How Do I Get My Reps to Follow the Sales Process?

Adherence scoring versus the common alternatives

Most teams reach for one of four fixes when reps stop running the play, and each fails in a predictable, diagnosable way. Understanding why each one breaks is what tells you where the weighted scorecard actually belongs — it is not a replacement for all of them, it is the layer that makes the others stick.

Mandatory CRM fields and validation rules. This is the most common first move because it is the cheapest: make the stage-exit field required, block the opportunity from advancing without a next step, force a close date and a competitor selection. It genuinely works for data hygiene, and it produces the clean inputs a scorecard later needs. What it does not do is produce behavior. A required "discovery completed" checkbox creates a rep who checks a box, not a rep who runs discovery. Within about two weeks of any new required field going live you will see the same pattern: the field is 98% populated and the underlying behavior is unchanged, because the enforcement point sits at the moment of data entry rather than at the moment of selling. Validation rules police keystrokes. They cannot police whether the economic buyer was actually in the room.

Methodology training. MEDDICC, MEDDPICC, Challenger, SPIN, Sandler, Command of the Message — pick one, run a two-day workshop, hand out the cards. Training changes vocabulary reliably and behavior temporarily. The well-documented pattern is a strong spike in usage for roughly two to four weeks, then a slow decay back toward baseline over the following quarter as deals get urgent and reps fall back on instinct. The reason is structural, not motivational: training installs a definition of good but installs no ongoing measurement of it, so the only feedback the rep gets after the workshop ends is whether the deal closed. Closed deals are a lagging, noisy, low-frequency signal. You cannot coach on a signal that fires once a month and is confounded by territory, timing, and luck.

Manager inspection and deal reviews. A disciplined manager running weekly pipeline reviews is genuinely the highest-leverage version of the non-scorecard approaches, and on a small team it can be enough. Its limits are arithmetic. A frontline manager with eight reps carrying 25 open opportunities each is looking at 200 deals; a real inspection of one opportunity — reading the notes, checking the mutual action plan, confirming the buying committee — takes eight to twelve minutes. Full inspection is a 30-hour week that manager does not have. So inspection collapses onto the biggest deals in the current quarter, which are exactly the deals where reps are already trying hardest. The process failures live in the early-stage, mid-size deals nobody looks at. Inspection also varies wildly by manager: two managers with the same rubric will disagree on the same deal roughly a third of the time unless you have written level definitions, which is precisely what the matrix forces you to write.

Leaderboards and gamification. Ranking reps by bookings, meetings set, or dials is energizing and cheap, and it is the single most common reason a broken process survives. A leaderboard measures one number, so it rewards whoever optimizes that number by any route available. The rep who is a level 5 on charisma and a level 1 on qualification and mutual action plans sits at the top of the board every month and gets held up in the all-hands as the model — while producing the highest-churn, lowest-forecast-accuracy book on the team. You do not find that out until renewals come due or until they leave and nobody can pick up their accounts, because none of the process artifacts exist.

The weighted matrix approach differs on one axis: it measures the *process itself* as a first-class object, at the same frequency as coaching, with weights the business controls. Each behavior a disciplined rep is supposed to run — qualification framework completed, discovery before demo, mutual action plan built, economic buyer engaged, stage-exit criteria met, required fields filled, agreed methodology followed — carries its own weight and gets a 1-to-5 level every cycle. The composite is the sum of (weight × level) across all lines. That single number is what makes the split between "closed something" and "ran the play" visible, weekly, in a form nobody can argue with.

The honest trade-off: the matrix requires manager judgment every week, and judgment costs time and can be inconsistent. Budget 10-15 minutes per rep per cycle for scoring, plus a one-time investment of four to six hours writing level definitions. If you will not spend that, the matrix decays into a scoring ritual nobody trusts — which is worse than no scorecard, because it burns the credibility you need for the next attempt. Validation rules and training cost less and demand less discipline. They also do not work on their own. The realistic architecture is all of them stacked: rules produce clean inputs, training defines good, the matrix measures it weekly, and comp or visibility supplies the teeth.

How to choose between them

The choice is not really between tools. It is between three questions asked in order: does a written definition of the process exist, where should the enforcement pressure sit, and can you afford the manager time the chosen approach demands.

Start with whether the play is written down. If you cannot produce a one-page document listing the behaviors that define a complete deal, no tool helps and no training sticks — you are trying to enforce something nobody has defined. Writing it is the highest-value four hours in this entire project because it forces leadership into a room to argue out what "good" means before any rep is graded. That argument is far easier in the abstract than during someone's review. Expect genuine disagreement: the VP will say discovery matters most, the enterprise AE lead will say economic-buyer access does, and finance will say forecast-field hygiene does. Resolve it with weights rather than by picking a winner.

Then decide where the teeth live. There are exactly three enforcement points and they have very different costs. *Visibility* — the score is published, everyone sees their levels and the distance to the next one — is the cheapest to install and works well on teams with strong peer culture and low tenure variance. *Pay* — the composite feeds a component of the commission plan — is the strongest lever and the slowest to install, because comp changes touch finance, legal, and sometimes employment agreements; assume a full quarter of lead time and a plan-year boundary. *Evidence* — conversation intelligence scoring what actually happened on the call rather than what got logged — is the most expensive and answers a question no CRM field can: did the step happen, or did the rep check the box. Most functional teams end up with visibility immediately, pay at the next plan year, and evidence only if the budget already exists for it.

Then check the manager-time math honestly. Scoring eight reps on nine lines at 12 minutes each is about 90 minutes a week per manager. If your frontline managers are also carrying a quota, that time does not exist and you should cut the matrix to four or five lines rather than pretend. A five-line matrix scored religiously beats a twelve-line matrix scored twice and abandoned. Sizing rule of thumb: one line per behavior you are genuinely willing to coach, never more than nine, and if two lines always move together, collapse them.

A final selection note that saves real money: prove the method free before buying anything. Build the matrix in a spreadsheet or a free scorecard tool, run one pod of three to five reps for two or three cycles, and see the adherence-versus-results split with your own team's names on it. Almost every team that does this discovers at least one rep the leaderboard was hiding and at least two matrix lines that turned out not to predict anything. Both discoveries are cheaper to make before you have signed an annual contract or rewritten a comp plan around lines that do not matter.

Costs, timelines, and expected impact

Direct tool cost. The honest range is zero to enterprise. A spreadsheet with a SUMPRODUCT rolling the composite costs nothing and is a completely legitimate way to run this for a team under about ten reps — the price is maintenance, and the failure mode is specific: someone stops updating it around week five, a stray inserted row silently breaks a formula, and by week eight the file is a historical artifact nobody opens. Free browser-based scorecard tools remove the formula fragility. Sales gamification and scorecard platforms generally sit in the low-to-mid tens of dollars per user per month; several publish pricing, several quote only. Your CRM likely already holds every raw input the composite needs, and building the scorecard from custom dashboards, report types, and formula fields is a real option that costs admin hours rather than license fees — the advantage being that the score lives inches from the pipeline it is grading, so nobody reconciles two sources of truth on Friday. Incentive-compensation platforms — the ones that model and pay multi-component plans with an audit trail — are the expensive tier and are almost always custom-quoted. Conversation intelligence is likewise a per-seat enterprise purchase. Check current published pricing directly with each vendor; it moves.

The cost that actually matters is time, and it is front-loaded. Budget roughly four to six hours to write the play and the level definitions, with the right people in the room. Budget another two to four hours to build the matrix and validate the weights against a handful of historical deals — score five closed-won and five closed-lost opportunities retroactively and see whether the composite separates them. If it does not, your weights are wrong, and you have learned that for the cost of an afternoon rather than a quarter. Then budget the standing cost: 10-15 minutes per rep per cycle for the manager, plus about 15 minutes of the rep's own one-on-one spent on the score rather than on deals.

A realistic rollout timeline. Week one: write the play, set weights, define what a 1 and a 5 look like on every line. Week two: score a single pod of three to five reps and do not show anyone the numbers yet — this is a calibration run to find where two managers disagree. Weeks three and four: score the pod again, share the scores with those reps, and run the coaching conversation off the lowest weighted line. Weeks five through eight: expand to the full team, publish the matrix so nobody is scored in the dark, and hold the weights steady so people can learn the game. Somewhere in the second quarter, if the composite is separating good process from lucky results, wire it into comp at the next plan boundary. Anyone promising floor-wide behavior change inside three weeks is selling something.

What impact to actually expect. Be skeptical of precise numbers here — public benchmarks on process adherence are thin and heavily confounded, and anyone quoting you a clean percentage lift is extrapolating from a case study. What you can expect with reasonable confidence, based on the mechanism rather than a study: the first visible change is in *dispersion*, not in average performance. Scores cluster tightly in cycle one because managers are conservative, then spread out over three or four cycles as level definitions sharpen. The second visible change is in forecast quality — deals that scored high on mutual action plan and economic-buyer access will close closer to their predicted date than deals that did not, and this becomes measurable within about two quarters. The third is in ramp time for new hires, because the matrix gives a new rep an explicit, ordered list of what to get good at instead of "shadow the top rep."

What will not change and you should say so upfront. The matrix does not make a bad territory good, it does not fix a broken ICP, and it does not create pipeline. If your problem is that reps are not following the process *because the process does not work*, scoring adherence harder just makes people comply with something ineffective faster. Test this before you build: ask three of your best reps which steps they skip and why. If they all skip the same step and all give the same reason, that step is the problem, not the reps.

A specific cost to plan for that teams miss: manager calibration. Two managers scoring the same rep will diverge until you make them score the same three deals independently and reconcile out loud. Do that once at launch and once a quarter, 60-90 minutes each time. Skip it and within two quarters you will have per-manager grading curves, reps will notice, and the composite loses the one thing that makes it work — the belief that it is fair.

Implementation and handoff details

Build the matrix in a fixed order; skipping steps is where these projects die.

Inventory the whole play, not the parts that are easy to count. Name the eight or nine behaviors that separate a complete rep from a lucky one: qualification framework completed, discovery run before the demo, mutual action plan built, economic buyer engaged, stage-exit criteria met, required fields filled, agreed methodology followed. Anything left off the matrix is functionally optional — it will not be coached and it will not get done. Be ruthless here: if you cannot describe how you would verify a behavior happened, either find a verification method or drop the line.

Weight the few that matter and level the rest. Hang a weight on each line with leadership present, then define levels concretely. A level 1 means the behavior is absent. A level 5 is the bar you want the whole floor clearing. The critical work is writing what a 3 looks like on a live call, in a sentence a manager can read aloud. Vague level definitions are the single largest source of grading drift. Instead of "tighten up discovery," the manager gets to say "move this line from a 2 to a 3, and a 3 means you named the current process, the cost of that process, and who owns fixing it before you demoed."

Point the money and the one-on-one at the composite. When both the commission and the weekly coaching conversation follow the composite rather than one flashy line, reps round themselves out without being nagged. It stops being an annual event and becomes a standing signal: everyone sees their own levels and the exact distance to the next tier, and the only route up is doing more of what the business needs.

Score the lowest weighted line first in every coaching conversation. Not the lowest raw level — the line where weight × gap is largest, because that is where a one-level improvement moves the composite most. This is a small mechanical rule that keeps coaching from wandering.

Keep the weights re-weightable. Adopt a new methodology, redraw stage definitions, react to a market shift — change the weights, and the floor re-aims by the next standup with no comp surgery and no confusion about what changed. Announce every re-weight explicitly and never re-weight retroactively; scores already given stay as given.

Handoff and ownership. RevOps owns the matrix definition, the weights, and the scoring cadence; the frontline manager owns the scores themselves; sales leadership owns the weights jointly with RevOps and signs off on every re-weight. Write down who does what before launch, because the most common organizational failure is RevOps building a beautiful scorecard that no manager feels responsible for filling in. Put the scoring on the manager's calendar as a recurring block, and put a scoring-completion check in whatever weekly ops review already exists — an unscored week should be as visible as an unforecasted deal.

Publish the matrix. Nobody gets scored in the dark. Reps should be able to read the weights, the level definitions, and their own history without asking. Publishing does two things at once: it removes the "this is arbitrary" objection, and it lets a rep self-coach between one-on-ones, which is where most of the actual improvement happens.

Watch for the three predictable failure modes. First, the matrix goes stale because nobody re-weights it after a strategy change — put a quarterly re-weight review on the calendar even if the answer is "no change." Second, managers drift apart — recalibrate quarterly. Third, someone tries to use the composite as a termination instrument in month two, everyone realizes it is a performance-management tool rather than a coaching tool, and honest scoring stops immediately. Decide upfront and say out loud how the composite relates to formal performance management; the safest answer for the first two quarters is that it does not, and that it feeds coaching and comp only.

What to do about the top rep who skips the process. They will post high on results and low on adherence, and that split is the feature. Do not lead with the dashboard. Lead with their interest: the process artifacts protect their renewals, their handoffs when they go on PTO, and their own forecast credibility when they need a discount approved. A process only one gifted rep can bypass does not survive their next vacation, let alone their resignation.

Related questions

How long before reps actually change behavior?

Expect vocabulary change in week one, visible score dispersion by cycle three or four, and durable behavior change over one to two quarters — and only if scoring stays weekly and never lapses. A matrix scored monthly behaves like a performance review, which is the slow ritual it is meant to replace.

Should the composite be tied to compensation immediately?

No. Run it for at least a quarter as a coaching-only signal so managers calibrate and reps trust the scoring. Wire it to comp at the next plan-year boundary, as one weighted component rather than a gate, and give finance a full quarter of lead time.

How many lines should the matrix have?

Four to nine. Fewer than four and it collapses into a leaderboard with extra steps; more than nine and managers stop scoring it honestly. One line per behavior you are genuinely willing to coach every week, and collapse any two lines that always move together.

What if reps say the process itself is broken?

Test it before dismissing it. Ask three strong reps which step they skip and why — if they name the same step for the same reason, fix the step. Scoring adherence to an ineffective process just produces faster compliance with something that does not work.

Can this work without a CRM buildout?

Yes. The matrix is a scoring method, not a system; a spreadsheet or a free scorecard tool runs it fine for a team under ten reps. CRM automation removes hand-keying and makes inputs auditable, but it is an efficiency upgrade, not a prerequisite.

FAQ

How do I score process adherence without micromanaging?

Grade the outcomes of the steps, not the keystrokes — did a mutual action plan actually exist, was discovery run before the demo, was the economic buyer in the room. Weight the two or three steps that genuinely predict a won deal heavily and let the trivia stay light, so the matrix reinforces the spine of the process instead of policing every field. Micromanagement is scoring effort; a good matrix scores results-of-behavior, which is the opposite of hovering.

What if a top rep closes deals while skipping the process?

They post high on results and low on adherence, and that split is exactly the conversation worth having out loud. A process only one gifted rep can bypass does not survive their next PTO week, let alone their resignation. Reframe it around their interest: following the play protects their renewals, their handoffs, and their pipeline predictability — not just the manager's dashboard.

Will this hurt my fastest rep?

No, it redirects them. A rep who is elite on one line and thin everywhere else scores high on a single KPI and mediocre overall, which is precisely the signal and the income opportunity to round out. In practice the strongest performers chase the composite hardest once both the paycheck and the praise follow it, because competitive people optimize for whatever is actually measured.

How does the matrix keep sales, RevOps, and the frontline manager aligned?

Everyone grades the same weighted KPIs, so "a good week" means one identical thing across all three roles and handoff meetings stop devolving into arguments about what counts. When you re-weight, all three re-aim on the same morning instead of drifting into three private definitions of quality that only collide at quarter-end.

How often should reps be re-scored?

On the same rhythm as coaching one-on-ones — weekly or biweekly — so the number stays a live signal rather than a dusty artifact. Frequent, low-drama scoring turns each conversation into "nudge this one line up" instead of "defend your whole year," which is what makes people willing to be scored at all.

What is the fastest way to prove this works before rolling it out to everyone?

Pick one pod of three to five reps, build the matrix free in a spreadsheet or a browser tool, and run it for two or three cycles alongside whatever you already track. You will see the adherence-versus-results split immediately, surface the reps the leaderboard was hiding, and have concrete before-and-after evidence for leadership before you touch the comp plan.

Sources

• Salesforce — sales process, stages, and reporting documentation: https://help.salesforce.com/
• HubSpot — sales process and pipeline management resources: https://knowledge.hubspot.com/
• Harvard Business Review — sales management and performance research: https://hbr.org/topic/sales
• MIT Sloan Management Review — sales force management: https://sloanreview.mit.edu/
• Gartner — sales practice research and terminology: https://www.gartner.com/en/sales
• WorldatWork — sales compensation design principles: https://worldatwork.org/
• SHRM — performance management and incentive practices: https://www.shrm.org/
• McKinsey & Company — growth and sales operations insights: https://www.mckinsey.com/capabilities/growth-marketing-and-sales

Related on PULSE

• [How Do I Get My Support Reps to Convert Chats to Sales?](/knowledge/tl0517)
• [How Do I Get My Distribution Sales Reps to Sell the Full Catalog?](/knowledge/tl0197)
• [How Do I Get My Sales Reps to Sell the Full Product Line Instead of Just One or Two Products?](/knowledge/tl0151)
• [How do I get my sales reps to sell the full product line without cannibalizing our flagship product in 2027?](/knowledge/tl21785)
• [How Many Sales Reps Do I Need to Hire for My Background Screening Company?](/knowledge/tl0248)