A founder of a 60-person B2B software company forwarded me a proposal in February. Sales training program, twelve reps, two days on site plus a quarter of follow-up sessions, $58,000. His note was two lines: "Win rate is down from 24% to 17% and the team feels rusty. Worth it?"
I asked him which stage the deals were dying in. He did not know. Nobody on his team knew, because their CRM had six pipeline stages and four of them were named after internal handoffs rather than buyer commitments, so every stalled deal sat in "Proposal" until someone closed it out at quarter end.
We spent two weeks pulling apart eighteen months of closed-lost deals. The win rate had not dropped because reps got worse at selling. It dropped because a competitor had cut their entry price and 40% of the losses were now happening at the pricing conversation, against one specific competitor, in one specific segment. That is not a training problem in the general sense. That is a two-hour objection clinic, a battlecard, and a pricing decision the founder had been avoiding.
He spent about $9,000 with a specialist instead of $58,000 on a general program. Win rate came back to 22% over the following two quarters. This post is about how to run that diagnosis yourself before you sign anything.
What a sales training program actually buys you
Strip the category down and there are only four things being sold, usually bundled and priced as one.
Most quotes I read are heavy on one and two, light on three, and almost silent on four. That ordering is backwards. The fourth item is the only one that determines whether the first three survive past the workshop, and it is the one item a vendor cannot deliver for you, because it lives inside your CRM and your managers' calendars.
The numbers cut both ways, and both sides are honest
I want to be fair to the category, because good sales training genuinely works and I have seen it work.
The $4.53 figure comes from ATD and CSO Insights benchmark work and gets quoted in nearly every vendor deck. It is real. So is the Gartner finding that organisations with formal enablement and embedded reinforcement report meaningfully higher attainment than those running one-off events.
And so is the third number. Only about 28% of B2B companies believe their training program had a significant effect on results. Those two facts are not in conflict. Training pays back beautifully when the diagnosis was right and the reinforcement existed. It returns nothing when a team bought a generic two-day event to fix a problem that was never a skill problem.
The forgetting data is the part founders underestimate. Without reinforcement, reps lose roughly 70% of what they learned inside a day and about 84% inside three months. Behaviour reverts inside 90 days. If you buy two days of training and change nothing about how the week runs afterwards, you have bought a morale event with a slide deck attached.
Share of training content reps lose within three months when nothing in the daily workflow reinforces it. This is the single number that should drive how you structure the spend.
Most skill gaps are not skill gaps
Here is the pattern I keep running into. A number drops, the number is a sales number, so the fix gets labelled a sales skill problem. Four out of five times the CRM says otherwise.
There are really four different diseases that all present as "the team needs training":
A targeting problem. Reps are working accounts that were never going to buy. Win rate falls, cycle length rises, and no amount of objection handling fixes it. This shows up as high loss rates concentrated in one segment, or a lot of losses to "no decision". If your ICP definition is a paragraph on a slide rather than a filter that runs against your database, start there.
A process problem. The stages do not mean anything, so managers cannot inspect deals, so nobody catches a deal going sideways until it is dead. This is what the founder above actually had. Fixing pipeline stages so each one is a buyer action rather than a rep intention costs a fortnight and no licence fee.
A market problem. Pricing changed, a competitor moved, budgets tightened. Training does not repair this, though targeted competitive work helps at the margin. Your win/loss analysis will separate this from the others faster than anything else.
An actual skill problem. Reps genuinely cannot run a discovery call, cannot hold price, cannot get past a gatekeeper. This exists and it is worth paying to fix. It is just less common than the invoice total across the training industry would suggest.
Training is the most expensive way to fix a problem that was never about skill.
Two weeks in your own CRM data will tell you which of the four diseases you have. Do that before you take a single vendor call, because the diagnosis also tells you what to buy.
The diagnosis: five queries to run before you take a vendor call
None of this needs a tool you do not already own. HubSpot, Salesforce, Pipedrive and Attio all do this natively, and if your data is messy enough that these break, that itself is your finding.
Loss reason distribution by stage. Pull twelve months of closed-lost deals. Group by the stage they died in. A team with a real skill gap loses everywhere. A team with a targeting problem loses early and in volume. A team with a process problem has a graveyard sitting in one middle stage. If your loss reasons are all "other" or blank, your first project is capture, not training, and that is a CRM data quality job.
Win rate by rep, against the median. Compute win rate per rep on deals that reached qualified stage. Now look at the spread. If your top rep converts at 31% and your bottom at 28%, the whole team has the same ceiling and something systemic is capping them. If it runs 34% to 9%, you have specific people who need specific help, and a general program aimed at everyone is the wrong shape of spend.
Stage conversion, ranked by drop. Take every stage-to-stage conversion rate and sort by the biggest fall. The worst one names your training brief. Discovery to demo falling off a cliff means qualification. Proposal to closed-won means pricing and negotiation. First meeting to discovery means the meeting is being booked on a promise the rep cannot land.
Activity against outcome. Not activity for its own sake. Compare the rep-level activity pattern of your top third against your bottom third. Are the low performers doing less, or doing the same and converting worse? Those are two different fixes and only one of them is training. Your existing sales KPIs should already answer this if they are wired correctly.
Ten call recordings from lost deals. Manual, unglamorous, and the highest-yield hour in the whole exercise. If you run conversation intelligence through Gong or HubSpot's own call tools, filter to closed-lost and listen. You will hear the actual gap in about forty minutes. Talk ratios above 65%, discovery calls with no budget question, three objections handled by discounting. Write down what you hear verbatim. That list is your scope of work.
By the end of this you have either a specific brief or a realisation that the money belongs somewhere else. Both outcomes are worth two weeks.
What to buy once you know the gap
Assume the diagnosis says skill. Now the shape of the purchase matters.
The pricing bands are worth knowing so you can tell a fair quote from a lazy one. Self-paced courses run roughly $200 to $2,000 per seat. Live workshops run $1,500 to $5,000 per person. Annual team memberships tend to land between $25,000 and $40,000 and start making sense somewhere north of twenty people. General L&D benchmarks put per-employee spend around $1,200 to $1,500 a year, and complex B2B sales teams run $1,500 to $3,000 per seller.
The methodology question matters less than people argue about. MEDDIC suits complex multi-stakeholder deals with procurement. SPIN and Sandler work well on shorter transactional cycles. Challenger fits when you are selling a change of approach rather than a replacement product. Pick the one closest to how your best rep already sells, because you are trying to make the other eight repeatable, not trying to convert the whole team to a philosophy.
Reinforcement is the product, and you build it
This is the part you cannot outsource, and it is where I spend most of my time with clients on this problem.
Step one and two are half a day of configuration in HubSpot and maybe two days in Salesforce. Step three is a calendar decision, not a budget decision. Step four is a dashboard.
We usually build this layer before the training runs, not after, so the vendor's material lands on a system that already expects it. When a client wants it automated, we wire the nudges through n8n so a rep gets a Slack prompt when a deal sits in qualified with empty fields, rather than a manager having to notice. That is the kind of thing our AI and automation work covers, and it costs a fraction of the training spend it protects.
The coaching cadence is the other half of this. Training is an event. Coaching is the weekly system that keeps the event alive, and teams that run real coaching alongside training see materially better attainment than teams that run either alone.
How to know six months later whether it worked
Set the measurement up before the training, because you cannot reconstruct a baseline afterwards. Three things, recorded the week before:
Stage conversion at the specific stage the training targets. Not overall win rate, which moves for a dozen reasons. If you bought discovery training, measure discovery to demo conversion and nothing else.
Rubric scores on ten calls per rep. Score them before, score them at 30 and 90 days. This is the only measure that separates "they learned it" from "they use it", and 90 days is the number that matters because that is where behaviour normally reverts.
Field completion rate on the new CRM fields. Crude, but it correlates with everything else and it takes zero effort to pull.
Then compute the return honestly. If twelve reps cost you $30,000 all in, and discovery to demo conversion moved from 38% to 46% on the same lead volume, work out the extra deals that represents at your average deal size and compare. If the number does not clear the spend by a comfortable margin, say so out loud rather than quietly renewing.
One more thing worth pricing in: turnover. B2B sales attrition runs 25% to 35% a year and replacing a rep costs well over $100,000 by the time you count recruiting, ramp and lost coverage. Teams with real development programs hold onto people longer. That is a legitimate line in the business case even if it never shows up in win rate. It also connects directly to how well your onboarding and ramp works, since the same reinforcement system serves both.
What I would do with $30,000 and twelve reps
Concretely, because founders keep asking me this exact question:
Two weeks and roughly $4,000 of effort on the diagnosis above. Five queries, ten call recordings, a written brief. Around $16,000 on a specialist who addresses the one gap the brief names, using your deals as the material, with live practice and recordings. About $6,000 building the reinforcement layer, fields, stage gates, dashboard, manager rubric. Hold back $4,000 for a follow-up session at 90 days aimed at whatever the rubric scores show did not stick.
That splits roughly half to the training and half to everything that makes the training survive. Almost nobody budgets that way, which is a large part of why the 28% number is what it is.
If the diagnosis instead comes back saying process or targeting, spend the whole thing on the CRM and the playbook, and revisit training in two quarters when you can actually see whether reps are the constraint.
Not sure whether you have a skill gap or a data gap?
Book a free 30-minute audit. We will run the five diagnostic queries against your CRM and tell you honestly which of the four problems you have before you spend anything on training.
Book an audit →FAQ
How much should a B2B sales training program cost per rep?
Budget $1,500 to $3,000 per seller per year for a complex B2B sale, which sits above the general L&D benchmark of roughly $1,200 to $1,500 per employee. Self-paced courses run $200 to $2,000 a seat and live workshops $1,500 to $5,000 a head. Team memberships at $25,000 to $40,000 a year usually beat per-seat pricing once you pass about twenty people. Whatever the total, hold back a third of it for reinforcement rather than spending it all on the event.
Should we build sales training in house or buy it?
Buy the method, build the content. External programs are good at teaching a repeatable qualification or negotiation framework, and they are worth paying for because your team has probably never been taught one properly. They are bad at knowing your product, your buyers and your competitors. The talk tracks, battlecards and objection responses should come from your own closed-won and closed-lost calls, because that material is specific enough to actually change a conversation.
How long before sales training shows up in the numbers?
Rubric scores on recorded calls move within 30 days if the practice was real. Stage conversion at the targeted stage usually moves within one sales cycle plus 30 days, so for a 60-day cycle expect a readable signal at around 90 days. Win rate and revenue lag by a full cycle beyond that. Anyone promising a revenue lift inside a quarter on a three-month sales cycle is selling you arithmetic that does not work.
Does sales training work for a team of five?
A generic program for five people rarely pays back, because the per-seat economics are bad and the diagnosis is usually simpler than a program can address. At that size, spend the money on conversation intelligence, a weekly call review with the founder or head of sales, and one specialist session on the single gap you can name. Formal programs start earning their keep somewhere around ten to twelve reps, when the manager can no longer coach everyone personally every week.
What is the difference between sales training and sales enablement?
Training is a set of events that teach skill and method. Enablement is the ongoing system of content, tooling and process that makes the skill usable in a live deal. Training without enablement decays inside 90 days. Enablement without training just distributes material to people who do not know what to do with it. If you have to sequence them, build the enablement system first, because it is cheaper and it makes the training measurable when it arrives. Our view on the tooling side of enablement covers where a platform helps and where it is premature.
Diagnose first, then spend
The uncomfortable version of this post is that a good chunk of sales training budget gets spent to avoid a harder conversation. Rewriting pipeline stages, admitting the ICP is wrong, or telling a manager their deal reviews are theatre is more awkward than booking a vendor for two days in a hotel room. Training feels like doing something. Sometimes it is.
But go run the five queries. Two weeks, no licence fee, and at the end you either have a sharp brief that makes the training three times more effective, or you save yourself $50,000 and fix the thing that was actually broken.
If you want help running that diagnosis, or building the CRM layer that makes any training stick, that is most of what we do. Have a look at our CRM and RevOps work and our go-to-market work, or get in touch and we will take a look at your pipeline data together.