We scored 10 real B2B positioning statements. One passed.

A positioning statement does four jobs at once. It names who you're for, the problem you lead with, what you actually do, and the language only you own. Drop one of the four and the statement still reads fine to your own team, because your team fills the gap from memory. It stops working on strangers, and it stops working on the AI engines that now answer "who should we hire for this" before a human ever loads your homepage.

We scored 302 real healthtech company homepages against a 19-criterion rubric on 2026-09-01. On those four positioning elements, exactly one company out of 302 earned full marks. 173 of the 302 took a hard zero on at least one of the four.

Here are ten of those cases with their real homepage language and their real scores, followed by the failure pattern, which turns out to be the same one almost every time. Every company below has its own scored page in the public 2026 Healthtech Messaging Index, so you can check our work.

What are the four parts of a positioning statement?

Each part scores 0, 1, or 2, for eight points total. The four aren't a writing formula. They're the four questions a stranger has to answer correctly after one pass at your page.

• Who it's for. A named buyer, stage, and context. "Growing companies" and "modern teams" score 1. An org type or a role a reader can match themselves against scores 2.
• The problem you lead with. The buyer's pain, in the buyer's own words, arriving before your product does. A problem mentioned generically scores 1. A page that opens in solution language scores 0.
• What you actually do. A visitor can repeat your mechanism in one sentence without rereading. This is the part almost everybody already has.
• The language you own. A named term, category, or framework that belongs to you and that an AI engine can quote back and attribute. Commodity category language scores 0 no matter how polished it is.

Eight out of eight means a stranger learns who should buy, what's broken, what you do about it, and what to call it. That's the same bar as the Three Questions Test, scored instead of eyeballed.

What do 10 real B2B positioning statements score?

Scores below are out of 8 on the four positioning elements. The overall rubric score out of 38 is in brackets, since a company can position well and still lose points on proof, alternatives, or conversion.

1. 1[Mend](/healthtech-messaging-index-2026/mend) scores 8 of 8 [30/38]. The only complete positioning statement in 302 companies. Buyer: "Built for CCBHCs, CMHCs, and outpatient behavioral health organizations." Problem first: "Without Mend, this is what every week looks like." Owned language: The Operational Gap, Patient Readiness Agent. Notice it isn't the highest overall scorer. It's the only one missing nothing.
2. 2[Theator](/healthtech-messaging-index-2026/theator) scores 7 of 8 [32/38, the top overall score]. Owns "Surgery-to-Text" as a registered term and "The Intelligence Layer for Surgery" as a category. Loses a point on buyer: "leading health systems performing 100,000+ procedures annually" tells you the org and never the role signing the contract.
3. 3[SmarterDx](/healthtech-messaging-index-2026/smarterdx) scores 7 of 8 [31/38]. A problem line most B2B pages would kill for: "Revenue is lost when teams only see part of the patient story." Same missing point as Theator. Health systems are obvious, the CFO or VP of Revenue Cycle is never named.
4. 4[BrightInsight](/healthtech-messaging-index-2026/brightinsight) scores 7 of 8 [30/38]. Leads with a quantified problem, "fewer than 1 in 3 patients still on a novel therapy at 12 months," and owns "the leaky bucket" as a frame. The buyer is a pharma brand, with no role or company stage attached.
5. 5[Capital Rx](/healthtech-messaging-index-2026/capital-rx) scores 7 of 8 [30/38]. Owns "fiduciary PBM" and "The Judi Rx Model," and names three buyer segments outright. The hero gestures at the pain instead of stating it: "A model built to lower your costs, not hide them" is a solution frame wearing a problem's clothes.
6. 6[DeepScribe](/healthtech-messaging-index-2026/deepscribe) scores 7 of 8 [30/38]. Unmistakable buyer (community oncology) and owned language ("Ambient Operating System for oncology"). The problem arrives as a benefit: "Reduce the cognitive burden with automation and intelligence across the oncology workflow."
7. 7[LeanTaaS](/healthtech-messaging-index-2026/leantaas) scores 7 of 8 [30/38]. Owns iQueue, names hospitals and the specific units. The hero is pure outcome: "Maximize healthcare capacity with AI and prescriptive analytics." The wasted capacity and the burned-out staff show up below the fold, after the reader has already decided.
8. 8[Abridge](/healthtech-messaging-index-2026/abridge) scores 5 of 8 [25/38]. Excellent buyer clarity, clear mechanism, and a zero on problem. The page opens "Enterprise-Grade AI / For every moment of care" and never states a pain before the pitch.
9. 9[Sully.ai](/healthtech-messaging-index-2026/sully-ai) scores 5 of 8 [25/38]. Strong coined language ("AI-employees," a named Consensus Mechanism) and another zero on problem. "Superhuman team of AI-employees for healthcare" is the product, offered to a reader who hasn't been told what's wrong yet.
10. 10[tab32](/healthtech-messaging-index-2026/tab32) scores 6 of 8 [24/38], and it's the interesting one. It has the single best problem line in the set: "You opened your practice to focus on patients. Instead, you're chasing claims, managing no-shows, and toggling between systems that don't talk to each other." Clear buyer, clear mechanism, and zero owned language. Every term on the page is commodity vocabulary any competitor could paste onto their own site. tab32 is the only company in all 302 with that exact shape.

Which part do most companies get wrong?

The same one. Across all 302 homepages, Solution Clarity averaged 1.58 out of 2 with exactly one zero in the entire set. Problem Leadership averaged 0.59 out of 2 with 154 zeros.

Read those two numbers next to each other. Nearly every company can explain what it does. More than half can't name the problem they solve before they start selling. That gap is Solution-Centric Marketing with a number attached to it, measured on real companies with real funding and real marketing teams.

Owned language is the second casualty: 45 of 302 scored zero, which means one company in seven publishes nothing an AI engine could quote and attribute to them specifically.

Why does a missing problem cost more in 2026 than it used to?

Because your positioning statement is now the raw material for an answer you don't write. When a founder asks an AI engine "who helps healthcare systems with revenue leakage," the engine needs a page that connected a named problem to a named company. A page that opens with a category and a capability gives it nothing to work with, and the engine quotes whoever did the naming.

This is the part most positioning advice hasn't caught up to. You can have a beautifully written statement, a clean brand, and a real product, and still be structurally unquotable, because the problem sentence that would have made you the answer was never on the page. Owned language is the same story. "Surgery-to-Text" gets attributed to Theator. "Cloud-based practice management software" gets attributed to nobody.

How do you score your own positioning statement in ten minutes?

Open your homepage. Read only what's visible before the first scroll, and answer in order. Give yourself 0, 1, or 2 on each.

1. 1Who is this for? If your answer needs the words "anyone who" or "companies that want to," score 0. If you can name the org type and the role, score 2.
2. 2What problem arrives first? Find the first sentence that describes something wrong in the buyer's world. If it comes after your product name, score 1. If there isn't one above the fold, score 0.
3. 3What do you do? Say it out loud in one sentence without reading again. Most companies score 2 here, and it's the point that tempts you into thinking the statement works.
4. 4What term is yours? Find one phrase a competitor couldn't paste onto their site by tonight. No phrase, score 0.

Under 6 out of 8, the problem is structural and rewriting the headline won't move it. The free Brand Signal Score runs the full 19-criterion version of this on your homepage, which is the same rubric that produced every number in this article. If you want the four parts written out in order first, what a strong B2B positioning statement looks like covers the anchors.

What this means for you

Ten real companies, and nine of them are missing a part. These aren't struggling businesses. Several are category leaders with funding and marketing teams, and they still publish statements a stranger can't complete.

The reason is almost never writing ability. A missing problem sentence usually means the company never agreed internally on which problem it leads with, and missing owned language usually means nobody was willing to name a category and defend it. Both gaps get papered over by competent copy, which is why a statement can read well for years while quietly failing on every stranger and every engine that meets it.

Score your own before you rewrite anything. If you come up short on the problem line or the owned term, you have a decision to make rather than a paragraph to polish.