The Sales Organization

What has to be in place for a sales team to keep selling on a bad Tuesday, when the head of sales is tired and nobody has given a speech?

I've had to answer that at every company where I've run a team. At Eatigo I built the acquisition team that signed 600 restaurants in Jakarta in under six months, door to door, the painful kind of selling. At Google I was an account manager on a $6M-a-quarter book, the other kind. Across ZALORA, Eatigo and JD.com I have run sales teams from 3 to 80 reps.

The framework below has three layers. A team needs all three to win, and can survive on two for longer than most people expect.

Operations and structure: CRM, rules, training, knowledge, all of it simple enough that the laziest rep can execute. Performance culture: standups, a weekly meeting, visible results, praise, promotion, and the exit of persistent low performers. Sales incentives: for outbound-oriented teams, a commission at 80-100% on top of base, attainable about seven times out of ten, paid on a short cycle, uncapped, individual. Then the triangulation. Salary is set by the market, the 80-100% by human psychology, payback by market fit. The company has to find the point where the three close.

The percentages are one operator's sample and I label them as such throughout. The piece takes the layers in order (02 to 04), then the arithmetic that decides whether the incentive layer is affordable (05), then the second triangle of process, culture and money (06). Counterevidence in 07, takeaways in 08.

The test is that the laziest, least able rep can win. It's less cynical than it sounds. A process built for the best rep on a good day is abandoned on a bad one. A process built for a tired rep on a Tuesday afternoon gets used. The table below is what that means for the CRM, qualification, rules, training and knowledge. Each element has one job: remove a decision the rep would otherwise make badly at four in the afternoon.

The cost of getting this wrong is in the market data. Salesforce's State of Sales has reps spending 28% of their week selling; the rest goes to deal management, data entry and finding things.[1] The Bridge Group's 2026 survey has ramp at 6.2 months, the longest it has recorded, and the experience required at hire up from 2.7 to 3.7 years.[2]

Companies that can't train buy the training from someone else's payroll. A rep who ramps in six months and leaves after two years has spent a quarter of her tenure not selling. The operations layer is where that quarter is won or lost.

Most teams overdo it. Every added field, approval and dashboard is paid for in selling time, out of the 28% of the week that reps already spend on it. My rule is the smallest process that reliably prevents the mistake. With AI, the whole knowledge base sits behind natural language: a rep asks how a discount works or what happens after a trial and gets the current rule with its source.

Complexity has a cost, and most people refuse to accept it. Bloated operations are paid for in speed, because every step is a queue; in time, because every field has to be filled and every approval chased; and in morale, because a rep who spends the afternoon feeding a system she doesn't believe in stops believing in the people who built it.

Communication is expensive too. Each new rule has to be explained, remembered and enforced, and each exception argued. The people who add process rarely carry that bill, so they treat it as free. It isn't, and the reps pay it first.

Limits: this is the framework I run, not a measured standard; the details change with the job and I have not used every element at every company.

Humans compare, and they are social. The whole layer runs on that. People notice who gets praised, who gets promoted and what happens to someone who keeps missing the standard. They adjust to what they see, not to the values slide. The head of sales shapes that environment with ordinary decisions, especially the ones made when a good seller behaves badly or a popular colleague keeps failing.

My routine is a short daily standup and a weekly sales meeting. Performance is transparent, with territory, tenure and ramp visible, so a new rep and a ramped rep aren't graded on the same test.

I praise the best work publicly, promote people who can carry more, and fire persistent low performers once training, expectations and a fair chance have been given. I don't automatically fire whoever ranks last. Even an excellent team has a last place, and an empty territory isn't a character defect.

Steenburgh and Ahearne found that social pressure is what moves laggards, and that reps with a bench behind them perform better than reps without one.[3] Lazear's Safelite result has a culture finding inside it: half of the 44% gain came from better workers applying once pay followed performance.[4]

Culture can substitute for money, at a price. The extreme case is multi-level marketing. The FTC's 2024 review of 70 income disclosure statements found most participants made $1,000 or less a year before expenses, and in at least 17 of the MLMs most made nothing.[5]

The commission is good on paper and for over-achievers with a network. For the average new joiner the math doesn't work, and the events, the rallies and the language supply the motivation the money can't. It's a performance culture doing the whole job. Someone with enough charisma to run one may prefer the sports cars.

I'd rather have a team that can calculate its commission, because a head of sales who has to manufacture belief every morning is a single point of failure. The reverse also holds: if the head of sales leaves or loses energy, operations and incentives keep the team selling for a long time without the culture layer.

The routine is a useful theatre. Some rep will say, more or less politely, "fuck off, read the CRM instead of meeting me every day". We should not tell the reps the routine is a theatre, but we should know it is. A theatre, in my sense, is a meeting that conveys energy and uses human psychology to the organization's advantage.

As information transfer, a sales meeting is hilarious and obnoxious: everything it transfers should already be in the CRM, and nobody has ever enjoyed the information transfer of a sales meeting. You still need it. Even with no transfer of information, there is an important transfer of emotions and energy, and that is what the meeting is for.

Limits: my routine, not a studied one. The evidence in this section is about social pressure and sorting in general, not about standups.

For outbound-oriented teams, a commission motivates at 80-100% on top of base; 20-40% rarely moves anyone. The choices: uncapped usually beats capped, monthly beats quarterly where the cycle allows, individual beats team, team commissions are bonuses and gifts.

Pay a little and you get less than paying nothing. The cleanest experiment on this is twenty-six years old. Uri Gneezy and Aldo Rustichini sent Israeli high-school students door to door collecting for charity. Pairs paid nothing collected 239 shekels on average. Pairs promised 1% of what they raised collected 154. Pairs promised 10% collected 219, and the top collectors in the 10% group out-collected everyone.[6]

The paper's title is the finding: pay enough or don't pay at all. A small commission did worse than none because it turned a favour into a badly paid job.

The 80-100% is my experience, not a measured constant. Where those were the percentages at stake I saw the magic happen, in the form of extreme motivation, and where the commission sat at 20-40% I saw people treat it as weather.

The percentages are relative. 100% of a salary in Jakarta is a different amount of money from 100% of a salary in London, and I saw precisely the same behaviour in both places, driven not by the dollars but by the increase against the base. Gneezy's data gives the shape of the curve; the field studies in the chart below give the direction when pay moves by a lot.[4][7][8][9][10]

The market already pays this way. Hunters sit at a 50/50 mix, which is a commission at target equal to base, and the median SaaS account executive at variable worth 89% of base.[11][12] One warning travels with the rule: at 100% of base you have delegated the management of the rep to the scheme. That is fine for acquisition and a mistake for account management, where the work that matters is the work the plan can't see and 20-30% is usually enough.[13]

The design choices below are defaults, with the situations in which I would do the opposite. One number to keep in view: Alexander Group's typical 3x leverage means the top rep earns three times the target incentive, and the plan has to survive that year.[14]

scroll the table sideways →

Limits: every study in the table is one firm; none is a B2B software sales team on an annual quota, and the calls are mine.

Salary is a market function and can't be changed. The 80-100% commission seems to be a constraint from human psychology. Payback depends on market fit. The game is to triangulate salary, incentive and payback period into a point that works for the company.

Salary first, because it's the number you don't control. What a rep costs is set by the industry, the geography and the seniority you hire at, in roughly that order. Selling complex software or financial products to enterprises buys expensive people. The median B2B account executive in the US sits at $200K on-target earnings, mid-market SaaS reps at $160K to $220K, enterprise reps at $230K to $270K.[2][19]

Selling simple SaaS to small businesses buys cheaper people, and the same job in a cheaper geography costs a fraction of the US number. European plans run on more base and less variable than American ones.[19] You can move the cost a little with junior hires, but the market is moving the other way. Experience demanded at hire went from 2.7 to 3.7 years between 2022 and 2026, because a junior rep on a 6.2-month ramp is the most expensive rep there is.[2]

I ran acquisition teams in Jakarta and an account book in Dublin. The salary bands were different worlds; the shape of the commission that moved people was the same.

Limits: US-centred survey data; the Southeast Asian differential is one operator's sample; salaries move with the cost of capital and did in 2022.

Attainable second, because impossible targets mean no targets. A commission at 100% of base on a quota the rep believes is unreachable produces the motivation of no commission at all, at full cost. My calibration is a target a ramped rep hits about seven months out of ten. The market is far from that: 48% of B2B account executives hit quota in 2026, down from 66% in 2022.[2]

Payback third, and it depends on market fit. Salary is fixed by the market and the 80-100% by psychology. The only number that moves is what a rep produces, and that depends on the product, the price and the demand, not on the plan. If gross margin per rep covers salary plus commission inside a payback the balance sheet can carry, the 80-100% is affordable. If it doesn't, you have a problem, or you accept a not so aggressive sales force, which for many businesses is fine.

David Skok's rule that acquisition cost is recovered inside 12 months is the ceiling investors screen on.[20] The direction of the tradeoff is the opposite of what a spreadsheet suggests. A 100% plan pays back later, not sooner, because the rep is guaranteed the commission through a ramp that runs four to six months, so the cash hole is deeper and longer. What it buys is volume once the ramp ends.

A 30% plan carries no guarantee and a short ramp, so it pays back in a few months on moderate volume, and the volume stays moderate. The three scenarios below share the same market salary.

scroll the table sideways →

Read across the table and the tradeoff shows. The not so aggressive scenario pays back in month 4 and is the safest line on the chart. It is also the one that never produces the magic line, because a rep on 30% of base doesn't make the hundredth call.

The magic scenario is a better market, not a better plan. It digs the same hole as the average scenario, five months of base plus guaranteed commission, and climbs out of it in month 6 on volume the other two never see. The average scenario digs the hole and climbs out in month 9, a good business if capital is cheap and a slow bleed if it isn't.

When salary, an attainable 80-100% and a payback the business can carry all triangulate, magic usually happens. When they don't, because market fit is missing or capital is, you land somewhere still good but not perfect. The mistake is to write the plan as if you were in the first scenario. Stingy commissions born of a spreadsheet that assumed the magic margin are how plans get rewritten in month nine, with the reps watching.

The second triangle is process, culture and money. Every corner buys something and costs something, and I have worked in all four environments on the map. The one I aim for is the most expensive point on it.[5]

The strongest objection is that money doesn't move people at all. Cala, Havranek and colleagues corrected 44 economics experiments for publication bias and found a negligible effect.[21] Deci, Koestner and Ryan's meta-analysis found that performance-contingent rewards reduce intrinsic motivation.[22] Weibel and colleagues found a negative effect on interesting tasks.[23] Alfie Kohn argued in 1993 that incentive plans cannot work,[24] and Justin Roff-Marsh built a sales method on removing commissions, which I've reviewed separately.

My answer is narrow. The crowding-out results are strongest on interesting tasks with output that is hard to measure. Cold outbound acquisition is neither, and the field studies that show large effects are all about measurable, unpleasant work. Where the job is interesting or unmeasured, Deci and Kohn are right and the rule doesn't apply. I have no experiment that says 70% fails and 80% works, and I doubt one exists.

• Three layers: operations and structure, performance culture, sales incentives. Two win for a while. Three outlive the head of sales.
• Operations: the smallest process that prevents the mistake, executable by the laziest rep. Complexity is paid for in speed, time and morale, and the reps pay first.
• Culture: standup, weekly meeting, transparent results, praise, promotion, exit for persistent low performers. It's a theatre, and the energy it transfers is real.
• Incentives for outbound teams: 80-100% on top of base, attainable seven times out of ten, uncapped, monthly, individual. 20-40% rarely moves anyone.
• Salary is the market's, the 80-100% is psychology's, payback is the product's. When the three close, magic. When they don't, write the plan for where you are.
• Process plus money without culture works, less than expected. Process plus culture with low money is cheap and fragile. Culture alone with bad money, and you are running an MLM.

Cash flows and human psychology have not changed since homo sapiens appeared; know what is negotiable and what is not. Most plans get this backwards. They negotiate the constants, the market salary and the threshold at which money motivates, and they treat the spreadsheet as fixed. The reps notice before finance does, and the reps who notice fastest are the ones you hired for exactly that skill.