Buying signals, in writing.
And their lookalikes.
Published · every example on this page is written for it; none are real posts
A buying signal is something a person says or does that shows they are considering a purchase: asking what something costs, naming a deadline, comparing two products. Sales training teaches reps to hear them on a call. The same signals now arrive in writing, earlier and in far larger numbers, in public threads, replies, emails, support tickets and chat. Writing makes them easy to collect and hard to read, because a buyer’s words are also the words of a joke, a complaint and an advert.
Guides such as Cognism’s rank buying signals by strength, and most of the signals they rank are behaviour: demo requests, page visits, webinar sign-ups. This page covers the written ones, strongest first, then the signals that point the other way, and the lookalikes that fool keyword alerts.
Written buying signals, strongest first
Replacing a named tool
“Our CRM contract ends in June and we’re not renewing. What did your team move to?”
A current tool, a decision already made and a date. The question is no longer whether to buy, only what.
A deadline
“We need online booking live before the clinic opens on the 1st. Recommendations?”
Timing is the hardest thing to infer and the easiest to state. A date turns interest into urgency.
Budget or sign-off
“Finance approved about $300 a month for this. What’s actually worth it at that price?”
Someone with a budget can buy this month. Someone without one is still building the case.
A shortlist
“Down to two tools for our onboarding emails. Has anyone used both at around 20k contacts?”
Research is over and evaluation has started. The open question is who wins.
How to switch
“How long did it take you to move your help articles over? We have about 400.”
Questions about migration, imports and integrations picture ownership. Nobody asks how to move in unless they are moving.
A trial in progress
“Day four of the trial. Is the paid tier worth it for a team of six?”
Already inside a product and weighing the price. The decision is days away.
Requirements
“Need something that invoices in three currencies and syncs with our accounting software. Under $100 a month.”
A specification, not a daydream. Strong once a timeline is attached; on its own it can still be research.
Price and plans
“Is it per seat or flat? We’re 25 people.”
Interest, but pricing questions also come from the curious, from people writing comparisons, and from complaints in the making.
Asking for proof
“Anyone running it with 50 agents? Does it hold up?”
Evaluation with a team size attached. Usually the step before a shortlist.
General pain
“Our onboarding emails are a mess.”
Aware of a problem, not shopping for a fix. Worth a helpful answer; a pitch would be early.
An open question
“What does everyone use for customer surveys?”
Could be someone about to buy, someone curious, or someone writing a listicle. The rest of the thread decides.
Signals stack. A price question on its own is weak; a price question with a team size and a renewal date is a buyer. Read the whole message, and the thread around it, before judging any one phrase.
Leaving is a buying signal too
“Third billing error this quarter. If it happens again, we’re out.”
To a competitor, that is a strong buying signal. To the vendor being left, it is a churn signal, and it usually arrives in a support ticket. The same reading catches both: churn detection.
Negative buying signals
Negative buying signals say the opposite: not now, or not you. In a live deal they are often silent: slow replies, meetings that slip. In writing they are usually explicit, which makes them the easiest signals to act on. Stop, or note when to come back.
“We just signed a two-year contract, so we’re set for now.”
Locked in. Worth noting the date, not a reply.
“No budget for this until next year.”
Not now. A date to come back, not a lead.
“Honestly, the spreadsheet works fine for us.”
No felt pain, so nothing to sell into.
“Just curious what people use. Not looking to switch.”
Said plainly. Take them at their word.
The lookalikes
Some text has the vocabulary of a buying signal and none of the intent. A keyword alert forwards all of it. These are the patterns that fool alerts most often.
“Anyone hiring a CRM admin? Five years’ experience, open to remote.”
Job seeker Names the tool and wants a job, not a product.
“Writing my dissertation on help-desk adoption. What do small teams use?”
Research A real question with no purchase behind it.
“We switched email tools last spring and haven’t looked back.”
Already bought Past tense. The decision is made; this is a review.
“Oh good, a fourth dashboard to check every morning. Just what I needed.”
Sarcasm A complaint dressed as enthusiasm.
“If we ever get to 100 customers I guess we’d need a real CRM?”
Hypothetical Someday, not now.
“What’s everyone using for social scheduling? I’ve been trying one that’s great, happy to share the link.”
Seller asking The question is there to set up a pitch.
“Looking for a payroll provider for 15 staff in two states. Budget approved.”
Wrong product Every strong signal at once, for payroll. Worth nothing unless you sell payroll.
The last one is the hardest, because nothing about it is fake. In the 158 public posts our own scorer put forward and a human reviewer then rejected, 32 were genuine buyers of a different product, alongside 65 too vague to act on and 61 about something else. A buying signal only counts relative to what you sell.
Verbal and written buying signals
Swipe the table sideways to compare.
| On a call | In writing | |
|---|---|---|
| Who they are talking to | You | Peers, a support desk, or nobody in particular |
| When | Late: they are already in your pipeline | Early: often before they have contacted anyone |
| How much you can tell | Tone of voice, pauses, and a follow-up question | One message, no tone, usually no follow-up |
| How many | A few calls a week | Thousands of posts, emails and tickets |
| What a misread costs | An awkward moment on a call | A public reply to someone who was joking |
Spoken signals are rarer and easier to read. Written ones come earlier and in far greater numbers, and a misreading happens in public. HubSpot’s guide covers the spoken side well; the written side is where the volume now is.
How to respond to a written buying signal
- Answer the question they asked, first. Someone asking about migration wants a migration answer, not a demo link.
- Match the reply to the strength. A strong signal earns a direct answer and a link. A weak one earns a useful answer and no pitch.
- Reply where they asked, as yourself. A public question gets a public answer that says who you work for. Moving an unsolicited pitch into private messages is the quickest way to be reported.
- Be quick. Buying signals go stale. The person who needs booking live by the 1st will have chosen by the 2nd.
- Keep a person in the loop. Anything sent in your name should be something you would have written yourself.
Spotting buying signals in text at scale
Reading every post, email and ticket by hand stops working at a few dozen a day. Keyword alerts catch the words and forward the lookalikes with them. A single LLM prompt catches more, and is just as confident when it is wrong. A panel of three judges, a Skeptic, an Analyst and an Optimist, reads each borderline message independently. When they agree, you can act; when they split, you see it, so you read only the cases that need a person.
Your agent, bot or scraper already has the text. It can send it to SignalPipe’s API or MCP server and get back a verdict, and whether the judges agreed. How that compares with keyword alerts and single prompts, and how buying signals relate to intent data, is on intent signals vs intent data.
Test a buying signal.
Paste a post, email or ticket into the live panel and see whether the judges call it a buyer. Free, and no account needed.
Questions
What is a buying signal?
Something a person says or does that shows they are considering a purchase, such as asking about price, naming a deadline, comparing two products or asking how to switch. In sales conversations they are spoken. They increasingly appear in writing first: in public posts, emails, support tickets and chat.
What are examples of buying signals?
Asking what others moved to from a named tool; stating a renewal date or a launch deadline; mentioning an approved budget; comparing two options; asking how to migrate or integrate; being partway through a trial and asking whether the paid plan is worth it. Pricing questions and requests for proof are weaker on their own.
What are verbal buying signals?
Statements a prospect makes in conversation that show purchase interest: questions about price, implementation, timelines or contract terms, or complaints about a current vendor. Their written equivalents appear in posts, emails and tickets, usually earlier and without the tone of voice that helps you read them.
What is the strongest buying signal?
A stated timeline together with a reason to act, such as a contract ending on a known date after a decision to leave. A specific migration question or an approved budget comes close. Two signals together are stronger than any one alone.
What are negative buying signals?
Signs that a sale is unlikely or not yet possible: a contract with a competitor just renewed, no budget until next year, satisfaction with the current setup, or saying plainly that they are only curious. In writing they are often explicit, which makes them easy to act on: stop, or note when to come back.
How do you identify buying signals in text?
Look for specifics: a named tool, a date, a budget, a team size, a shortlist, a migration question. Then rule out the lookalikes: sarcasm, venting, hypotheticals, job posts, research questions, past purchases, sellers asking leading questions, and real buyers of a different product. At volume, a panel of independent judges that says when it disagrees tells you which verdicts need a second look; a keyword or a single prompt does not.
Are buying signals the same as intent data?
No. Intent data is aggregated from what people at a company read and research, usually reported per account. A buying signal in writing is one person saying what they need, in their own words. The intent signals page on this site compares the two in detail.
Sources
- HubSpot: Buying signals in sales
- Cognism: 13 buying signals and how strong they are
- SignalPipe: Buying-intent hard negatives, CC BY 4.0
- SignalPipe: Intent signals vs intent data vs what people say