An intent signal is a single observed action showing that a specific person or company has a need right now; intent data is many such actions, aggregated and scored over time, usually at company level. Vendors swap the phrases, but they answer different questions. A signal tells you who to message today and what to say first. Intent data tells you which accounts researched your category this month. The intent signals vs intent data choice comes down to which question you need answered.
What is the difference between intent signals and intent data?
The difference is granularity and timing. A signal is one event with a date, a source and usually a name attached. Intent data is a trend: it only exists once you have collected enough events to compare this week against last month.
| Intent signal | Intent data | |
|---|---|---|
| Unit | One action ("asked for CRM recommendations on LinkedIn, Tuesday") | A score or stage per company ("Account X surging on topic Y") |
| Names | Usually a person | Usually a company |
| Timing | The moment it happens | Computed over weeks, delivered on the provider's schedule |
| Can you check it? | Yes, open the post or the profile | No, you receive a number |
| First line of the message | Quotes the action | Cannot quote anything specific |
| Where it comes from | Public posts, job changes, funding news, your own site, a reply | Publisher networks, review sites, your own site at volume |
| Typical buyer | A founder or a team of 1 to 20 reps | A revenue team with RevOps and an ABM budget |
| Fails when | You cannot see the action, or you reply late | The score names nobody, or the research was weeks ago |
The lesson on what intent data is covers the account-level versus person-level split in detail. For what a buying signal is, the short version is that a buying signal is one kind of intent signal: the kind that implies a purchase rather than curiosity.
How do you tell which one a vendor is selling?
Ask one question of the sales page: can I quote the action and name the person? If the product shows you something like "head of sales at a 40-person agency commented 'we are comparing tools for this' under your competitor's post at 09:12" (an invented example), that is a signal. If it shows "Acme Corp: Decision stage, High activity" or "Acme Corp: 74 on 'sales intelligence'", that is intent data, however the vendor labels it.
SalesIntel's comparison calls the same split probabilistic versus deterministic: intent data says an account might be researching, a signal confirms something happened. Common Room's article lists the four limits that follow from a modelled score: no identity, no context, delay, and no clear next step. Both are vendors with a product to sell, but the distinction holds up when you check it against what the big providers actually output.
What do Bombora and G2 actually deliver?
The two best-known intent data products are useful here because their own documentation shows how far a score sits from a quotable action.
Bombora's Company Surge compares an account's content consumption over the most recent three weeks against a 12-week baseline for that same account, across more than 25,300 topics, and Bombora's own wording is that "account-level intent shows you where to look" (Bombora, Company Surge). A company counts as spiking on a topic when its score reaches 60 or more, and Bombora's recommended floor for action is that same 60 (Bombora customer documentation). So the output is: this company read more about this topic than it usually does. Not who read it, not what they wrote.
G2 Buyer Intent reports at the buyer-organisation level and, since 2022, gives each company a Buying Stage (Awareness, Consideration or Decision) and an Activity Level (Low, Medium or High), updated daily, from pageviews on G2 plus Capterra, Software Advice and GetApp (G2 documentation). The company name is real and the research is real. You still do not know which of the eight people in that office compared you with a rival, so the first message has to open with a guess.
That gap, from "this company is researching" to "this person said this", is the whole intent signals vs intent data debate in one sentence. A score is a reason to look. A signal is a reason to write.
Why does the gap matter more than it used to?
Because buyers decide earlier and talk to sellers later. In 6sense's 2025 Buyer Experience Report, a survey of more than 4,000 B2B buyers, 94% of buying groups already had a preferred vendor before they contacted any seller, 77% went on to buy from that preliminary favourite, the research-to-seller time split had moved to 60/40, and the average cycle had shortened to about 10 months (6sense, Business Wire, 12 November 2025).
Read that against the two definitions. An account-level score that rises over three weeks tells you a shortlist is being written somewhere inside the company. A signal, a named person asking a public question or engaging with a competitor's post, tells you who is writing it and gives you a reason to be in the conversation before the preferred vendor is chosen. If you cannot name the person, the shortlist closes without you either way.
Where is each one the wrong choice?
Every page on this topic ends with "use both". That advice is only useful if you know which one to drop when you cannot afford both, so here is where each fails.
Intent data is the wrong buy when your deals close in under 90 days, your buyer is one person rather than a committee, or nobody on your team will own the weekly list. The score arrives at company level; someone still has to find the person, write the message and send it before the surge is stale. If that someone is the founder, the contract goes unused by month three. The intent data tools for small business comparison goes through the options and their minimum team sizes.
Signals are the wrong choice when your buyer never acts in public. A hospital procurement team does not comment on LinkedIn posts, so post engagement will find nobody, and you would be better served by job-change and hiring signals or by first-party data from your own site. Signals also need a fast reply: a public question gets its answers within a day or two, so a team that batches outreach weekly loses most of the value.
Public LinkedIn signals, which is what Intentsly watches, are the wrong choice when your buyers are not on LinkedIn, when you sell to consumers, or when you need email or Reddit monitoring, which we do not offer. We watch post engagement, keyword posts, competitor page followers and hashtag engagement on LinkedIn, and nothing else. The LinkedIn buying signals guide shows what you can see with a free account before paying anyone.
Is a LinkedIn comment an intent signal or intent data?
A single comment is an intent signal. It names a person, it has a timestamp, and you can read it. Fifty comments from the same company over a quarter, counted and ranked, would be intent data you built yourself.
That is the useful loophole for a small team: you can produce your own intent data from signals, but you cannot recover signals from someone else's intent data. The score strips the names out. The lesson on first-party intent data explains why public signals sit outside both the first-party and third-party categories: you do not own LinkedIn, and nobody aggregated the comment for you.
Saw your question under Maria's post about which tool handles reply tracking. We hit the same wall last year. Happy to send the two-line answer here, no call needed. What's the volume you're at?
That message is only possible because a signal, not a score, was the starting point. The example above is a hypothetical opener, not one we sent; it shows the shape.
How do you use intent signals and intent data together?
If you have both, the order of operations is: data decides the list, signals decide the message and the day. Pull the accounts that are surging or in Decision stage, then watch the people at those accounts for a public action, and write to the person who acted, quoting what they did.
- Set the account filter first. Companies that fit your ideal customer profile by size, industry and geography. Without this, both signals and scores flood you with people who cannot buy.
- Pull intent data for the list, if you have it. Mark the accounts that are surging or in a Consideration or Decision stage this week.
- Watch for signals at those accounts. Post engagement with competitors, questions that match your category, a new hire in the buying role, a follow of a rival's page.
- Score the person, not just the account. Right role, right seniority, acted in the last 48 hours. Intentsly does this step for LinkedIn: it finds people who engaged with competitor posts or wrote matching posts, scores them against your ideal customer profile, and sends the first LinkedIn message, so the daily watching goes away.
- Write from the signal. First line quotes the action. Second line offers something useful. No pitch until they reply.
- Measure reply rate on signal-led messages against your cold baseline over the same four weeks. That number, not a vendor's, is the one to trust.
If you have only signals, skip step 2 and let the account filter do the prioritising. If you have only intent data, step 4 becomes manual: open the account, find the likely buyer on LinkedIn, check their recent activity, and accept that your first line will be generic.
What did we measure when we ran signals on ourselves?
One first-hand number, labelled as what it is. When Intentsly launched, we had no intent data contract and no list. We ran our own LinkedIn signals and AI SDR on founders who were posting about cold-outreach burnout; the system found the posts, scored the people, sent the first message and handled the replies. The result was 14 qualified demos in 14 days with zero manual outreach, and a reply rate 4x our previous cold campaigns, with most leads coming from signals caught in real time (case study).
That is one company over two weeks, selling a sales tool to founders who live on LinkedIn, which is the easiest possible case for public signals. Treat it as a before-and-after shape to copy, not a benchmark to forecast from.
Which one should a small team use?
The rule in one minute:
- Under 200 employees, deals under six figures, buyer is one or two people, buyers active on LinkedIn: work from signals. Start with the free methods in the signal-based selling lesson, then automate the watching when the daily searching eats more than an hour.
- Deals over six figures, committees of five or more, cycles over six months, a RevOps owner: intent data earns its contract, because you need to know which accounts to work before anyone acts in public. Add signals on top for the message.
- Buyers who never act in public: neither. Use first-party data from your own site and product, plus hiring and job-change signals from company pages and job boards.
The free Buying Signal Playbook lists the 12 buying signals worth chasing and the 8 to ignore, so you can see which ones your buyers actually produce before you pay for anything. If your buyers are on LinkedIn, try the AI shows you matching people with live signals from this week, no account needed, which is the fastest way to find out whether signals exist for your market at all.