Signal-based selling replaces the contact list with a trigger: an observable event showing that a company has just entered a buying window. Three families matter (company, technology, engagement) and they are not equal: a signal is judged on its freshness, its scarcity, and how close it sits to your problem. A signal acted on within 48 hours is worth ten acted on three weeks later.
Signal-based selling is a sales approach in which every first contact is triggered by an observable event at the prospect’s company, a buying signal, rather than by the prospect’s presence on a list. Applied to prospecting, it is usually called signal-based outbound: the signal decides which account you contact, when you contact it, and what the first message talks about.
Mass outbound is not dying for moral reasons. It is dying because it no longer produces: reply rates are collapsing, inboxes are armoured, and B2B buyers have learned to recognise an automated sequence from its first line. The usual reflex, raising volume, accelerates the decline rather than slowing it.
Signal-based selling inverts the logic. Instead of starting from a list and hoping to land at the right moment, you start from an observable event showing that a company has just entered a buying window. Volume goes down, yield goes up. Here is how such a machine is actually built.
The problem is not volume, it is timing
At any given moment, a very small share of your market has an active project. Everyone else has no budget, no priority, or no felt problem. Sending the same message to all of them means paying to be ignored by the majority, and burning your deliverability and your reputation along the way.
The right message at the wrong time is still a bad message. That is why the best written sequences plateau: they try to convince someone who has nothing to decide. A signal changes the nature of the approach, because it makes you arrive while the question is being asked, not before and not after.
The three families of signals
Company signals describe a change of state: a funding round, a new executive, a market opening, rapid headcount growth, an announced reorganisation. They are public, easy to capture, and they open long windows, from several weeks to several months.
Technology signals describe the prospect’s environment: adoption of a tool that complements yours, a migration under way, a competitor being dropped, job postings naming a specific technology. These are the most underused, although they are often the most predictive: a company hiring a dedicated profile for a topic has just decided to invest in it.
Engagement signals come from your own side: repeated visits to high-intent pages, a look at the pricing page, a mid-funnel piece of content opened, interaction with your posts. These are the warmest, and they are precisely the ones most teams leave sleeping inside a tool without acting on them.
Examples of buying signals
The families stay abstract until you name the events. Here are the signals we see most often in B2B, with what each one may indicate and where to find it. None of them is a purchase intention on its own: each is a hypothesis to check.
| Signal | Family | What it may indicate | Where to spot it |
|---|---|---|---|
| Job posting for a role tied to your problem | Technology | A budget has been decided on the topic | Careers pages, job boards |
| New executive in the buying function | Company | Suppliers and tools about to be reviewed | LinkedIn, press releases |
| Funding round | Company | Money available, not yet a need | Press, funding databases |
| New office, country or business unit | Company | New processes and tools to set up | Press, legal announcements |
| Merger or acquisition | Company | Systems and teams to consolidate | Press, company registries |
| Adoption or removal of a tool next to yours | Technology | A migration, or a gap you can fill | Job postings, technology detection |
| A former customer contact joins a new company | Engagement | A buyer who already knows your value | LinkedIn job changes |
| Repeat visits to your pricing or comparison pages | Engagement | An evaluation under way | Your web analytics |
| An old proposal reopened | Engagement | A shelved project coming back | Your proposal or e-signature tool |
| An executive posting about the problem you solve | Engagement | The pain is felt and spoken about | LinkedIn, podcasts, conferences |
For companies selling into the public sector, published tenders play the same role. In France they appear on the BOAMP, and legal announcements such as registrations, transfers or mergers appear on the BODACC, both available free of charge through official APIs.
Not all signals are equal
This is what most implementations get wrong. Everything gets collected, sources pile up, and the team ends up facing a queue it no longer works. A signal should be judged on three criteria.
Freshness. A signal from this week is worth ten times the same signal from last quarter. Beyond three or four weeks the window has usually closed: someone else has been consulted, a provisional decision has been made.
Scarcity. If ten competitors watch the same funding announcement, the prospect will receive ten near-identical messages in the same week, and your signal loses all its value. The signals that pay best are the ones your market is not watching yet.
Proximity to your problem. A funding round indicates money, not a need. Hiring a revenue operations lead indicates a precise need. Always prefer the specific signal to the generic one, even when it fires less often.
Detection: start without a tool
The first mistake is buying an intent data platform before validating which signals matter in your market. The tool will then deliver a volume of data nobody knows how to qualify.
Start with public, free material: job postings, which give away almost every company priority, funding announcements, leadership changes, press releases, decision-makers’ posts. Half a day a week, a spreadsheet and two or three alerts are enough to produce the first twenty conversations.
Tooling comes next, once you know what you are looking for and manual volume becomes the limiting factor. It assumes a clean data foundation, without which you will get alerts that cannot be tied to any account or open deal, a point we cover in building your revenue operations foundation.
The 48-hour rule
A signal is a fresh product. Between the moment the event becomes public and the moment the window closes, you have a few days, sometimes a few hours on engagement signals.
That imposes an organisation, not just a tool. Someone has to look at the queue every day, not on Monday morning. The sequence has to be written before the signal arrives, not improvised when it lands. And the opening message should fit in three lines that name the event, state a hypothesis about what it implies, and propose a short conversation. Nothing about your product: the signal is the subject, you are only the context.
A signal-based machine that takes two weeks to react does not outperform classic prospecting. It simply costs more.
Signal-based outbound, step by step
Signal-based outbound is the prospecting side of the method. It runs as a short chain of five steps, and each step needs an owner, otherwise the chain breaks at the first handover.
1. Detect. An alert, a data feed or a person spots the event and logs it against an account in the CRM, with its date and its source.
2. Qualify. Is the account in your target? Is the signal fresh, scarce and close to your problem? If not, drop it. A discarded signal costs nothing, a signal worked badly costs reputation.
3. Route. The account goes to one named person, with a deadline counted in hours. A signal nobody owns is a signal lost.
4. Fire the sequence written for that signal. Each signal has its own sequence. The structure stays the same, the subject changes: the opening message names the event, offers a hypothesis about what it implies, and asks a short question. Your product comes later, if it comes at all.
5. Log the outcome. Reply or silence, meeting or not, recorded against the signal type. After a quarter you know which signals open conversations and which only produce noise, and you stop watching the second kind.
A concrete case. A target account publishes its first revenue operations job posting. The rep reacts the same day: a three-line message that mentions the posting, suggests that this kind of hire often comes when leadership can no longer read the pipeline from the CRM, and asks whether that is the trigger here. No deck, no demo request. If the hypothesis is right, the conversation starts on the prospect’s project. If it is wrong, the prospect usually says why, which is information too.
List-based outbound vs signal-based outbound
| List-based outbound | Signal-based outbound | |
|---|---|---|
| Entry point | A list of contacts | An event at the account |
| Timing | Your sending calendar | The prospect’s calendar |
| First line | Your product or your company | The event that triggered the contact |
| Volume | High and constant | Lower, and it varies with the signals |
| What you measure | Replies per email sent | Conversations per signal worked |
What it changes in the numbers
Send volume drops sharply, often by a factor of five to ten. Reply rate rises, because the message lands while the question is live. The cycle shortens, because the prospect was already in motion. And your brand is preserved, which is a long-term asset that is hard to rebuild once damaged.
A less expected effect matters just as much: the conversation changes in kind. A rep who enters through a signal talks about a live project, not about a product. They discover faster, qualify better, and document information the rest of the team can use. That is also what makes the approach transferable, provided it is written down, as we describe in a sales playbook that survives your best closers leaving.
The mistakes that kill a signal-based machine
The most common is keeping the volumes of the old model and adding a layer of signals on top. What you get is lightly personalised mass prospecting, with the costs of both approaches and the benefits of neither.
The second is watching too broad a segment, which drowns the relevant signals in an unmanageable queue. Signal-based selling assumes a tight perimeter, which is one more reason to settle the targeting question, as we cover in why an ICP that is too broad kills your growth.
The third is believing the signal replaces the message. It determines the moment, not the relevance: if your value proposition does not name a precise problem, arriving at the right time achieves nothing.
The Arkimoz method
We start by mapping the events that genuinely preceded a purchase among your recent won deals. This retrospective analysis eliminates the seductive but non-predictive signals, and it almost always surfaces one that nobody was watching.
We then tool up detection, starting with free sources, and write the sequence specific to each signal along with the daily routine that guarantees the reaction time. The deliverable is not a contact list: it is a system that tells you when it is time to talk, and knows what to say at that moment.
Where to start
Take your last twenty won deals and look, for each one, at what happened at their company in the three months before the first conversation. A pattern almost always appears. Pick the two most frequent signals, write a sequence for each, and work the queue every morning for a month.
After four weeks you will know whether the pattern holds, and you will have spent time rather than a tooling budget.
Frequently asked questions
- What is signal-based selling?
- It is an outbound approach where every first contact is triggered by an observable event at the prospect's company, rather than by their presence on a list. The signal can be a funding round, a hire, a tooling change, or repeated visits to a high-intent page. The opening message refers to that event, not to your product.
- What is signal-based outbound?
- It is signal-based selling applied to prospecting. An account only enters an outbound sequence when a qualified signal fires, such as a relevant job posting, a new executive or repeated visits to your pricing page, and the first message is about that event. Volume becomes a consequence of the signals you detect, never a target set in advance.
- What is the difference between signal-based selling and intent data?
- Intent data is raw material: bought or observed intent signals, usually aggregated at company level. Signal-based selling is the system that turns that material into conversations: choosing the signals that actually predict a purchase in your market, detecting them, and running the sequence each one triggers. Buying intent data without that system just produces another list.
- Do you need a paid tool to get started?
- No. The most profitable signals early on are public and free: job postings, funding announcements, leadership changes, press releases, executives' own posts. A spreadsheet, two or three well-configured alerts and half a day a week are enough to validate the approach before spending on detection software.
- How many signals should you track?
- Two or three, no more. A machine watching fifteen poorly qualified signals produces noise, and nobody works the queue. Pick the two events that genuinely preceded the purchase for your last twenty won deals, and build a sequence for each before adding a third.