
Affordable sales coaching starts inside your own team: manager call reviews, peer coaching and a library of recorded calls cost nothing but time, and in our experience they change more behaviour than most paid programs. When you do spend, the cheapest effective options are AI call scoring billed per seat and a freelance coach booked by the day; cohort programs and instructor-led methodology courses sit at the expensive end. The way to compare any of them is cost per behaviour changed, not price per seat, and the behaviours worth paying for are known: in our benchmark of 1,000,000 B2B cold calls in France, a trigger-based opener books meetings at 7-10% of conversations against 2-3% for a pitch.
Six places, three of them free, ranked by what they cost.

Coaching is listening to a call and changing one behaviour. The listening is free if the dialer records. The changing is a 15-minute conversation. What you buy, when you buy, is coverage (someone or something listens to more calls than you can), practice (a safe place to try the new sentence) and outside eyes.
A weekly 15-minute session per rep: one recorded call, one behaviour, one commitment. Plus the manager sitting in the 4pm block, when connect rate peaks at 15.8% in our benchmark against 9.8% at 8am, listening live to three calls per rep.
This is the coaching I received at Google Cloud. Five years as an AE, quota from 1.5 million to 10 million euros, and the entire coaching budget was my manager's Thursday afternoon. One call a week, played back, with one question: "what would you do differently?" It cost the company nothing and it is still the most valuable thing a manager has given me.
Good for: any team with a manager and a dialer that records. Limit: the manager hears about 2% of calls and picks which 2%.
Every Friday, each rep brings one call they are proud of and one they are not. The team listens to two or three. The rule: the rep who made the call speaks first, then peers say what they heard, then the manager. Thirty minutes.
Good for: teams of four or more, where the best opener on the floor can be heard by everyone. Limit: it drifts into storytelling without a scorecard; keep it to five behaviours (opener, "how" question, objection move, close, slot).
Tag the ten best calls on the floor by behaviour: three trigger openers, three "how" questions, two objection moves answered with "What do you mean?", two closes with a slot proposed. New reps listen to the library before their first block. It takes a manager one afternoon to build and ten minutes a month to refresh.
Good for: ramp. In our experience a new SDR who hears ten good openers before their first dial pitches less; without it, they join the 65% of reps who open by explaining what they sell. Limit: a library shows what good sounds like, it does not tell a rep what they did on their own call.
Software that transcribes and scores every call against your scorecard, billed per rep per month, ideally inside the dialer. What it buys is coverage: 100% of calls scored instead of 2% replayed, and the manager's 15 minutes spent on the moment that matters. Price drivers: number of seats, whether scoring works in your language (French, for us), whether it is bundled with the dialer or a separate tool.
Good for: teams of five or more reps where the manager cannot listen enough. Limit: a score is not coaching; someone still has to sit with the rep.
An experienced sales leader who comes in for a day, listens to calls, runs sessions, and leaves a one-page plan. Billed per day or half-day. Price drivers: the coach's track record, whether they prepare from your recordings beforehand, whether they come back monthly.
Good for: a team without a manager who sells, or a manager who wants outside eyes once a month. Limit: a day changes the manager more than the reps; the follow-through is yours.
Instructor-led courses, per participant, over several days or weeks, usually teaching a methodology. Price drivers: group size, live or recorded, certification, whether managers are trained to reinforce it afterwards.
Good for: a team adopting a new approach at once. Limit: knowledge fades without weekly reinforcement, and the program rarely hears your reps' real calls.
We do not publish price bands, because they move every quarter and because the number that matters is not the price. The four cost models are stable: per seat per month for software, per day for a freelance coach, per participant for a cohort program, and by quote for anything enterprise.
What drives the price inside each model is predictable. Seats: more reps, higher total, lower per seat. Language: scoring in French or German costs more than English-only, and is worth it. Human time: anything with a person in the room costs more than anything without. Reinforcement: a program that trains your managers to coach costs more up front and holds longer.

Compare on cost per behaviour changed. Take the yearly cost of the option, divide it by the number of reps who, four weeks later, do the coached behaviour on most of their calls. A free Friday ritual that moves three reps from pitch to trigger opener beats a cohort program that moves none. Set the baseline before you buy: count the pitches in each rep's last ten calls, then count again a month in.
Our own practice: Sonik's sales team trains three times a week, on our own recordings, with the AI scoring inside our dialer and no external program. Our coaching spend is the manager's time.
In France, every company is attached to an OPCO, the sector skills operator that can fund part of professional training. Sales training delivered by a registered training organisation can be eligible. We will not quote amounts or conditions here, they depend on the sector, the company size and the current rules, but the question costs nothing: ask your OPCO adviser whether a sales coaching or cold-calling program is eligible and what the provider needs to supply. In our experience most sales managers never ask. One practical point: software billed per seat is rarely treated as training; a coach or a cohort program delivered by a registered organisation is worth asking about.
Zero budget, one manager, one dialer that records. Four weeks, one behaviour per week, measured on calls.

Week 1, baseline. Pull each rep's last ten connected calls, write down the first sentence, tag where the call died. Count the pitches. Build the library from the best calls you hear.
Week 2, the opener. Each rep gets one trigger per prospect from the CRM before dialing. Target: 80% of calls open on a trigger. The gap you are chasing is 7-10% meeting rate against 2-3%.
Week 3, the question. Three "how" questions per persona, closed questions banned in the first minute. Our data: +190% conversion.
Week 4, the objection and the close. "What do you mean?" before any answer (+197%), and one slot proposed, not "when suits you?". Friday: count all four behaviours across the week's calls and compare with week 1.
Throughout: the 4pm block is protected, the manager listens live to three calls per rep per block, and Friday is the call of the week. That is the whole plan. It costs four afternoons.
Sonik is an AI sales dialer, and the coaching is part of the seat rather than a separate line on the invoice. Every call made through the dialer is transcribed and scored, in French, on the five behaviours above, so a manager with no coaching budget gets 100% coverage of the floor's calls.

During a block the manager sees every live call and its status, can listen in on any of them and prompt the rep. After it, each rep's page carries one behaviour to work on, chosen from their own calls, and the trend on that behaviour week over week.
The 30-day plan above runs as is on Sonik; week 1's counting takes the AI minutes instead of an afternoon.
Yes, when it is specific. A weekly 15-minute session on one recorded call, with one behaviour and one commitment, changes what a rep says within a month. What makes free coaching fail is vagueness ("be more confident") and inconsistency (skipped when the quarter gets loud). The behaviours with the largest measured effects in our data, the trigger opener, the "how" question and "What do you mean?", need no software to coach.
Freelance coaches bill by the day or half-day, and the rate depends on their track record, on whether they prepare from your recordings and on whether they come back monthly. Ask for a one-page plan and a baseline count after the first day; judge the second day on whether the first changed a behaviour, not on the rate.
Possibly. French companies are attached to an OPCO that can fund part of professional training, and sales training delivered by a registered training organisation can be eligible. Conditions and amounts depend on the sector and the company, so ask your OPCO adviser directly before signing with a provider. Software seats are rarely treated as training.
A recorded-call library of ten good calls tagged by behaviour, listened to before the first block, plus the manager live in the new rep's first 4pm sessions. It costs an afternoon to build. In our experience the fastest ramps are the reps whose first 50 connected calls were listened to.
Affordable sales coaching is mostly free: manager rituals, peer reviews and a call library cover the behaviours with the largest effects in 1,000,000 cold calls. Spend on coverage (AI scoring per seat) or outside eyes (a coach by the day) only once the free tier runs weekly without fail, and compare every option on cost per behaviour changed. This week, count the pitches in each rep's last ten calls. That number is your baseline and your business case.