Most small sales teams do not have a targeting problem. They have a timing problem. The list is usually fine. The ICP is roughly right. The messaging has been through four rounds of edits. What is missing is any sense of whether the person receiving the email has a reason to care this week rather than some week in the next two years. Send the same message to the same company at the wrong moment and you get silence. Send it three weeks after they hired a head of sales, or the day after they announced a new market, and the same message reads as timely. That gap between a good list and a good moment is what signal-based selling is trying to close. The idea is not complicated. Instead of working alphabetically through a list of companies that match your firmographic filters, you watch for observable events that suggest a company has just entered a buying window, and you prioritise those companies first. The problem is that most of the writing about signal-based selling assumes you have a revenue operations team, a six-figure data budget and an engineer who can wire webhooks between five platforms. If you are a team of four selling into companies under 200 people, that advice is not useful. So this is the version for small teams: which signals are actually worth watching, how to collect them without buying an enterprise data stack, and how to turn them into outreach that does not read like a robot noticed something on LinkedIn.
What counts as a signal, and what is just noise
A signal is an event, not an attribute. "Uses HubSpot" is an attribute. "Switched from HubSpot to Salesforce last month" is a signal. "Is a 60-person logistics company in Queensland" is an attribute. "Just posted three logistics coordinator roles in Queensland" is a signal. The distinction matters because attributes tell you whether someone could buy and signals tell you whether they might buy now. Attributes are how you build the list. Signals are how you order it. The second thing worth being clear about is the difference between signals you can verify and intent data you are asked to trust. Verified signals are things that happened in public and that you can point at: a funding announcement, a job advertisement, a leadership change, a new office, a product launch, a review left on a comparison site, a pricing page visit from a named account. Third-party intent data is a vendor telling you that someone at Acme Corp has been reading about your category, usually based on IP-level inference across a publisher network. Both have a place. But verified signals are cheaper, more defensible and far easier to reference in an email without sounding like you have been surveilling someone. There is a lot of enthusiasm in the market for the returns on this approach, and a fair amount of that enthusiasm comes from vendors with something to sell. Reported uplifts vary wildly depending on who is doing the counting and what they are counting against. Treat the specific numbers with suspicion. The underlying logic holds up without them: contacting a company in the fortnight after something changed will beat contacting them in a quarter when nothing has.
The signals that pay for a small team
Not all signals are equally useful, and the useful ones depend heavily on what you sell. A team selling recruitment software cares enormously about hiring volume. A team selling warehouse hardware does not. So rather than a generic list, here is how to work out which signals matter to you, followed by the ones that tend to travel well. Start by looking at your last ten closed deals and asking a single question about each: what changed in that business in the ninety days before they started looking? Not what pain did they have, because the pain was probably there for years. What changed. Someone new arrived. A contract ended. They lost a big customer. They won a big customer. They opened a second location. An audit went badly. A spreadsheet broke. Do that ten times and patterns show up quickly. Usually three or four change events account for most of your closed business, and those are your signals. Everything else is a distraction you can safely ignore for now. With that said, a handful of signals earn their keep across most B2B categories. Leadership changes in the function you sell to are the strongest single signal in the set, because new leaders arrive with a mandate, a budget conversation and no loyalty to the incumbent tool. Somebody who has been in the seat for six weeks is actively deciding what to keep and what to replace. Somebody who has been there four years already decided. Hiring activity is the next most reliable, and it is genuinely underused by small teams because it takes some interpretation. A company advertising for its first sales operations hire is telling you something specific about the state of its sales operations. A company that has posted eleven roles in a quarter after posting two in the previous year has just had a funding event or a big win, and its processes are about to break in predictable places. Funding rounds are obvious enough that everyone chases them, which is exactly why the window matters. The week of the announcement, every vendor in the category arrives at once. Six to ten weeks later, when the spending actually starts and the noise has died down, is a far better time to be in the inbox. Then there is the category of signals about your own product that most small teams already have and do not use. Repeat visits to your pricing page from the same company. A trial that stalled at step three. A comparison page view. A webinar registration from a business that never showed up. Somebody from a target account viewing three team members' LinkedIn profiles in a week. These are the highest-quality signals you will ever get because the intent is directed at you specifically, and they cost nothing beyond paying attention. Finally, competitor movement is worth watching if you can do it without becoming obsessive. Public complaints, pricing changes, an acquisition, a product sunset, a support outage that spills onto social media. When a competitor makes a decision that annoys its customers, the fortnight afterwards is the cheapest pipeline you will find all quarter.
Collecting signals without buying a data platform
Here is the part where most advice becomes unhelpful. You do not need a platform to start. You need a routine, somewhere to put what you find, and a rule for what happens next. The routine matters more than the tooling. Most teams that try this fail because signal collection becomes a project rather than a habit. Someone spends a fortnight building an elaborate system, uses it enthusiastically for nine days, then stops because it is January and there are more urgent things. A modest system that runs every Monday morning for twelve months beats a sophisticated one that runs for three weeks. The cheapest useful setup is a saved search habit. Job boards let you save searches and email you results. LinkedIn will notify you when people change roles if you follow the right accounts. Google Alerts still works for funding and expansion news, badly but for free. Government business registers publish new registrations and address changes. Industry publications in most verticals still send a weekly email that half the market ignores. Set up eight of these, funnel them into one folder, and spend forty minutes a week reading them with your list open next to you. For product-side signals, your existing tools almost certainly already have the data. Website analytics can be configured to show you repeat visits by company where the traffic is identifiable. Your email platform knows who opened three times and never replied. Your product knows who signed up and abandoned onboarding at the same step as your last four customers. Nobody needs to buy anything to read what they already collect. Where paid tools genuinely help is in two places. The first is contact discovery once you have identified a company worth approaching, because finding the right person and a working email address by hand is slow and demoralising work. The second is scale, once you have proven that a particular signal converts and want to catch every instance of it rather than the ones you happened to notice. Buy for those two problems, in that order, and only after you know which signals matter. Buying a signal platform before you know which signals matter is how teams end up paying monthly for a dashboard nobody opens. This is roughly the sequence Empiraa Signal was built around. Prospect Spark finds and enriches companies that match a defined profile, and the sequencing sits alongside the pipeline rather than in a separate tool, which removes the copy-and-paste step that kills most manual signal workflows. It is worth pointing out that the tool does not decide which signals matter to your business. That part is still yours. Small business sales teams can use the same workflow without adding enterprise software or a dedicated research function.
Storing what you find so it is still useful in three weeks
A signal has a shelf life. A leadership change is interesting for about a quarter. A funding announcement is interesting for about two months. A pricing page visit is interesting for about five days. If your notes do not capture when you saw something, you will end up referencing a "recent" change that happened last March, which is worse than not referencing it at all. So whatever you use to store signals, record three things: the company, the event with a date, and the source. The source matters because you will need to check it before you send. Nothing undermines a good opener faster than congratulating someone on a promotion that turned out to be a job title correction. Beyond that, keep it plain. A single table with a column for the signal type, one for the date observed, one for status, and one for who owns the follow-up is enough for a team of five working a few hundred accounts. Teams routinely over-engineer this, build fourteen custom fields, and then find that nobody fills them in. If your signal record cannot be updated in fifteen seconds, it will not be updated. The one piece of structure worth adding is a decay rule. Anything older than its useful window either gets actioned or gets cleared. A signal list that only grows becomes a graveyard, and a graveyard is indistinguishable from a normal cold list within about two months.
Turning a signal into a message that does not sound automated
This is where signal-based selling most often falls over in practice, and it is a writing problem rather than a data problem. The failure mode is easy to recognise because your inbox is full of it. An email opens by naming a signal, then pivots immediately into a pitch that has nothing to do with the signal. "Congratulations on the Series A. Have you considered how our platform can transform your approach to compliance?" The signal is decoration. Everyone can tell. A signal is only worth mentioning if it changes what you are saying. The test is simple: if you removed the first sentence, would the rest of the email still make sense? If yes, the signal is doing nothing and you should either connect it properly or leave it out. Connecting it properly means naming the consequence, not the event. The event is that they hired a head of sales. The consequence is that the new head of sales will spend their first month trying to work out why the pipeline number in the board deck does not match the pipeline number in the CRM. That is the sentence worth writing, because it demonstrates you understand what the event creates rather than that you can read an announcement. The same applies to hiring signals. Anyone can say "I saw you are hiring three SDRs." The useful version identifies what happens next: three new SDRs means three new people needing a prospecting list on day one, and whoever built the last list is about to be asked to build three more. Keep the ask small and specific to the moment. Signal-triggered outreach is a poor place for a demo request because you are arriving mid-change, when the person is busy and has not yet decided what they need. A question they can answer in one line, or an offer of something genuinely useful with no meeting attached, converts better because it matches the state they are actually in. And send from a human. Signal-based outreach lives or dies on the recipient believing a person noticed something. Anything that breaks that impression, including a sending domain nobody recognises, a footer with four logos or a subject line with their company name awkwardly inserted, undoes the work.
Volume, restraint and the thing nobody mentions
There is an uncomfortable tension in this approach that vendors tend to skip past. Signal-based selling produces fewer prospects than list-based prospecting. That is the entire point. But it means your daily send volume goes down, and if you have a team measured on activity, that is a difficult conversation. The honest framing is that you are trading volume for reply rate, and the maths only works if the reply rate genuinely improves. So measure it properly before you commit. Run signal-triggered outreach as a separate sequence against a control for six weeks, keep the messaging as close as you can between the two, and compare replies and meetings booked rather than opens. If the signal cohort does not clearly beat the control, your signals are wrong, not the approach. There is also a deliverability argument for the lower volume that has become considerably more relevant. Mailbox providers have tightened requirements for high-volume senders, and complaint rate thresholds are now low enough that a broad, poorly matched campaign can do lasting damage to a sending domain. Sending fewer, better-matched emails is no longer only a conversion strategy. It is increasingly a survival strategy for anyone doing outbound from a domain they also use for customer email. The last thing worth saying is about patience. Signals are events, and events happen on their own schedule. In the first month of running this properly, you will not find many. The list of companies with a genuine change event in the last thirty days is much shorter than the list of companies that match your ICP, and that is disconcerting if you are used to a large list feeling like progress. The value compounds. By month four you have a routine, a set of signals you trust, a library of openers that work for each one, and a much clearer sense of what a real buying window looks like in your market. That is a better position than a bigger list.
Four mistakes that show up in almost every first attempt
The first is confusing a signal with a trigger for a template. Teams build one email per signal type, which sounds sensible, then discover they have created a template library that produces obviously templated email. The signal should inform the first two sentences and the specific relevance of what follows. It should not be a slot in a form letter. The second is chasing signals that are visible to everyone. Funding announcements and job postings are public, which means they are public to your competitors too. If you only work signals that anyone can see on a free job board, you are competing on speed against much better resourced teams. The signals with genuine advantage are usually the ones specific to your product, your customers and your category knowledge, because nobody else can see them. Somebody abandoning your onboarding at step four is a signal only you have. The third is letting the signal set the whole agenda. A change event tells you a company is more likely to be receptive. It does not tell you they are a good fit. Plenty of teams end up with a pipeline full of businesses that had an interesting event and no budget, because the signal overrode the qualification criteria. Keep the fit filter first and use the signal to order what passes it, not to override it. The fourth is not writing down what happened. Six months in, the useful asset is not the signal list. It is the knowledge of which signals converted, at what rate, with what opener, and how long the window stayed open. That knowledge only exists if somebody recorded the outcome against the signal type. Most teams do not, which means every quarter they re-learn the same lesson about which signals waste their time.
How this connects to the rest of the funnel
One consequence worth planning for: signal-triggered outreach changes what happens after the reply. A prospect who responded because something just changed in their business is arriving with a live problem and a compressed timeline. They are less interested in a general product overview and more interested in whether you can help with the specific thing that changed. That means your discovery call needs a different shape. Instead of the standard qualification sequence, the first question is usually about the change itself, and the demo, if there is one, should be built around the workflow the change disrupted. Teams that win with signal-based outbound and then hand off to a generic demo script lose most of the advantage they just created. It also changes what a good follow-up looks like. If the window is closing, a four-week nurture sequence is the wrong instrument. Two or three touches inside a fortnight, then a clean exit and a note to revisit in a quarter, respects both the timeline and the relationship.
Where to start this week
If you want one concrete place to begin, do the ten-deal review. Pull your last ten closed-won deals, work out what changed in each business in the ninety days before they engaged, and write the answers in one place. It takes about ninety minutes and it will tell you more about your own market than any amount of reading about signals in general. Then pick the single most common change event from that list and set up one way to watch for it. One saved search, one alert, one weekly habit. Work that one signal for a month before adding a second. Small teams do not fail at signal-based selling because the concept is too advanced. They fail because they try to watch fifteen signals at once, build a system to manage all fifteen, and abandon it in week three. One signal, watched consistently, with outreach that actually references what changed, will outperform that every time.
