Choose prospecting tools by your motion, meaning the channel and the trigger that produce most of your meetings, not by comparing feature lists. A phone-led team, an email-led team and a signal-led team need three different stacks, and buying the "best" tool for the wrong motion is how companies end up with eight subscriptions and five conversations a day per rep. Here is the decision tree we use with French and European teams, and the stack we recommend for each branch.

Why feature lists produce the wrong stack

Feature comparisons assume every team does the same job. They do not. A team selling to founders of 20-person companies lives on the phone at 10am. A team selling to enterprise IT runs twelve-touch email sequences. A team with 5,000 closed-lost deals in the CRM should not be cold-prospecting at all.

Three facts from our analysis of 1,000,000 B2B cold calls in France shape the choice:

• The phone works: 11.3% average connect rate, up to 15.8% at 4pm and above 15% for sales and leadership personas.
• The trigger matters more than the channel: trigger-based openers reach a 7-10% meeting rate against 2-3% for pitch-first openers.
• Known contacts convert far better than strangers: "How have you been?" to a dormant lead reaches about 10% success against roughly 1.5% for a standard cold open.

So the first question is not "which tool" but "which motion produces our meetings today, and which one should".

The decision tree: three questions

Question 1: where did your last 30 meetings come from?

Pull them from the CRM. Tag each one: phone conversation, email reply, inbound or signal (website visit, job change, old deal reactivated), LinkedIn. The biggest bucket is your current motion. Do not guess; teams are often wrong about this by a factor of two.

Question 2: does your persona pick up?

Founders, sales leaders, HR and marketing in SMB and mid-market pick up at 11-20% in our data. Enterprise procurement and IT security often do not. If your persona answers the phone, a phone-led motion will outperform email on meetings per hour. If it does not, email and LinkedIn carry the first touch and the phone comes second.

Question 3: how much do you already know?

Count the contacts in your CRM with a history: closed-lost, churned, dormant leads, past website visitors, contacts who changed jobs. If that is more than a few thousand, you have a signal-led motion waiting. Reps act on only about 20% of the signals already in their CRM, and half of Sonik's own pipeline comes from signals that were already there.

The three stacks, one per motion

Branch A: the phone-led stack

You are here if most meetings come from conversations and your persona picks up.

• Data: mobile-first. Test enrichment providers on 200 real targets and count mobiles found. For France, Germany and the UK, a waterfall of two or three providers beats any single database.
• Dialer: parallel dialing with local numbers, caller ID reputation monitoring, voicemail detection, automatic dispositions, list building by persona and time slot. This is the lever: 5x more conversations per rep, with 20 conversations a day as the target.
• Sequencer: a follow-up layer, not the engine. It sends the email the rep promised on the call and books the next attempt. In our experience, most connects happen within the first three attempts, so three call steps minimum.
• Coaching: inside the dialer. Score openers, "how" questions and objection handling from real calls.

Mistake to avoid: buying an enterprise engagement platform for its sequencing and running calls through its click-to-call. The phone is your engine; buy the engine.

Branch B: the email-led stack

You are here if replies drive most meetings and your persona rarely answers the phone, or if you sell internationally across time zones.

• Infrastructure: secondary domains, warm-up, inbox rotation, sending limits. Deliverability is the whole game; one burned domain costs a quarter.
• Sequencer: a European multichannel tool with LinkedIn steps and GDPR defaults, or CRM-native sequences for small teams.
• Data: verified emails first, mobiles second.
• Dialer: small and targeted. Call the people who opened three times, clicked, replied "not now", or visited the site. A call to someone who just engaged is a warm call, and it converts like one.

Mistake to avoid: scaling volume when reply rates drop. Buyers filter generic email; more of it makes it worse. Add the phone to the engaged segment instead.

Branch C: the signal-led stack

You are here if your CRM holds years of history, or if your website gets meaningful traffic from target accounts.

• Plumbing first: website visitor identification, job-change tracking, closed-lost and dormant-lead views in the CRM. Most of this is configuration, not a new vendor.
• Dialer that acts on the signal the same day: the visit happened at 9:40, the call happens at 10:15 with the trigger on screen. A signal that reaches the rep on Friday's dashboard is a cold call again.
• Sequencer: short, trigger-specific sequences (job change, pricing page visit, anniversary of a lost deal), not generic cadences.
• Third-party intent data: last, and only once first-party signals are exhausted. It is noisier and more expensive than what you already own.

Mistake to avoid: buying intent data before mining your own CRM. The cheapest signals are the ones you already paid for.

Running the choice: a two-week test per branch

1. Pick the branch from the three questions above. Most teams are a primary motion plus a secondary one; buy for the primary.
2. Shortlist two tools for the layer that is weakest in that branch. Not five.
3. Run two reps on each for two weeks, on your CRM, your numbers, your list.
4. Count conversations per rep per day and meetings per rep. Ignore activity counts and dashboards.
5. Write down who owns which data field before go-live. The CRM owns the contact; every other tool writes activities.

How Sonik approaches tool choice by motion

Sonik is built for branches A and C. It is an AI sales dialer that reads the signals already in HubSpot or Salesforce (website visitors, job changes, closed-lost, dormant leads, social intent), turns them into a prioritised calling list by persona and time slot, dials in parallel with local European numbers, and shows the trigger on screen before the prospect answers.

For branch B teams, it is the targeted dialer for the engaged segment: repliers, clickers and visitors get a call the same day. We do not sell data or email sequences; teams keep their existing tools for those layers. See the dialer page and the signals page.

Conclusion

Choose prospecting tools by motion: find where your meetings come from, check whether your persona picks up, count the signals you already own, then buy for the weakest layer of that branch. This week, tag your last 30 meetings by source. The result will probably surprise you, and it will tell you which stack to build.