A CEO I worked with last November had the whole next year mapped out in a board deck. Twelve million in new ARR. Eight account executives. He divided one by the other, added a bit on top because two more hires were coming in Q2, and told me the quota was set. One point six million per rep. Done in about ninety seconds.
I asked him one question. How many of your eight reps hit their number last year?
Two.
So the plan for next year was to take a target that six out of eight people missed, raise it, and hand it to a team that included three people who had not started yet. He was not being reckless. He was doing what almost every founder does, because dividing the target by headcount feels like math. It looks like a model. It is really just arithmetic wearing a model's clothes, and it is the single most common reason B2B revenue plans fall apart by May.
I have been building and fixing quota models for B2B teams for over ten years. The good ones are built from the bottom up, from what your team can actually produce, and then reconciled against the number the board wants. The bad ones start at the board number and work backwards until the spreadsheet balances. Same output format. Completely different outcome.
The attainment numbers should scare you into changing the method
Start with what is actually happening across B2B sales teams, because the gap between plan and reality is wider than most founders realise.
RepVue tracks quota attainment across hundreds of software companies and tens of thousands of quota-carrying reps. Their Q2 2025 read put average attainment at roughly 43 percent across about 246 cloud and software companies. Salesforce's 2025 State of Sales reported that only around 28 percent of reps hit their annual number. Pre-2022, the same benchmarks sat in the 60 to 65 percent range.
Average quota attainment across roughly 246 software companies in RepVue's Q2 2025 data. If your capacity model assumes 100 percent, you have already built a plan that misses by more than half.
Read that as a planning input, not as a complaint about rep quality. If the median rep in your market lands somewhere near half of quota, then a model that multiplies headcount by full quota is not optimistic. It is wrong by construction. You will hire against revenue that never arrives, commit to a burn rate you cannot cover, and spend Q3 explaining to your board why the pipeline looks fine but the bookings do not.
The healthy target most operators aim for is 55 to 65 percent of reps at or above 100 percent, with the median rep landing near 90 percent and the top fifth pulling 130 percent or more. That distribution tells you the quota is set right. Everyone missing means the quota is fiction. Everyone clearing it easily means you left money on the table and overpaid on commission.
Why divide-by-headcount produces a number nobody believes
The top-down method has four failure points, and they compound.
It counts bodies instead of selling capacity. A rep who starts in March is not a rep for the year. They are maybe half a rep, depending on your ramp curve. Counting them as one is the fastest way to inflate a plan.
It assumes full attainment. Nobody says this out loud, but dividing 12 million by 8 reps assumes each of them delivers 1.5 million. Your actual blended attainment is probably 60 to 70 percent on a good team.
It ignores territory quality. Two reps carrying the same number where one has 40 named accounts in a mature segment and the other has 300 unqualified logos in a new vertical are not carrying the same quota. They are carrying the same digit.
It has no relationship to what a rep earns. Quota and on-target earnings have to sit in a sane ratio or the plan stops being a plan and becomes a base salary with theatre attached.
The second column takes about a day to build. I have never seen a team regret that day.
Build the capacity number first
Here is the model I use. It is deliberately simple, because a quota model that only one person understands is a quota model that quietly rots.
Count ramped rep equivalents, not headcount
A rep is not productive on day one. Ramp time runs about 3 to 4 months for SMB motions, 4 to 6 for mid-market, and 6 to 9 for enterprise. The way to handle this is a ramped quota schedule: a common six-month curve runs 0 percent in month one, then 25, 50, 75, 90, and full quota in month six.
Now convert that into capacity. A rep starting in January on a six-month ramp contributes roughly 0.85 of a full year. One starting in July contributes about 0.4. Add those fractions across your whole roster and you get ramped rep equivalents, which is the only headcount number worth putting in a model.
For the CEO with eight reps and two hires coming in Q2, the honest count was not ten. It was about 8.9.
Apply real attainment, not aspirational attainment
Take your own historical blended attainment. Not the average of the people who hit quota, the average across everyone who carried one, including the person who left in August. If you do not have history, plan at 65 to 70 percent for an established team and 55 to 60 percent for a team in its first year of a new motion.
The formula is nothing fancy:
Capacity = ramped rep equivalents x quota per rep x expected blended attainment
If that produces a number below your target, you have found the actual conversation. You need more reps, a higher quota per rep, better conversion, or a smaller target. Those are the only four levers. Pretending the gap does not exist is not a fifth one.
Sanity check quota against on-target earnings
The ratio between quota and OTE is the fastest way to tell whether a number is set at a level humans can reach. The benchmark most SaaS teams work to is 4x to 6x, with 5x treated as steady state. SMB motions with smaller deals and higher volume sit nearer 4x. Enterprise motions with longer cycles sit at 5x to 6x. High-velocity, low-ASP models can run higher.
An AE on 200,000 OTE at a 5x ratio carries a 1 million quota. If you want that same rep on 1.6 million, you are running an 8x ratio, and unless your deal sizes and win rates support it, you have just designed a plan where the variable half of their pay is decorative. I wrote about the rest of that math in the guide on sales compensation plan design, because quota and comp are one system, not two.
Check that each territory can hold the number
This is the step teams skip and then wonder why two reps carry the whole team every quarter. A quota is only real if the accounts assigned to that rep contain enough addressable revenue to produce it at your normal win rate.
Run the check per rep. Take the number of accounts in their patch, multiply by the share you expect to open in a year, then by your win rate, then by average deal size. If a rep has 120 accounts, opens 20 percent, wins 22 percent, at a 45,000 average, that patch produces about 238,000 of realistic bookings. Handing that rep a 1 million quota is not ambition. It is a resignation letter with a twelve-month delay.
When patches come out uneven, fix the patch, not the quota. That is what territory planning is for, and it is far easier to rebalance accounts in January than to explain in October why one rep got a winnable book and another got a desert.
Then reconcile with the board number, out loud
Almost every time, the capacity model produces a smaller number than the target. That is not a failure of the model. That is the model doing its job.
The move is to put the gap on one slide and name the levers that close it. More ramped capacity means hiring earlier, not more, because a hire in month nine barely moves the year. Higher conversion means fixing a specific stage, and you should be able to point at which one. Bigger deals mean pricing or segment changes. A longer runway on the target means a phased plan.
What you must not do is close the gap by raising per-rep quota until the spreadsheet balances. Raising quotas without raising coverage, conversion, or account potential does exactly one thing: it lowers the percentage of reps who hit plan. And when attainment collapses, your best people leave first, because they are the ones with options. The reps who stay are the ones who cannot move.
A quota is a forecast of what a rep can produce, not a statement of what you need them to produce.
The moment those two things diverge and you publish the second one anyway, the number stops being a target and becomes a message. Reps read it as: nobody here did the math.
The mechanics that decide whether reps trust it
Getting the number right is half the job. The rules around it decide whether people plan their year against it or quietly write it off.
Ramped quota for new hires needs to be written down before they sign, with the schedule and any draw spelled out. Reps compare notes. If one person got a three-month ramp and another got six for the same role, you will hear about it.
Quota relief needs a policy, not a case-by-case judgment call. When a rep gives up accounts mid-year in a rebalance, when a patch gets split for a new hire, when someone takes six weeks of parental leave, the adjustment rule should already exist. Deciding it in the moment means every decision looks political.
Annual quota with quarterly milestones beats pure quarterly quota for most B2B teams with cycles over 60 days. Pure quarterly targets in a 90-day sales cycle produce end-of-quarter discounting and deals pulled forward at the wrong price. If your cycle is short and transactional, quarterly is fine.
Attainment should be visible weekly, and it should come from the CRM, not from a spreadsheet a RevOps analyst rebuilds by hand. The moment reps maintain private trackers because they do not trust the dashboard, you have lost the thread.
Where the model meets the CRM, and usually breaks
Everything above assumes your historical numbers are real. In most of the audits I run, they are not, and that is the uncomfortable part.
Win rate is wrong because dead deals sit open in the pipeline for months, which inflates the denominator or hides it entirely. Average deal size is wrong because multi-year contracts get booked as total contract value in one record and annual value in the next. Attainment history is wrong because quota lives in a spreadsheet and was edited three times mid-year without a version. Territory capacity is unknowable because half your accounts have no segment or employee count populated.
You cannot build a capacity model on that. So the first job is usually not the quota, it is the CRM data quality underneath it. Practically that means quota stored as a property on the owner record with an effective date, deal amounts normalised to one definition, a stage exit rule that forces stale deals out, and firmographics filled in on every account so patches can be compared.
We do this work as part of CRM and RevOps builds, and a fair amount of it gets automated with n8n jobs that flag deals sitting past their stage limit or accounts missing the fields your model depends on. It is unglamorous. It is also the difference between a quota model and a guess with decimal places.
Once the data holds, the same inputs feed your forecasting model and your pipeline coverage targets, which is the point. Quota, coverage, and forecast should all be reading from one set of numbers. When they disagree, someone is running a private spreadsheet.
Not sure your quota model holds up?
We rebuild capacity models on real CRM data and show you the gap between the plan and what your team can actually produce. Free 30-minute audit, no deck required.
Book an audit →What I told that CEO
We rebuilt his model in a day. Ramped rep equivalents came to 8.9 rather than 10. Blended attainment over the prior two years was 61 percent. His median deal was 52,000 and his win rate from qualified opportunity was 24 percent.
At a 5x quota to OTE ratio, his reps should have been carrying about 1.05 million, not 1.6 million. Capacity at 61 percent attainment came to roughly 5.7 million. The board wanted 12.
That is a brutal slide to build. It was also the most useful conversation he had all quarter, because it turned an argument about rep effort into a specific list: hire the two AEs in January instead of April, fix the stage two to stage three conversion rate that was sitting at 31 percent, and move upmarket on ASP or accept a phased target.
He did not hit 12. He hit 8.4, which was 47 percent growth, and six of his ten reps cleared quota. Nobody quit over the number. That is what a working quota model buys you: not a bigger result on paper, a real one you can hire and spend against.
FAQ
What is a sales quota and how is it different from a target?
A quota is the revenue an individual rep is expected to produce in a period, and it carries pay consequences through commission. A target is the company number. They are related but not the same, and the sum of individual quotas is usually set 10 to 20 percent above the company target to absorb attrition and underperformance. That buffer is called overassignment, and it should be deliberate rather than accidental.
How do you set quota for a brand new sales team with no history?
Use the quota to OTE ratio as your anchor, since you have no attainment history to work from. Start at 4x to 5x OTE for a first team, plan attainment at 55 to 60 percent, and use a longer ramp than you think you need. Then review after two quarters with real win rate and cycle data, and adjust. Setting a first-year quota and refusing to revisit it is worse than setting it slightly wrong.
Should quota be annual or quarterly?
Annual quota with quarterly milestones works better for most B2B teams with sales cycles over 60 days, because it stops reps from discounting to force deals into an arbitrary quarter end. Quarterly quota suits high-velocity, transactional motions. The deciding factor is your cycle length, not your reporting cadence.
What is a healthy quota attainment rate?
Aim for 55 to 65 percent of reps at or above 100 percent, with the team median near 90 percent. If more than 80 percent of your reps clear quota, the number is too low and you are overpaying commission on revenue you would have booked anyway. If fewer than 40 percent clear it, the number is unreachable and you will lose your best people first.
How often should quotas be adjusted mid-year?
Only for structural changes: a territory rebalance, an extended leave, a segment reassignment, or a product change that removes something the rep was selling. Adjusting quota because the market got harder teaches reps that the number is negotiable, and next year they will negotiate. Write the relief policy before the year starts and apply it the same way every time.
If your quota model currently starts with a board number and a division sign, that is the thing to fix before the next planning cycle. We help B2B teams rebuild it from capacity, clean the CRM data it depends on, and wire the reporting so attainment is visible without a spreadsheet. Have a look at how we approach go-to-market and AI automation work, or get in touch and we will pressure test your current model.