Positioning is working when three things move together: reply rates on outbound climb past the cold-email baseline, sales starts repeating the same phrases back without being fed a script, and the company shows up when someone asks an AI chatbot to name tools in its category. Miss all three and a solo marketer isn't looking at proof. They're looking at a slide.
That distinction matters because most solo marketers at 30-person SaaS companies never actually check. They write the positioning doc once, during the first month on the job, then move on to the next fire. Nobody circles back to ask whether the doc is still true, or whether it was ever true. The doc becomes furniture.
The pain nobody names: positioning gets written once and never tested
A solo marketer at a company this size doesn't have a research team. There's no one to run a message-testing survey or commission a brand-lift study. So positioning gets drafted from founder instinct, a few competitor pages, and whatever copy converted on the landing page last quarter. It gets pasted into a Notion doc. Sales skims it once. Then everyone goes back to what they were doing.
Six months later, nobody can say whether that positioning is working, because nobody defined what "working" would look like. That's the actual problem: not bad positioning, but untested positioning. A statement that's never been checked against real buyer behavior isn't a strategy. It's a guess with a deadline attached.
The fix isn't writing better copy. It's instrumenting three signals that already exist in the business and watching whether they move in the same direction.
Signal one: reply rates on outbound, measured against a real baseline
Positioning shows up first in outbound, because outbound is the fastest feedback loop a solo marketer has. Cold outreach sent with generic, un-targeted messaging typically lands around a 1.6% reply rate. Write that number down. It's the baseline everything else gets compared against.
Outreach built on positioning that actually names the buyer's problem, sent to prospects already showing signal around that problem, should land meaningfully higher. A 3x lift over the cold-email baseline, something closer to 5% reply rates on ICP-matched sends, is a realistic bar for positioning doing its job. If reply rates sit at or near 1.6% no matter how the subject lines get rewritten, the problem isn't copywriting. It's that the positioning underneath the copy isn't describing a problem the prospect recognizes.
This is the fastest test available to a one-person marketing team. No new tooling, no waiting period longer than a two-week outbound sequence. COR's Prospect Scout builds ICP-matched, signal-based lists from the same Blueprint positioning was written from, so this test runs against prospects already primed for the message instead of a random export.
Signal two: sales repeats the positioning back without prompting
The second signal takes longer to observe but is harder to fake. Positioning is working when a rep, on a call with no marketing person present, uses the same language the marketer wrote, unprompted, in their own words, because it's the language that actually closes deals.
Positioning isn't working when sales quietly reverts to its own pitch on every call, or asks marketing to "resend that one-pager" before every demo because nobody retained it. That reversion is a tell. It means the positioning was never adopted, only distributed. A doc that gets forwarded but never internalized hasn't changed anything about how the company sells.
The way to check this without sitting in on every call: pull five recent closed-won deal notes and five closed-lost notes, and read them for language. If the phrases in the positioning doc show up in the notes, it traveled. If they don't, it stayed in the doc.
Signal three: showing up when AI chatbots get asked about the category
The newest signal, and the one most solo marketers aren't yet checking, is whether the company appears when someone asks ChatGPT, Claude, Gemini, or Perplexity to recommend tools in its category. Buyer research increasingly starts in a chatbot window, not a search bar, and a company that's never been asked whether it shows up there usually has no idea what the answer is.
A structured check on this runs the same handful of buyer-style questions, "what tools help with X," "best platform for Y at a company this size," across all four major chatbots, and compares the results side by side against named competitors. Roughly 20 queries across ChatGPT, Perplexity, Gemini, and Claude, with a side-by-side view of which competitors get named and which don't, turns a vague worry ("are we even findable?") into a specific list of gaps. A company that never appears in these answers isn't necessarily positioned wrong. It may just have no content built to answer the exact phrases buyers are typing into these tools, which is a fixable, specific problem, not an abstract branding one. This is the exact gap COR's AI Presence function is built to close, tracking where a company shows up, or doesn't, in AI-generated answers so this signal doesn't stay a blind spot.
One tactic worth naming directly: a single, tightly structured FAQ page of around 600 words, built around one specific query phrase, can close several AI visibility gaps at once if it directly answers the question the way a buyer phrased it. It's a much smaller lift than a full content overhaul, and it's measurable. Rerun the same query set a few weeks later and check whether the answer changed.
Why checking all three matters more than trusting any one of them
A marketer who only tracks outbound reply rates can be fooled by a great subject line that briefly lifts opens without any real positioning improvement underneath it. A marketer who only tracks whether sales repeats the language can be fooled by a persuasive one-pager that reps memorize but buyers never actually respond to. A marketer who only tracks AI visibility can show up in chatbot answers while still losing every deal on a sales call, because visibility isn't the same as resonance.
The three signals corroborate each other, and each one proves something different:
| Signal | What it proves | Bar for "working" |
|---|---|---|
| Outbound reply rate | The message reaches strangers who don't know the company yet | ICP-matched replies ~3x the ~1.6% cold baseline, toward ~5% |
| Sales language | The message survives contact with a live, skeptical buyer | Reps use the phrasing unprompted in closed-won and closed-lost notes |
| AI chatbot visibility | The message exists somewhere a buyer can find it without a rep in the room | Company gets named across a ~20-query set spanning ChatGPT, Claude, Gemini, and Perplexity |
Positioning that's actually working moves all three. Positioning that only feels good in a Notion doc moves none of them.
Quick Answer
A solo marketer at a 30-person SaaS company can tell if positioning is working by tracking three concrete signals together: outbound reply rates (cold-email baselines sit near 1.6%; positioning that resonates should push ICP-matched replies toward 5%, roughly a 3x lift), whether sales reps repeat the same language unprompted on calls, and whether the company appears when buyers ask ChatGPT, Claude, Gemini, or Perplexity to name tools in its category. If none of these move, the positioning has never actually been tested. It's only been written down.
Positioning that hasn't been checked against these three signals isn't a strategy. It's an assumption waiting to be disproven.