Martech category labels have stopped describing products, and a blind test shows buyers can't tell CRMs apart

Martech category labels have stopped describing what software actually does, according to a MarTech analysis published September 24 that pairs a vendor-rebranding audit with a blind test showing B2B buyers cannot tell rival CRMs apart from their homepage copy alone.

The piece, by Digital Mindshare principal Gene De Libero, audited 23 vendor homepages from 2020 through 2026 and found 19 had changed their self-described category at least once. One vendor cycled through six identities — Jamstack, composable DXP, a composition platform, a visual workspace, a plain DXP, and now agentic DXP — while four vendors kept the same label the entire six years. Separately, B2B message-testing firm Wynter ranked 445 category leaders across 61 B2B categories by how much of their homepage messaging duplicates a competitor's. The overlap averaged 57%. Ninety-nine percent of the software homepages claimed some form of AI. Eighty-six percent claimed to save the buyer time.

What changed

Wynter followed up with a blind test: it stripped the company names off value-proposition copy from the five largest CRM vendors and asked 100 B2B marketing leaders, 73% of them VPs or CMOs, to match each passage back to its brand. The group averaged 1.86 correct out of 5 against a random-guessing baseline of 1.0, and 43% got one match or zero.

"Your positioning isn't bad because it's wrong. It's bad because it's interchangeable."

That's Wynter CEO Peep Laja, and the piece pairs his data with a separate check on category-vendor fit: of 17 vendors named in Gartner's Magic Quadrant for Digital Experience Platforms, 11 led their marketing with AI or "agentic" language, but only 13 had a verifiable AI capability shipped. Eleven of the 17 showed reasonable alignment between label and product; three overstated what they'd built, three understated it.

The pattern isn't isolated to one columnist's read of two studies. Chiefmartec's Scott Brinker, who runs the annual State of Martech census, reported in May that the commercial martech landscape grew just 0.79% in 2026 to 15,505 products, after adding 1,488 and losing 1,367 — a level of churn well above the net growth. He called it a possible "peak martech" moment. MarTech's De Libero quotes Brinker separately telling a 2022 podcast that "everybody wants to be their own category creator and have a category of one. It's a mess" — describing a dynamic that, four years on, the Wynter data suggests has only gotten more crowded, not more differentiated.

Who it affects

This lands hardest on marketers in saturated categories — CRM, CDP, marketing automation, and now anything that can plausibly claim "agentic" — where the category label was supposed to do positioning work the product itself no longer does. If 11 of 17 analyst-recognized vendors in one category are all leading with the same AI framing, the label has stopped functioning as a filter for buyers, and it has stopped functioning as a differentiator for marketing teams that lean on it in place of a real value proposition. A founder-led SaaS company competing against a well-funded incumbent loses the most here: the category label was one of the few free ways to signal parity, and Wynter's overlap number says buyers have already stopped reading it that way.

It also affects anyone budgeting analyst-relations time on category placement. A Magic Quadrant slot signals less than it used to if a third of the vendors in it are overstating or understating what they ship, per MarTech's read of the data — buyers doing diligence past the label will find the gap.

What to do differently

The direct implication of a 1.86-out-of-5 blind score is that category language is not carrying your differentiation, whatever your homepage currently assumes. The fix isn't a new label — Brinker's data shows the market already has one for everything — it's specificity that survives having the logo removed.

• Audit your last three years of homepage headlines for category-label churn — repeated rebranding is itself a signal you're chasing the market instead of defining it
• Replace generic AI and time-savings claims with the specific mechanism and the specific number behind them
• Treat an analyst category placement as a distribution channel, not a positioning strategy
• Re-test messaging blind, without company names attached, at least once a year

None of this means categories disappear. Brinker's own view, cited in the piece, is that "categories suck, but the only thing that would suck worse than categories is having no categories." The realistic takeaway for a software marketer is narrower: the category label is doing less work than it used to, buyers are already discounting it, and the marketing budget that used to go toward owning a category name is better spent proving a specific claim a stripped-down blind test would still recognize.