Gavel Gavel
Experts Guides Skills
Cited from real sources 6 min read Updated August 2026

A framework by Mark Roberge

The Sales Acceleration Formula: Mark Roberge's System for Hiring and Scaling Sales

The Sales Acceleration Formula is Mark Roberge's five-part method for making revenue predictable: hire on the traits your own data correlates with success, train on the buyer journey instead of scripts, coach one skill a month off funnel metrics, keep demand generation matched to sales capacity, and instrument all of it. Roberge built it as HubSpot's fourth employee, running sales from zero to $100M.

What hiring on pedigree gets you

You end up with the bottom 25% performers in your industry.

Every founder asks for ten years selling a product like mine. The formula exists because that is the wrong request.

Mark Roberge 20VC with Harry Stebbings Watch at 05:45

The framework

Sales as an engineering problem

Roberge was an MIT engineer who fell into sales, and he ran HubSpot's revenue org the way he would have run a build. Five parts: hiring, training, management, demand generation, technology. Each has inputs you can measure and outputs you can correlate against. Sales is not a personality contest you win by finding a closer; it is a set of loops you can instrument.

Most founders do the opposite. They hire a sales leader, hand over the function, and inherit whatever comp plan came out of that person's last company. Roberge calls it the biggest pothole in the exit from founder-led selling:

Every time I ask them what are you looking for, they always say the same thing: 10 years of experience selling a product like mine to a customer like mine.
Roberge, on the request every founder makes Watch at 05:20

Pedigree is the one variable a founder can judge without knowing sales, so it absorbs the attention and the actual skills go unassessed. Roberge runs the other way: given a rep who has sold your customers and one who has closed your deal size, take the deal size. Champion mapping, procurement and legal take years. The vocabulary takes weeks. The formula replaces the question: not who has the right resume, but which measurable traits predict success here.

How to apply it

How do you run the Sales Acceleration Formula?

Seven moves. The first three build the hiring loop; the rest turn training and coaching into measurable work.

  1. 1

    Write down five to ten traits you believe predict success here.

    Roberge's list was coachability, curiosity, prior success, intelligence, work ethic. Yours is a hypothesis, not an answer.

  2. 2

    Score every candidate on every trait, one to ten, every interview.

    The scorecard is the data collection. Skip it once and the regression has a hole a year from now.

  3. 3

    Regress the scores against twelve months of performance, then cut what did nothing.

    This is where the surprises live. Aggression, the trait founders reach for first, rarely survives.

  4. 4

    Train on the buyer journey, not on the pitch.

    A rep who understands how buyers describe the problem adapts. A script stalls the moment a call goes sideways.

  5. 5

    Gate the first customer conversation behind an exam.

    Certify comprehension instead of counting training days. Time served proves nothing about holding a discovery call.

  6. 6

    Coach one skill a month, and pick it from funnel data.

    Find where the funnel breaks, calls to demos to close, then work that one skill for thirty days. A list of fifteen produces nothing.

  7. 7

    Tie lead volume to headcount with a written SLA both ways.

    Marketing commits to volume and quality, sales to follow-up speed. Without it both teams spend a quarter pointing at each other.

One rule sits under all seven: promote on metrics, never on tenure. A career path built on time served drains the meaning out of everything you measured.

The hiring half

The Sales Hiring Formula

Coachability keeps surviving the regression, and it is the trait founders weight least. Roberge tests it inside the interview rather than asking about it. He puts the candidate on a first call with a lead who downloaded an ebook yesterday, watches for open-ended questions and quantified need rather than a five-minute commercial, then stops and asks them to grade themselves.

So I have you self assess, and if you're like, that? I was awesome, I have no feedback for myself, then that's a red flag.
Roberge, on testing coachability in the interview Watch at 10:51

He gives one piece of praise and one correction, runs the role play again, and scores the delta rather than the second performance. The same test survives the hire: when a manager says a rep is failing, Roberge asks what they are coaching and how. Lift that holds is a slow ramp. Lift for a week then slide back is the cut.

Boundary conditions

When does the Sales Acceleration Formula fail?

Works best when

  • You have product-market fit and a repeatable ICP
  • You have hired enough reps for the correlations to mean anything
  • At least 80% of new customers look healthy inside 90 days

Fails when

  • You are pre-fit, where Roberge runs no commission plan at all
  • You have four reps, so any correlation is anecdote in a lab coat
  • You chase enterprise logos before the market pulls you upmarket

The pre-fit failure matters most. Before product-market fit Roberge pays flat salary plus equity and no commission: a rep sweating a mortgage payment through the third pivot pushes you toward revenue you should refuse.

The second precondition is a defined ICP. A funnel aimed at the wrong accounts only fails faster. Roberge sorts accounts green, yellow and red, and treats red as a fence:

Then there's reds where we have strong conviction we will not sell. In fact, you cannot sell them.
Roberge, on the green, yellow, red ICP split Watch at 38:51

The third is a health signal you can pay against, and the fix is the piece most teams never copy:

I'm going to give you half of the commission when the customer pays, and half the commission when they hit their leading indicator of retention, which usually happens in the first 30 days.
Roberge, on splitting commission against retention Watch at 16:51

Operators diverge on sequence. Jason Lemkin reads churn backwards, as evidence you sold the wrong ICP months ago. Roberge fences the ICP up front so the rep cannot make that sale.

The sources

Where Roberge discusses this

Roberge is now a partner at Stage 2 Capital. The clearest recent walkthrough is his 20VC appearance with Harry Stebbings, taking hiring, comp and ICP in the order a founder hits them.

05:20
Timestamp

Mark Roberge · 20VC, E1176

Recruiting and Assessing Sales Talent

"10 years of experience selling a product like mine to a customer like mine."
10:51
Timestamp

Mark Roberge · 20VC, E1176

Role-Play and Assessing Coachability

"I have no feedback for myself, then that's a red flag."
16:51
Timestamp

Mark Roberge · 20VC, E1176

Aligning Incentives with Customer Success

"half the commission when they hit their leading indicator of retention."
38:51
Timestamp

Mark Roberge · 20VC, E1176

Refining ICP with the Green-Yellow-Red Framework

"we have strong conviction we will not sell."

Useful? Send it to whoever is about to make their first sales hire.

Want the full playbook?

Get 94 sales & revenue frameworks.

14 frameworks 27 rules 50 heuristics & principles 2 operators

From Mark Roberge and Jeb Blount. Drop one .md into Claude, Cursor, or ChatGPT. Your AI cites practitioners, not guesses.

See the pack

Instant .md download · One-time purchase · No subscription

Related frameworks

Sales

Four Go-to-Market Playbooks

Geoffrey Moore

One playbook per market phase. The tell for which one you are in is budget, not revenue.

Sales

Hook-Meat-CTA Ad Framework

Alex Hormozi

Failed ads die at the hook. Name the pain in three words, deliver the meat, then make one clear ask.

Sales

Cold Email Method

Sam Parr

Pick 100 specific names, 30 minutes a day for a month, follow up 10 times each. Expect 10 percent to reply.

Negotiation

Tactical Empathy

Chris Voss

Demonstrating you understand their position, not agreeing with it. Push until they say that's right, never you're right.