The Signal-Based ABM Stats Everyone Cites Have No Source

"32% win rate versus 13%. A 94 day sales cycle versus 151. A 4.2x pipeline-to-close ratio versus 1.8x."

That's the cluster. And it's everywhere right now: decks, LinkedIn posts, agency blogs, all citing the same "2026 benchmark of 94 B2B companies running signal-based motions vs list-based ABM." No report attached to it. No publisher named. Just the sentence, repeated with enough confidence that nobody stops to ask where it came from before they use it.

So I stopped. This is what four numbers look like once you trace them back one fetch at a time, and what's still standing once you do.

The claim, as it circulates

The numbers travel as a set. Signal-based ABM converts at a 32% win rate; list-based ABM converts at 13%. Signal-based deals close in 94 days; list-based deals take 151. Signal-based pipeline converts to closed revenue at 4.2x; list-based pipeline converts at 1.8x. And every version carries the same attribution line, word for word: "a 2026 benchmark of 94 B2B companies running signal-based motions vs list-based ABM."

No report. No publisher. No link to a methodology, a survey instrument, or a single respondent. Just the sentence.

A fourth figure, something about the share of revenue that's marketing-sourced under a signal-based outbound motion, rides along in a few versions of this cluster. It changes value depending on which site you're reading, and that's its own tell. So I'm leaving it out and sticking to the three numbers that hold constant across every copy I found.

Where it spread

Three pages carry this cluster where I could confirm it on the page itself, by direct fetch, not by trusting what a search snippet said was there.

• thesmarketers.com, "Signal-Based Selling: The Evolution of ABM in 2026." This is the earliest carrier and the likely origin. That post is self-attributed, "from our 2026 benchmark," authored by Enoch Pakanati, CEO of The Smarketers, a B2B marketing agency and not a research or analyst firm, published April 22, 2026. The site carries no linked methodology page, no sample list, and no data appendix anywhere on it.
• growthspreeofficial.com, "10 Best B2B SaaS Signal-Based ABM Agencies (2026)." It repeats the full cluster and, to its credit, correctly cites The Smarketers post in a Sources section. Fair citation practice, wrong assumption underneath it: a footnote isn't the same thing as a checked source.
• growthspreeofficial.com, "Signal-Based ABM for B2B and B2B SaaS." A second page on the same domain, repeating just the 32% and 13% figures, relabeled "Industry Benchmark." No link. No source name. Not even the one attribution the original post carried.

Beyond those three, the pattern holds without needing individual names. A wider field of ABM statistics roundups, agency listicles, and platform blog posts surface for this exact cluster in search results. But fetched directly, most of them just don't contain the numbers at all.

What the search layer told us versus what the pages say

There's a second failure sitting on top of the first one.

Two receipts, by URL. Ciente, "Signal-Based ABM: Why Your Account List Is Only Half The Strategy," is named by search results as a source for the 32% versus 13% figure. And it isn't on the page. Salesmotion, "Account-Based Selling: The Signal-Driven Playbook for 2026," is named as a source for the 94-day versus 151-day cycle comparison. It's not there either, anywhere in the full article.

And I checked roughly fifteen more sites the same way, all pulled from the same search results: digitalapplied.com, geisheker.com, kokasexton.com, motionabx.com, devcommx.com, demandloops.substack.com, two articles on prospeo.io, two pages on usergems.com, landbase.com, intelligentresourcing.co, recotap.com, digitallitmus.com, and autobound.ai. Same result every time. I'm not going to characterize any of those sites beyond that one fact, because the fact is the whole point of this exercise: checked, absent, next.

So the citation web for this cluster is only partly real. Part of it is three pages that actually repeat the number on the page, and part of it is a search-summarization artifact: aggregation asserting a citation that the page underneath just doesn't support. Those are two different failures stacked on top of each other, and nothing else ranking for this query bothers to separate them.

The search for a primary source

The Smarketers is a B2B marketing agency, founded in 2015, a HubSpot and SEMrush partner, and an ITSMA Gold Award winner for its ABM work. Not a research firm. Not an analyst house. An agency, the same category of business as the readers deciding whether to trust the number.

The self-attribution line, quoted in full: "From our 2026 benchmark of 94 B2B companies running signal-based motions vs list-based ABM." I checked the site for anything that would back that up: a linked report, a methodology section, a survey instrument, a data appendix. And nothing turned up.

There is one on-page link relevant to the claim. The anchor text reads "The Smarketers ABM Benchmark 2026," and it points to a toolkit page. But fetched directly, that page is a gated lead-magnet form, with no methodology, sample list, data appendix, or downloadable report; it cites zero statistics. If your objection here is "but don't they link to their benchmark," that link goes to a form asking for your email address, not a benchmark.

What we found

Here's the full chain, laid flat. An agency published three numbers as "our benchmark," with zero underlying data made public anywhere on its own site. Two other pages picked the claim up downstream: one kept the attribution intact, one stripped it down to "Industry Benchmark" with no name attached at all. And search aggregation then told searchers the claim showed up on pages where, once you actually read those pages, it says nothing of the kind.

So that's the complete trail, start to finish. There's no fourth layer where a primary study is sitting, waiting to be found. I looked for one.

What defensible signal-based versus list-based evidence exists

None of that means the underlying direction, prioritizing accounts by buying signal instead of a static list, is wrong. But it just means the specific multiples everyone quotes aren't the evidence for it. Real evidence exists. It's thinner and more hedged than a clean 32-versus-13 split, which is probably exactly why nobody quotes it instead.

The Starr Conspiracy, another agency, ran its own ABM Operations Audit across 47 deployments between January and October 2024, spanning B2B SaaS, HR tech, and fintech. Its own writeup discloses that the sample skews mid-market, 200 to 1,000 employees, and states outright that it isn't statistically representative. I'm not citing this for a specific multiple. I'm citing the disclosure, because that's what an honest sample note looks like sitting next to a benchmark that discloses nothing at all.

The same page attributes a conversion figure onward to Forrester's 2024 B2B Buying Study. That's one hop closer to a credentialed source than "our benchmark" ever gets, but I didn't independently fetch the original Forrester report, so I'm not treating that figure as confirmed here. It's a different tier of sourcing than The Smarketers' claim. It still isn't a number I'd put in front of a client without going to Forrester directly first.

And there's a real counterweight worth naming. Andrei Zinkevich and Vladimir Blagojevic at Full Funnel argued, in May 2025, that third-party intent data without genuine first-party engagement can reproduce the exact ineffective outreach it's supposed to fix. One firm's argument, not a benchmark. But it's a useful check on the assumption that "signal-based" automatically beats "list-based," which is the assumption the fake 32/13 split exists to confirm without anyone doing the underlying work.

Put together: the direction has real support, with disclosed limits attached to it. But the specific numbers used to sell that direction do not, and how intent data providers actually differ matters more to your outcome than which multiple ends up on a slide.

What this means for how you build a target account list

You don't need a manufactured stat to justify moving off a static target account list toward live signal scoring. The defensible evidence above, thin as it is, already supports the direction on its own terms. What you don't get to do is cite the 32/13 split, or its cycle-time and pipeline-ratio companions, in a deck or a client proposal without the caveat just built above.

And that matters more than it sounds like it should. The same failure that let an unsourced number travel through three pages and a search summary will travel through your own account scoring model just as easily, if nobody on your side is checking. General skepticism doesn't fix that. Scoring buying intent without a black box does, because the weight you put on a given signal traces back to a rule you wrote and can defend, not a stat somebody else published and nobody checked.

A short checklist before you cite anyone's benchmark

Before the next number like this lands in your inbox or your feed, run it through the same checks I used to trace this one.

Fail any one of those and the number doesn't go in the deck. We run a version of this check on every stat before it goes in front of a client. And it kills more numbers than it passes. That's how GTM engineering approaches ABM generally: check the mechanism before you trust the output, whether the output is a benchmark or a model score.