The most common mistakes when looking for buying signals are buying third-party intent data before using the signals already in your CRM, treating one weak signal as proof of intent, and acting on a signal a week after it fired. The next four are quieter: openers that never mention the signal, call lists with no prioritisation, reps who miss the live signals the prospect gives them on the call, and teams that measure signal volume instead of meeting rate. We see all seven on our own floor and in the CRMs of the teams we onboard, and each one has a cost we can put a number on.

What are some common mistakes to avoid when looking for buying signals?

Two numbers from our benchmark of 1,000,000 B2B cold calls in France frame everything below. Reps use about 20% of the signals already sitting in their CRM. And trigger-based openers reach a 7-10% meeting rate against 2-3% for pitch-first openers.

So the signal is rarely the problem. The problem is what happens between the signal firing and the rep saying the first sentence. Here are the seven places where that chain breaks, with the cost and the fix for each.

Mistake 1: buying intent data before using the 20% you already own

A sales manager at a 40-person SaaS in Lyon showed us a quote for a third-party intent feed last spring. The same week, her HubSpot held 300 closed-lost deals older than 11 months, 1,200 contacts with no activity in a year, and a website-visitor integration nobody had switched on.

That is the pattern. 75% of the SMB HubSpot users we meet have no intent data at all, and most of them also ignore the first-party signals they already pay for: closed-lost anniversaries, job changes of past champions, dormant leads who reopen an email, repeat visits to the pricing page.

I learned this the expensive way. In five years as an AE at Google Cloud my quota went from 1.5 million to 10 million euros, and the quarters that carried it came from accounts already in the CRM, with a lost deal or a champion who had moved, far more than from the lists we bought. That is the gap I built Salespresso, then Referly, then Sonik to close.

The cost: 50% of Sonik's own pipeline comes from signals that were already in the CRM. Buy third-party data before you have worked that half and you pay twice for the same meeting.

The fix: pull the closed-lost list from 10 to 14 months ago and the contacts with a job change this quarter. Call those for two weeks before you sign anything.

Mistake 2: treating a single weak signal as intent

One person at a 500-employee company read one blog post. A junior analyst followed your company page. A domain appeared once in a topic feed.

None of those is intent. In our experience a lone weak signal converts about like a cold list, and the rep who opens with "I saw you read our article" sounds like a tracker, not a seller.

The cost is the rep's time and the account's patience. The fix is a rule we use on our own floor: a weak signal earns a call only when a second one lands on the same account within the month, or when the contact is someone you already know. Dormant leads prove the point. A known contact called with "How have you been?" reaches about 10% success in our data, against roughly 1.5% for a standard cold open, because the relationship is the signal.

Mistake 3: calling a week late

Signals decay. A demo request answered the same hour and the same demo request answered on Thursday are two different leads. A job change is still fresh after three weeks; a pricing-page visit is cold after three days.

Most teams lose the window to process, not to laziness. The signal lands in a marketing dashboard, gets exported on Friday, is uploaded to a sequence on Monday, and the rep sees it on Tuesday as line 214 of a list.

The fix: decide the response window per signal type before you collect any, and route the fast ones straight into the day's calling block. Inbound and pricing visits today. Job changes and funding this week. Closed-lost anniversaries and renewals this month.

Mistake 4: a signal with no reason to call in the opener

This is the one that hurts the most, because the hard part is already done. The signal is real, the timing is right, and the rep opens with "I am calling from Sonik, we help sales teams make more calls."

65% of reps in our benchmark open by explaining what they sell. The signal was on their screen. It never reached the first sentence.

The cost is the whole gap between 2-3% and 7-10%: the same conversation, three times fewer meetings. The fix is mechanical. Write the opener before the session, one line per signal type: "You posted for two SDRs last week, how are you planning to ramp them?" The signal is the reason to call. The pitch can wait until they ask.

Mistake 5: no stacking, no priority, one flat list

Three hundred accounts, each with one signal, sorted alphabetically. That is what most "signal-based" call lists look like when we open them.

The accounts worth the 10am slot are the ones where two or three signals stack: a new VP Sales plus a job post for SDRs plus a closed-lost from last year. In our experience those accounts book at a multiple of the single-signal ones, and they are rarely more than 10% of the list.

The fix: score by the count of independent signals in the last 30 days, call the top 10% first and at the best slot for the persona (founders around 10am, marketing around 4pm, HR around 1pm), and let the parallel dialer work through the long tail in the afternoon.

Mistake 6: ignoring the signals the prospect gives you live

The pre-call signal gets you the conversation. The in-call signals get you the meeting, and reps who read a script miss them.

When a prospect asks how it works for a team their size, names a colleague, or describes their current process, they are signalling. The answer is a question, not a feature. Discovery questions starting with "how" lift conversion by 190% in our data. When an objection comes, asking "What do you mean?" before answering lifts the outcome by 197%.

The cost: a conversation that was already won, handed back. The fix: listen to your last ten recorded conversations and mark every moment the prospect asked a question about their own situation. Count how many times the rep answered with a pitch.

Mistake 7: measuring signal volume instead of meeting rate

"We detected 4,000 signals this month." Fine. How many meetings?

A signal programme has one output: meetings per conversation, compared with the no-signal baseline, on the same reps and the same slots. A typical Sonik session runs 127 dials, 10 conversations and 0.66 meetings an hour. If the signal list does not beat the cold list on meetings per conversation, the signal is noise, whatever the dashboard says.

The fix: tag every call with its signal type and report meeting rate by tag every Friday. Drop the tags that sit at 2-3%. Double the volume on the ones at 7% and above.

How Sonik approaches buying signal mistakes

Sonik reads the first-party signals in HubSpot or Salesforce (website visitors, job changes, closed-lost, dormant leads, social intent), stacks them per account and builds the day's call list from them, so mistakes 1, 3 and 5 are handled before the rep sits down.

On the contact view, the signal and the account intelligence are on the same screen as the call button, with the suggested opener written from the signal, which is our answer to mistake 4. During the session the parallel dialer keeps the rep in conversations, around 16 an hour against 3 by hand, and the AI coaching flags the live cues from mistake 6.

Every call is tagged with its signal type, so the Friday report in mistake 7 is a filter, not a spreadsheet. Founder-led cold calling on these signals brings 50% of Sonik's own pipeline.

FAQ

What is the most common buying signal mistake?

Paying for third-party intent before using the first-party signals already in the CRM. Reps use about 20% of those, and half of Sonik's pipeline comes from the other 80%. Work the closed-lost list, the job changes and the dormant contacts first; buy data for the gaps you can name.

How fast should you act on a buying signal?

Same day for inbound and pricing-page visits, within the week for job changes and funding, within the month for closed-lost anniversaries and contract renewals. The exact decay varies by signal, but every day in a spreadsheet costs meetings, so route the fast signals into the calling block instead of a sequence.

Can a buying signal be misleading?

Yes, and single weak signals are misleading most of the time. One page view or one follow from a junior profile says little. Two or three independent signals on the same account within 30 days, or one signal on a contact you already know, is what changes the meeting rate.

How do you know if your buying signals are working?

Compare meeting rate per conversation on signal-sourced calls with the cold baseline, same reps, same slots. Trigger-based openers reach 7-10% in our benchmark against 2-3% pitch-first. If your signal calls sit at 3%, the signal is not reaching the opener.

Conclusion

Buying signals fail in the handoff, not in the detection. Most of the seven mistakes above are process: who routes the signal, how fast, into whose opener. This week, export your closed-lost deals from 10 to 14 months ago, write one opener line per signal, and call them during a single two-hour block. Then count meetings, not signals.