connector, AI usage connections, renewal alerts — is labeled "built and waiting" or "coming next," not shipped.”
These landed. Keep the wording when you edit around it.
The problem and audience land within the first screen
“the hero line "One living record of every tool you pay for" plus "Upload your contracts and a finance sheet... you have a living record of every tool you pay for — what you own, what you pay, who owns it, and when it renews" tells me exactly what it does and it's aimed squarely at finance/procurement/IT teams managing SaaS spend, which is me. I didn't have to hunt”
'Verified vs. declared' is the one line respondents could repeat back as a differentiator
“The one thing that would actually pull me toward this one is the "verified vs declared" labeling on every number - "we never show you a wrong one" is a real point of difference from a generic spend dashboard”
Why: The verified-versus-declared labelling is the one thing readers repeat back as the reason to pick this over dashboards, consultant decks and spreadsheets — yet it is buried mid-paragraph as 'Every figure is labeled by how well we can prove it.' Replace or extend the H1 'The system of record and agent for your SaaS stack' so verified-versus-declared is in the first few words, and make it a named, repeatable label rather than a sentence a scanning reader will skim past.
2 of 15 raised this
“the page never names what ServiceNow's contract/renewal module can't do, so I had to infer the differentiation from scattered lines like 'agent,' 'living record,' and 'overlap sized in dollars' rather than being told it directly”
Why: The section labels — 'Record plus agent', 'Start with the number', 'Where this goes' — read as internal shorthand and give a scanning reader no reason to prefer this page to any other SaaS management page. Rewrite each so the heading alone carries the distinguishing idea: contracts read and labelled by evidence, overlap sized in dollars, renewals dated from first upload. Headings are where differentiation is either made or lost.
2 of 15 raised this
“the page never names what ServiceNow's contract/renewal module can't do, so I had to infer the differentiation from scattered lines like 'agent,' 'living record,' and 'overlap sized in dollars' rather than being told it directly”
Why: The page repeats in-development flags inline — 'Claude is the first connector and is not switched on in production yet', 'names what it still needs' — so the reader assembles a picture of half a product from scattered admissions. Consolidate this into a single short 'What is live today' line under the product section: contracts, terms, renewal dates, overlap sizing and spend are live on first upload; usage connectors and identity data arrive on a stated date. Naming what works today, in one…
6 of 15 raised this
“half the value — Okta connector, AI usage connections, renewal alerts — is labeled "built and waiting" or "coming next," not shipped.”
Why: Because so much is flagged as forthcoming, any number on the page reads as projection. Instead of asserting savings, state exactly what the free audit returns from what already works: number of tools found, dollars of overlap sized, contracts read, renewals in the next 90 days. 'Start the free audit' should be followed by a line naming the deliverable and the timeframe already claimed elsewhere — about ten minutes from a contract set and a finance sheet — so value is verifiable rather than…
6 of 15 raised this
“half the value — Okta connector, AI usage connections, renewal alerts — is labeled "built and waiting" or "coming next," not shipped.”
No specific edits needed here — this layer held up.
No specific edits needed here — this layer held up.
Why: 'The system of record and agent for your SaaS stack' asks the reader to carry an undefined term through the whole page, and it is unclear whether they are buying one system or three bundled components. Add a clause at first use saying plainly what the agent does — answers questions about your stack on demand from the record — and make the product section state that the record and the agent are one product, not modules.
A deliberately adversarial read of the same answers. Each claim was checked back against what the personas said and dropped if nothing supported it.
The page's clarity win is cosmetic: respondents understand the pitch fast and then stall on whether the product exists.
Seven respondents got problem and audience inside ten seconds (theme 5), but eight said the connectors, Okta identity data and the AI piece are flagged as in development and treated the savings numbers as unverifiable, several saying this blocks deal consideration outright (theme 2). Fast comprehension is being spent on a product respondents conclude is half-built. Three more could not tell whether it is one system or three bundled components (theme 0).
The only differentiator that lands is undermined by the same page that makes it.
Six respondents could repeat 'verified vs. declared' back as the thing that sets the product apart (theme 7). But eight flagged that the live usage connectors and identity data that would produce verification are not in production (theme 2), and four found no customer names, references or case studies anywhere on the page (theme 3). The differentiator is a claim about proof with no proof behind it.
Credibility rests entirely on assertion, and the page offers no substitute path to it.
Four respondents found no customer names, references or case studies anywhere (theme 3), one specifically wanting a manufacturing reference running real contract data before assigning value. Another said a real audit against their own company data would be required before investing further time (theme 6). Both routes to belief — someone else's evidence or their own — are closed off by the page.
The roadmap disclosure is the single largest drag on the page and outweighs every positive reaction combined.
Eight respondents raised the in-development functionality and unverifiable savings, several stating it blocks deal consideration or real validation today (theme 2). That is more respondents than named the clarity win (seven, theme 5) or the one working differentiator (six, theme 7). The strongest signal on the page is a reason not to buy yet.
Buyers are left to do the competitive analysis themselves, and they conclude they already own this.
ServiceNow is named once and overlap with tools respondents already run is never addressed, so differentiation has to be inferred (theme 4). One respondent said the contract and renewal data is already tracked in their existing Coupa contract. When the page's own differentiator depends on verification that is not yet shipping (theme 2), 'we already have this' is the default conclusion.
Undefined load-bearing language makes the headcount and savings implications unsupportable.
Three respondents said 'agent' and 'verified' carry weight without definition, one reducing the offering to a SaaS audit dashboard with chat and noting there is no proof of the headcount replacement being implied (theme 0). The savings numbers were separately treated as unverifiable because of the roadmap disclosure (theme 2). Two independent routes lead to the same place: the quantified claims cannot be checked.
Adjacent incumbents are never named, so overlap has to be guessed
2 of 15
“the page never names what ServiceNow's contract/renewal module can't do, so I had to infer the differentiation from scattered lines like 'agent,' 'living record,' and 'overlap sized in dollars' rather than being told it directly”
'Verified vs. declared' is the one line respondents could repeat back as a differentiator
5 of 15 · what worked
“The one thing that would actually pull me toward this one is the "verified vs declared" labeling on every number - "we never show you a wrong one" is a real point of difference from a generic spend dashboard”
“the "labeled by how well we can prove it" framing — verified vs. declared, and the line "when we can't back a number, we say so — we never show you a wrong one." That's a concrete, checkable claim about data integrity that a spend dashboard or a generic consultant deck doesn't usually make”
“The thing that would actually pull me toward this one is the "verified vs. declared" labelling and the explicit line "we never show you a wrong one" — most competitors just show a blended savings number”
“The thing that would make me pick it is the "verified vs. declared" labeling on every dollar figure — "we never show you a wrong number" plus the caveat that missing contracts stay blank instead of guessed”
“Every figure is labeled by how well we can prove it. When we can't back a number, we say so — we never show you a wrong one.”
“the real change is Okta plus the AI usage connectors turning "declared" seat counts into actual observed usage — that's the number I'd take to a renewal conversation, not a spreadsheet someone half-updated eight months ago”
Roadmap disclosure reads as half the product not existing yet, and it blocks consideration
6 of 15
“half the value — Okta connector, AI usage connections, renewal alerts — is labeled "built and waiting" or "coming next," not shipped.”
“The "in development," "coming next," "not switched on in production yet" language all over the connectors section is the tell; a mature vendor wouldn't ship a page that honest about gaps.”
“it's explicitly an "illustrative sample," Okta and directory integration (the thing that turns seat counts into real unused-seat dollars) is "next," not live, and Claude is "not switched on in production yet" — so today it's mostly a nicer read of data I already have in ServiceNow plus finance exports”
“the mechanics that would actually produce those numbers for me — Okta identity connection, Entra, Workspace, ChatGPT/Gemini connectors — are flagged "in development" or "coming next." So no, I wouldn't take a meeting yet”
“Whether it actually delivers depends on whether Okta and the AI usage connectors (currently only Claude, and not "switched on in production" per their own admission) actually work as promised”
No named customers or case studies, so credibility rests on assertion
3 of 15
“Small, early-stage vendor — feels like a seed or Series A startup, maybe a dozen people, still building half the connectors they're selling.”
Load-bearing terms and the product's shape are left undefined
3 of 15
“the tabs listing 'System of record / AI usage / Contracts & renewals' side by side as if they're three separate things, which made me pause to work out whether this is one product or three bundled together”
“Same with "verified" — verified against what exactly, a contract PDF a human still has to read, or an API check? Those two words are load-bearing and neither is defined precisely enough for me to know what I'm buying.”
“Basically a glorified SaaS audit dashboard with some agentic chat bolted on, built around dollar figures it labels "verified" versus "declared."”
“Whether it actually delivers depends on whether Okta and the AI usage connectors (currently only Claude, and not "switched on in production" per their own admission) actually work as promised”
The problem and audience land within the first screen
5 of 15 · what worked
“the hero line "One living record of every tool you pay for" plus "Upload your contracts and a finance sheet... you have a living record of every tool you pay for — what you own, what you pay, who owns it, and when it renews" tells me exactly what it does and it's aimed squarely at finance/procurement/IT teams managing SaaS spend, which is me. I didn't have to hunt”
“the subhead "The system of record and agent for your SaaS stack" plus "One living record of every tool you pay for" told me the problem (sprawl, unowned renewals, unmonitored AI spend) within the first screen”
“It was obvious within the first screen — "the system of record and agent for your SaaS stack" plus "One living record of every tool you pay for" tells you exactly what it does before you even scroll.”
ERP-centric and manufacturing buyers do not see themselves addressed
2 of 15
“A line naming SAP or an ERP-adjacent workflow explicitly — something like 'sits alongside SAP Ariba/vendor master' — plus one manufacturing or industrial reference logo would do it”
“What's less clear is company size or industry fit; there's no "built for mid-market healthcare" or similar, so I'm assuming general applicability rather than being told it.”
Respondents want an audit on their own data before spending time
1 of 15
15 AI-simulated personas matched to your target market. Each answered independently, without seeing your goal, the scoring criteria, or each other’s answers. Attribution is role, industry and company size only.
Every answer on this page was written by an AI model role-playing a buyer profile, scored on Wynter’s B2B Message Layers framework. The personas were sampled in code across role, industry, company size and behavioral traits; the model wrote only the answers. Scores arrive through fixed verdict categories and the counts are computed in our own code, so no number here was written by a model.
The count is how many personas cleared the bar on each question. A yes can be unhesitating or come with reservations; the scorecard counts both as a yes, and this is the only place the difference is shown. Per layer:
These answers are AI-simulated and directional. Validate anything you’re betting on with real buyers, your ICPs.
A detailed, section-by-section message test report from verified B2B professionals who are actually in-market for what you sell.







