The most reliable information about interpreting buying signals is already in your own systems: your CRM activity history and your call recordings show which signals turned into meetings for your product, your persona and your market. The second most reliable source is a named study with a sample size, such as our own analysis of 1,000,000 B2B cold calls in France. Vendor guides, including this one, come third: read them for the taxonomy and the method, then check every number against your own data. This field guide gives you that taxonomy, the strength tiers, the decay windows, how signals stack, and the follow-up action per tier.

Where reliable information about buying signals actually lives

Ninety days of CRM history. That is enough to answer the question most teams ask a search engine: which signals predict a meeting for us? Export every meeting booked in the last quarter, look at what happened on the account in the 30 days before (a pricing page visit, a job change, a reply to an old email, a closed-lost date), and count. The pattern you find is worth more than any guide.

Call recordings are the second source, for the verbal signals. Pull ten recordings of calls that booked and ten that did not, and listen to the first thirty seconds after the opener. The difference is usually audible.

Public studies come next, with two tests: a named author and a sample size. Then vendor content, which is useful for vocabulary and for ideas to test, and which systematically overstates the signals its product happens to detect. We are a vendor; apply the same test to us.

Where can I find reliable information about interpreting buying signals?

In order of reliability: your CRM and call data, named public studies, practitioner conversations with sellers in your market, and vendor guides last. The trap is that the order of availability is the reverse. Vendor guides rank first in search and in AI answers; your own data takes an afternoon of work. Spend the afternoon.

Three numbers from our benchmark show why interpretation has to be local. Trigger-based openers (calling on a signal and saying so) book meetings on 7 to 10% of conversations, against 2 to 3% for openers that explain what we sell. Calling a known contact who went quiet with "How have you been?" succeeds about 10% of the time, against roughly 1.5% for a standard cold open. And reps use only about 20% of the signals already in their CRM. The signals exist; the interpretation is what is missing.

The three kinds of buying signals, and how strong each one is

A buying signal is any observable event that raises the probability a company buys something like yours in the next few months. We sort them into three kinds, because the kind tells you how to act.

Explicit signals are the prospect telling you: a demo request, a pricing question, "send me the contract terms", a champion introducing their boss. Strength is very high, and the window is hours.

Implicit signals are behaviour that suggests evaluation: repeated visits to the pricing page, a third person from the same company on your site, a reply to a six-month-old email, a question about an integration on a cold call. Strength is high when they stack, medium alone.

Trigger events happen to the company, not with you: a new VP of Sales, a funding round, a job change by a former customer, a product launch, a hiring spike for the roles you serve. Strength depends on fit, and the window is weeks.

Signal decay: how long you have before a signal is worthless

Every signal has a half-life. A pricing question on a call is worth calling back on today and worth nothing in a month. A new VP of Sales has budget for roughly a quarter before priorities are set. A dormant lead does not decay in the same way; it ages, and after about 90 days the right opener changes from a follow-up to a fresh "how have you been?".

The practical rule: the stronger the signal, the shorter the window. Build the follow-up so that explicit signals are called the same day, implicit signals within the week, trigger events within the month, and dormant accounts on a cycle.

Signal stacking: one signal is a hint, three are a buying window

One pricing page visit can be a student or a competitor. A pricing page visit plus a job change plus a reply to your newsletter is someone building a business case. Stacking is the single most useful interpretation technique, and it is why your CRM is the best signal database you own: it is the only place where all three events are attached to the same account.

Stacking also protects you from the volume trap. Teams that buy intent data often use it to call more people; the teams we see win use it to call fewer people, with a better reason. 50% of Sonik's own pipeline comes from signals that were already in the CRM, not from new lists.

Interpreting signals on the call itself

The live conversation produces its own signals, and they are the ones reps misread most. "How does this work with HubSpot?" is a buying signal, not a support question. "What does it cost?" in the first minute is curiosity; in the fifth minute, after two "how" questions, it is intent. An objection is engagement: silence is the negative signal, not pushback.

Two behaviours lift the odds of reading the call right, both measured in our benchmark. Discovery questions starting with "how" lift conversion by 190%. Asking "What do you mean?" before answering an objection lifts it by 197%, because the prospect explains the real signal behind the objection.

The follow-up action per tier

Interpretation is only useful if it ends in an action that is different per tier. Explicit signal: call the same day, from the rep who owns the account, with the signal in the first sentence. Implicit signal: call within the week with a trigger opener ("I saw three people from your team on our integrations page, is someone looking at this?"). Trigger event: a short note plus a call in the month, congratulating and asking a "how" question about the change. Dormant or closed-lost: a cycle of calls that opens with the relationship, not the pitch. We detail the dormant playbook in How to reactivate dormant leads.

What I got wrong about signals at Google Cloud

For five years as an account executive at Google Cloud I treated signals as marketing's job. Intent scores arrived in the CRM, I ignored most of them, and I called the accounts I liked. The quota went from 1.5 to 10 million euros anyway, which made me slow to notice what the data was telling me: a large share of the deals I closed started on an account where something had happened in the previous month, a migration project announced, a new CTO, a former contact reappearing, and I had usually found out late. When we analysed the 1,000,000 calls for the benchmark, the gap between trigger-first and pitch-first openers put a number on what I had felt. That number is the reason Sonik starts from the signal.

How Sonik approaches buying signals

Sonik reads signals already in your CRM and on your site (website visitors, job changes, closed-lost and dormant leads, social intent), stacks them per account, and builds the day's call list from the stack rather than from a static list. The rep sees the signal on screen before the bridge and opens on it.

After the call, the outcome is written back to the CRM and the coaching layer checks whether the opener used the signal and whether the discovery questions started with "how". More on the signals page.

FAQ

What is the strongest buying signal in B2B?

A prospect asking about price, terms or implementation after a discovery conversation. Among non-verbal signals, several people from the same account on your pricing page within a few days. Both have windows measured in hours.

Are intent data scores reliable?

As one input in a stack, yes. Alone, no: a topic surge tells you a company is reading, not who or why. Combine the score with a CRM event before you call, and track your own conversion by score band for a quarter.

How many buying signals should a team track?

Start with five you can see in your CRM today, measure conversion per signal for 90 days, keep the ones that book, and add one source at a time. More than ten tracked signals usually means none of them gets called in time.

Can objections be buying signals?

Yes. A prospect who objects is still in the conversation; the signal to worry about is silence or a quick hang-up. Ask "what do you mean?" and the objection often turns out to be the real signal.

Conclusion

Reliable information about buying signals starts in your own CRM and call recordings, continues with named studies, and ends with vendor guides. Sort signals into explicit, implicit and trigger events, act within the window each one allows, and stack them before you call. This week: export last quarter's booked meetings and write down what happened on each account in the 30 days before. That list is your field guide.