How do I structure AE compensation across regions with different cost-of-living and market rates in 2027?

The two models you are actually choosing between

Every company expanding past its home market lands on the same fork, whether or not anyone names it out loud. Option one is the flat global plan: one on-target earnings number, one quota, applied everywhere, on the theory that identical pay for identical titles is the fairest thing a company can do. Option two is the geo-tiered, quota-normalized plan: a small number of pay tiers indexed to local cost-of-labor, with quotas built bottom-up from each region's actual selling conditions and a currency policy sitting underneath both.

The flat plan is not stupid. It is what almost every company does by default, because it is what you get when you take the headquarters plan and copy it into a new offer letter. It has real advantages: nothing to explain, nothing to defend, no awkward conversation about why a rep in one city earns less than a rep in another. It survives a Glassdoor leak without incident. For a company with eleven AEs across three countries, it is very often the right answer, and the counter-case section below says so plainly.

What kills the flat plan is scale plus asymmetry. A $180,000 OTE is generous in Lisbon, roughly at-market in Austin, and thirty to forty percent below market in San Francisco or Zurich. Apply it everywhere and you have not achieved fairness — you have achieved a systematic overpayment of your lowest-leverage markets funded by a systematic underpayment of your highest-leverage ones. And the underpaid markets are, without exception, the ones with the largest deals, the deepest competitive sets, and the most aggressive recruiters. The flat plan does not remove unfairness from the system. It relocates all of it onto the markets the plan's author cannot see from their desk, which is precisely why the author never notices.

The geo-tiered plan costs more to design and far more to explain. Its advantage is that it is defensible from every seat rather than just the headquarters seat. A rep in any market can ask "why is my number this number?" and get an answer grounded in something external and checkable — what three named competitors would pay to hire them locally — rather than an answer grounded in "that's what everyone gets." The design principle underneath it is a redefinition of fairness: equal opportunity to earn at the same effort percentile, not equal dollars. Two AEs who both land at the 60th percentile of their quota should both earn the 60th-percentile slice of their region's OTE. That is a stricter fairness standard than flat pay, not a looser one, and stating it that way is what wins the rollout.

There is a third position worth naming because a surprising number of companies drift into it accidentally: the half-tiered plan, where OTE gets regionalized but quota does not. This is the worst of the three. It takes the hard, unpopular part of geo-tiering — telling a rep their number is smaller — and pairs it with none of the compensating adjustment that makes the smaller number legitimate. A Mexico City AE on a tiered-down package carrying a flat London quota has been handed the compensation of an emerging market and the expectations of a mature one. Reps read this instantly and correctly as the company taking the savings and skipping the work. If you are not going to normalize quota, do not tier OTE either; stay flat and accept the cost.

Cost-of-living versus cost-of-labor, and why the distinction decides everything

The single conceptual error that produces bad regional plans is anchoring pay to cost-of-living. Cost-of-living measures what it costs an employee to live somewhere. Cost-of-labor measures what it costs an employer to hire someone there. They correlate loosely. They are not the same number, and compensation must track the second one.

Consider a senior AE in Toronto. The cost of living in Toronto is meaningfully lower than New York's — the housing delta alone is substantial. But the cost of labor, meaning the package a competing software company would put in front of that AE to poach them, is only modestly lower, because the enterprise-AE talent pool in Toronto is thin and globally mobile. A plan that discounts Toronto by the cost-of-living gap rather than the cost-of-labor gap will lose its Toronto AEs to the first competitor that prices labor correctly, and it will lose them fast, because the gap between those two discount rates is exactly the size of a compelling counteroffer.

The reason this matters operationally rather than semantically is that the two numbers diverge most in exactly the markets where being wrong is most expensive. In a low-cost, deep-talent market, cost-of-living and cost-of-labor sit close together, so anchoring on the wrong one barely hurts. In a low-cost-of-living, scarce-talent market — Toronto, Tel Aviv, increasingly Lisbon, and Bangalore for genuinely senior enterprise sellers — the gap is wide, and it is wide in the direction that bites. Cost-of-living says "pay them less." Cost-of-labor says "pay them roughly market." A company anchored to cost-of-living builds a plan that is cheapest precisely where talent is hardest to hold. That is not a philosophical error. It is a targeting error that aims the company's underpayment directly at its most fragile hires.

Purchasing power parity belongs in the same discard pile as cost-of-living. It is useful macro context and a terrible pay input, for the same reason: it describes the employee's side of the transaction, and pay is set on the employer's side. The operational expression of cost-of-labor is the local market rate for the role at the level — median total compensation for a senior enterprise AE in that metro. That is the number you buy from a benchmark vendor, and that is your OTE.

A discipline worth institutionalizing: make the cost-of-labor question literal in every compensation review. Name the three companies most likely to poach this AE, and the package those three would actually put on the table. If RevOps cannot answer that for a given market, the company does not have a defensible number for that market — it has a guess wearing a spreadsheet. Cost-of-labor is not a dataset you buy once and file. It is a question you commit to keeping answerable.

This same logic travels upward and downward through the org, which is worth knowing because regional AE comp is rarely designed in isolation. Sales engineers, customer success managers, and SDRs all face the same cost-of-labor-versus-cost-of-living question, and the answer is the same, but the *tier placement* frequently differs by role. A metro can be a premium engineering market and a mid-tier sales market simultaneously — Tel Aviv is the textbook case. Pulling a generic city cost index and applying it across every function is the shortcut that produces a plan wrong for most of the people it covers. Always pull the role-and-level-specific cut, and expect the resulting tier maps to disagree across functions.

How to decide between flat and tiered

The decision is not a matter of taste. There are four tests, and they resolve cleanly in most cases.

Test one: population per region. Below roughly five AEs in a region, a tier model is administrative overhead with no statistical payoff — you cannot run an attainment-distribution audit on four people and learn anything. Pay each AE at their individual local market rate, hold mix and curve constant for fairness, and formalize when the population is large enough that distributions carry signal. A company with eleven AEs across three countries does not have a regional compensation problem. It has a hiring-negotiation problem, and it should solve that one instead.

Test two: is the talent market genuinely single? Some fully remote companies recruit senior enterprise AEs who could live anywhere and frequently relocate mid-tenure. If the company competes for the same candidate regardless of city, then cost-of-labor genuinely is uniform and tiering is a fiction the data does not support. This test is empirical, not philosophical: look at offer-acceptance rates by proposed tier. If your Tier 4 offers get rejected at the same rate as your Tier 1 offers, the tiers are not real, and you should pay one global band and stop pretending.

Test three: will leadership defend the model in public? In a transparency-forward culture, geo-tiering reads as "paying people less for where they sleep" unless someone is willing to stand up and make the cost-of-labor argument repeatedly and without flinching. The defense is available and it is strong — the opportunity is identical, the quota is normalized, the pay is at-market for each person — but it has to actually be made. A hidden tier system is strictly worse than no tier system. If the exec team will not defend it openly, stay flat.

Test four: currency volatility. If a market's currency is volatile enough that a re-rating band would trip every quarter, the banded model breaks down, because constant re-rating is itself destabilizing. In those handful of frontier markets, the honest answers are a hard-currency package at higher cost, or partner coverage instead of a quota-bearing AE.

Notice that the final node applies to every branch. Whatever you decide about OTE, the pay mix and the accelerator curve stay constant. That is the one piece of the architecture that does not vary by any of the four tests.

The spine: what stays constant no matter which model you pick

Two things are identical across every region, in every version of the plan: the pay mix — the base-to-variable split — and the accelerator curve, meaning how commission rates change above and below quota. A 60/40 mix in Chicago is a 60/40 mix in Chennai. A 2x accelerator above 100% attainment in Chicago is a 2x accelerator in Chennai.

Three reasons this is non-negotiable. The mix encodes the role's risk profile, and a 50/50 AE is a genuinely different job than a 70/30 AE — the job should not change because the AE crossed a border. A constant curve means a sales leader can compare attainment distributions across regions cleanly, because the incentive gradient is identical everywhere, which turns the distribution into a diagnostic instead of noise. And constant mix and curve are what make the plan feel fair to reps who compare notes across regions, which in a remote-first company they absolutely do, usually within a week of plan documents going out.

The mix itself should be set by sales motion, not geography. A transactional, high-velocity motion — small deals, short cycles, many at-bats — runs well at 60/40 or even 70/30, because volume means variable pay tracks effort tightly. An enterprise motion — six-figure deals, nine-month cycles, three or four closes a year — runs better at 50/50, because one slipped deal in a low-at-bat job should not wipe out a rep's mortgage payment. If a company runs both motions in one office, which is extremely common, it should run two mixes in that office rather than forcing one onto both. Geography did not create the difference; the motion did.

Conflating those two axes is a common and expensive error. A leader looks at the London office, sees enterprise reps out-earning commercial reps, and concludes "London compensation is high." London compensation is not high. London simply hosts more enterprise headcount, and enterprise compensation is high everywhere. If the company then "fixes" London by trimming OTEs, it has cut enterprise pay below market for a reason that has nothing to do with enterprise. The clean mental model is a grid: geo tier on one axis, sales motion on the other. A rep's plan sits at the intersection. OTE comes from the tier; mix and curve come from the motion. Keep them separate on paper and every review stays legible.

A defensible curve for a 50/50 enterprise AE runs decelerated below half attainment — roughly 0.6x of pro-rata, which protects margin on a bad year without being punitive — then linear from 50% to target, then a 2x marginal multiplier from 100% to about 130%, then 3x beyond that, uncapped, with a review trigger on genuinely outsized deals. Those four numbers apply identically in every region. Only the dollar value of "variable target" changes, because OTE changes.

The curve stays uncapped on purpose. Capping commission tells your best reps to stop selling in Q4 and sandbag into next year, which is the most expensive behavior any compensation plan can buy. And the curve stays constant across regions for a subtler reason: the accelerator curve is the company's public statement about how much it values overperformance. If that statement varies by region — 2x in New York, 1.5x in Bangalore — the company is saying an overperforming Bangalore AE is worth less than an overperforming New York AE. Reps detect this immediately and read it correctly as "we expect less of you here." The best emerging-market AEs, the ones most able to leave, read it fastest. Ambition cannot be regionally discounted.

The mirror-image error is a regionally *steeper* curve introduced to "motivate" a lagging market. If the market lags because its quota is mis-normalized, a steeper curve does not fix the quota — it makes an unfair quota slightly more lucrative for the few who beat it and does nothing for the majority who cannot. The curve is not the lever for regional underperformance. The quota is.

The concrete numbers behind each option

Tier count first: three to five, never per-city. Five is the practical maximum, beyond which tiers blur into per-city pay and you lose the entire administrative advantage.

Per-city pay is seductive because it feels rigorous — Austin gets the Austin number, Denver the Denver number, Raleigh the Raleigh number. In practice it is quicksand. It means maintaining and defending a live market rate for every city where you employ even one AE. It means every internal transfer becomes a compensation event with a winner and a loser. It means a rep relocating from Denver to Austin for personal reasons opens a negotiation about a raise or a cut that has nothing to do with their performance. It means the annual benchmark refresh is a multi-week reconciliation of dozens of independently drifting numbers rather than a re-rate of a handful of indices. And it means you are perpetually one forum post away from a rep in city A discovering they earn four percent less than a near-identical peer in city B, with no clean story for why. Each is a small tax; together they are a structural drag. GitLab publishes its entire compensation calculator on the open web using a small number of location factors rather than hundreds of city rates, and Stripe publicly consolidated toward fewer bands after finding granular city pay generated friction without measurable retention benefit. Tiers beat cities.

A four-tier structure covers most global software companies. Working from a Tier 1 reference OTE of $260,000 for a senior enterprise AE at a 50/50 mix, with indices of 1.00 for premium talent markets (San Francisco, New York, Zurich, London), 0.85 for major markets (Austin, Boston, Toronto, Sydney, Singapore), 0.72 for secondary markets (Denver, Lisbon, Dublin, Tel Aviv), and 0.55 for emerging or cost-advantaged markets (Bangalore, Kraków, Mexico City, Manila), you get: New York at $260,000 split $130,000 base and $130,000 variable; Toronto at $221,000 split $110,500 and $110,500; Lisbon at $187,200; Bangalore at $143,000. Every one of those reps runs the same 50/50 mix and the same 0.6 / 1.0 / 2.0 / 3.0 curve. A plan refresh updates one reference number and four indices, and every plan re-derives automatically — a half-day exercise rather than a multi-week project.

One detail in that index spacing is real rather than illustrative: the gaps are wider at the bottom than the top. Tier 1 to Tier 2 is fifteen points; Tier 3 to Tier 4 is seventeen. Cost-of-labor compresses at the high end, because premium markets converge as senior talent becomes globally mobile and globally priced, and it spreads at the low end. A company that imposes even spacing — 1.00, 0.83, 0.66, 0.49 — will tend to overpay Tier 2 and underpay Tier 4 relative to what the benchmark data actually says. Let the data set the spacing; do not impose arithmetic regularity on a distribution that is not regular.

Now the quota side, which is where the flat-plan-versus-tiered-plan comparison gets decided. Geo-tiered OTE without geo-normalized quota is a broken plan in a nice suit. A quota is a claim about how much revenue one AE can reasonably produce in a year, and that number is a function of four local variables: addressable market, win rate, average deal size, and sales-cycle length. Those spread across regions by factors of two to five. Addressable market alone commonly varies three to five times between a mature and an emerging region. Win rate varies one and a half to two times. Deal size, two to three. Cycle length, one and a half to two times, with an inverse effect on capacity — longer cycles mean fewer deals fit in a compensation year.

A workable normalization: regional capacity equals working selling days divided by average cycle length, multiplied by win rate, multiplied by local average deal size; regional quota equals that capacity times a coverage ratio. The coverage ratio — typically 3.0 to 3.5 — stays constant across regions, because it is a company-level risk parameter, not a regional one. Working selling days stay constant too. Only cycle length, win rate, and deal size vary. That is the same hold-constant / let-float discipline from the pay-mix spine, applied to quota instead of pay.

Run it on two real-shaped cases. A London senior enterprise AE: 230 selling days, 95-day average cycle, 24% win rate, $88,000 average deal — roughly $2.42M of capacity, times a 3.2x coverage ratio inverted into quota, lands near $1.18M. A Mexico City AE in the same role: 230 selling days, 130-day cycle, 19% win rate, $61,000 average deal — roughly $2.05M capacity, same 3.2x, lands near $810,000. A flat global plan would have handed both a $1.0M number, overshooting Mexico City by about 23% and undershooting London.

The fastest sanity check on the whole system is the quota-to-OTE ratio, which should land in a similar band across regions — three to five times for enterprise software. London at $221,000 OTE against $1.18M quota is 5.3x, top of band but defensible. Mexico City at $143,000 against $810,000 is 5.7x, in band, internally consistent. Now break it deliberately: Mexico City at $143,000 OTE against a flat $1.18M quota is 8.3x, screamingly out of band. That single number catches the most expensive class of error in the entire system — the half-tiered plan — before any rep ever sees a plan document. It costs one spreadsheet column. The ratio works as a diagnostic because it closes the loop on the two numbers that are supposed to vary: a smaller market justifying a lower OTE also justifies a proportionally lower quota, for the same underlying reason, so the ratio should stay roughly stable when both moved correctly. Divergence means one moved and the other did not.

The aggregate budget guardrail sits above all of this: a well-run software company spends roughly nine to eleven percent of regional revenue on AE variable compensation, and that ratio should hold across regions. Geo-tiering changes the *allocation* of that spend, not the total. A Tier 4 region running at six percent is underpaying and will bleed talent; a region at fifteen percent has either a quota set too low or an OTE set too high. Check the ratio per region and you catch drift that the tier indices alone will not surface.

Finally, currency, which is where otherwise-good plans quietly fail. Reps are paid in local currency, always — an AE in Tokyo in yen, in Warsaw in złoty. Paying a local employee in dollars exports the company's FX risk straight onto their paycheck, and their rent is not denominated in dollars. It is frequently a legal and tax requirement besides. But quotas usually must be denominated in dollars for consolidated forecasting, which means a strong-dollar quarter silently makes a euro-region AE's job ten percent harder through no fault of their own.

The fix is a banded re-rating policy: within plus or minus five percent of the plan-set rate on a trailing-90-day average, nothing changes and the company absorbs the noise; five to seven percent is a watch band flagged in the quarterly review; beyond seven percent, quota and OTE re-rate to the new trailing average. High-volatility currencies run a tighter collar, around three percent. The band is a deliberate compromise between two failure modes. Re-rate on every spot move and the plan changes monthly, destroying the one thing a comp plan must supply — a stable target a rep can run a six-month enterprise cycle against. Have no policy at all and the quota freezes in dollars while the local economy moves underneath it. The band gives stability inside the collar and correction outside it.

Two operational details keep the band from becoming a grievance. It must be symmetric — if the dollar weakens past the band, quota re-rates up as mechanically as it re-rates down when the dollar strengthens. A company that re-rates up in a strong-dollar year and forgets to re-rate down in a weak one has built a one-way ratchet, and reps notice within a cycle. And it must be automatic rather than discretionary, because discretion is where trust leaks. Alongside the band, the division of labor on hedging is clean: corporate finance hedges the aggregate regional compensation exposure with forwards or natural hedges; the rep hedges nothing. Finance can manage that risk at portfolio scale for a fraction of the morale cost of pushing it onto individuals. High-inflation markets get one more adjustment — a six-month benchmark refresh instead of annual, because local market rates erode mid-year and an annual cadence leaves those reps behind for two quarters at a time.

Implementation and sequencing

Moving from a flat plan to a geo-tiered, quota-normalized, FX-banded one should be sequenced. Doing it all at once produces a plan nobody trusts, for a reason that goes beyond project hygiene: if you simultaneously re-tier OTE, re-normalize quota, change the mix, and introduce FX bands, and attainment then moves, nobody — not the reps, not RevOps, not the board — can attribute the movement to anything. The plan becomes unfalsifiable, and an unfalsifiable compensation plan cannot be tuned. Sequencing is what makes the feedback loop legible.

Phase one — data and tiers. Buy or refresh AE-specific, level-specific cost-of-labor data, then place metros into tiers. Budget two to four weeks plus a benchmark subscription. Two placement rules govern: use the role-specific cut rather than a generic city index, and place on cost-of-labor while sanity-checking against talent depth. If the data says Tier 3 but hiring managers can only close candidates at Tier 2 money, the pool is thin and the metro is functionally Tier 2. Markets self-correct slowly; the plan should respect what recruiting actually experiences.

Phase two — spine and quotas. Lock mix and curve first, then build normalized quotas region by region. This is the analytically heaviest phase, and regional sales leaders must be in the room, because they hold the ground truth on cycle length and win rate that the formula consumes. One caution on inputs: they must be measured, not borrowed. The most common normalization failure is using headquarters win rate and cycle length as a stand-in for a new region because the new region "doesn't have enough data yet." That assumption runs exactly backward — a young region almost always has a longer cycle and lower win rate, because brand is weaker and the team is less seasoned. Porting headquarters inputs produces a quota that is too high, which produces missed attainment, which produces a board that concludes the market is weak and cuts its headcount plan, which thins the pipeline for the next rep. A quota-design error can quietly kill an entire continental expansion thesis. When a region genuinely lacks data, set a deliberately conservative ramp quota and re-normalize the moment real local numbers exist.

Phase three — FX and rollout. Set the bands, assign the hedge, and write per-rep plan documents that state tier, OTE, mix, quota, curve, and FX band explicitly. Communicate in the rep's terms, though. A document leading with tier taxonomy and indexing methodology loses the reader on page one. A document leading with *their* number, *their* quota, *their* curve, and a plain-language paragraph on why their number is what it is — anchored to their market's cost-of-labor — lands. Make the full machinery available to anyone who asks, in complete transparency, but do not make it the headline. The headline is always: here is your number, here is your opportunity, here is the honest reason it is that number.

Phase four — audit and iterate. Two quarters in, plot the distribution of quota attainment by region. In a healthy plan, every region's distribution has roughly the same median and shape, because the curve is identical and the quota is normalized — attainment should be region-agnostic. Signals to read: regional medians within about eight points is healthy, a twenty-plus-point gap means the quota model is wrong; a bimodal distribution in one region means territory or quota inequity inside it; elevated voluntary attrition in one region means OTE is below local market; a quota-to-OTE outlier means a math error somewhere in that region's model; one region overloaded in the 0–50% attainment band means quota exceeds local capacity.

Commit to the interpretation rule *before* you run the audit: a persistent regional median gap is read as a quota error first, and only as a talent or territory issue after the quota model has been re-checked and cleared. The lazy reading runs the other direction — a leader sees low attainment and concludes the region has weak reps or is a hard market, because that conclusion requires no rework. It is occasionally right, usually wrong, and self-fulfilling when wrong. Forcing the quota hypothesis to be tested and rejected first is the procedural safeguard that keeps the audit honest. Be precise about "healthy," too: some spread is expected. A brand-new region mid-ramp sitting below a mature one is a ramp effect, not a defect. The audit is hunting for persistent, mature-state gaps — regions running several quarters with stable headcount that still show a twenty-point median gap. A young region's low attainment is a footnote. A three-year-old region's low attainment is a finding.

Ownership should be written down alongside the model, because ambiguity here is where plans rot between cycles. RevOps owns the model, the tiers, the normalization formula, and the audit. Finance owns the compensation budget, the FX hedge, and the cost ratio. People/HR owns the benchmark subscriptions and level definitions. Regional sales leaders own per-rep quota review and surface ground truth on talent depth. The CRO owns the framework and signs off on tier placement and the annual refresh. Territory assignment inside a region is a related governance question and should be answered consistently across regions even though the quotas themselves differ.

Related questions

Should quotas be denominated in local currency instead of dollars?

Usually not — consolidated forecasting needs one currency. Denominate quota in the reporting currency, pay in local currency, and let the FX band absorb the mismatch. Local-currency quotas make regional roll-ups messy and shift translation risk into the forecast rather than removing it.

How do you handle an AE who relocates between tiers mid-year?

Grandfather the current year's OTE and quota through the plan period, then re-rate at the next annual cycle. Mid-year re-rating creates a negotiation and disrupts an in-flight quota. Publish the relocation policy in advance so nobody discovers it during a move.

Do SDRs and sales engineers use the same tier map as AEs?

Same methodology, frequently different tier placement. Pull role-and-level-specific cost-of-labor data per function. A metro can be premium for engineering and mid-tier for sales, so a shared city index applied across functions is wrong for most of the people it covers.

What if benchmark data does not cover a market at all?

Triangulate from adjacent metros in the same labor market, then validate against actual offer-acceptance data from your own recruiting. Treat the placement as provisional, flag it for review at the next cycle, and correct once you have real hiring evidence rather than defending a guess.

How often should tier indices themselves change?

Annually for most markets, semiannually for high-inflation ones. Tier *placement* for a metro should change rarely — only when hiring evidence contradicts the data over multiple cycles. Frequent tier reshuffling destroys the stability that makes the model administrable.

FAQ

Does geo-tiering mean paying people less for the same work?

In a trivial sense the number is smaller; in every sense that matters, no. The work is not identical — a rep in a different region sells into a different market, against a different competitive set, with a different deal-size distribution, and carries a quota normalized to that reality. Each rep's pay is at-market for them, meaning neither could improve their situation by walking across the street. The opportunity — mix, curve, percentile mechanics, upside — is identical. What differs is a dollar figure that would mean radically different things in two economies. A company that can articulate this confidently keeps its rollout; one that cannot loses to the soundbite.

Can we cap commissions in high-cost regions to control spend?

No. Capping tells your best reps to stop selling once they approach the cap and push deals into next year, which is the single most expensive behavior a compensation plan can purchase. If regional spend is running hot, the problem is a quota set too low or an OTE set too high — fix the input. The nine-to-eleven-percent-of-revenue budget guardrail is the right control, applied per region, and it works without touching the curve.

How many geo tiers is too many?

Beyond five, tiers stop being tiers and become per-city pay with extra steps. Each additional tier adds benchmark maintenance, more edge-case placements to defend, and more relocation events that trigger comp changes, while the retention benefit over a well-designed four-tier model is not measurable in the published operator experience. Three tiers is plenty for a US-only company; four covers most global software companies.

What is the fastest way to tell whether our current regional plan is broken?

Compute quota divided by OTE for every region and look for anything outside a three-to-five-times band, then plot regional attainment medians and look for gaps above twenty points among regions with at least two quarters of stable headcount. Those two checks take an afternoon and catch the large majority of structural errors, including the half-tiered plan, which is the most common and most damaging failure mode.

Who should own the FX hedge — finance or RevOps?

Finance, without ambiguity. Corporate treasury can hedge aggregate regional compensation exposure with forwards or natural offsets at portfolio scale. RevOps owns the re-rating band that protects the individual rep's quota and OTE. The rep owns none of it. Splitting it any other way pushes macro risk onto people who have no instrument to manage it.

Should the accelerator curve ever differ between an enterprise team and a commercial team in the same office?

Yes — curve and mix follow the sales motion, not the location. Two teams in one office running different motions should run different mixes and can reasonably run differently shaped curves, because at-bat counts differ dramatically. What must not vary is the curve between the *same* motion in *different* regions. Motion is a legitimate axis for curve differences; geography is not.

Sources

• https://about.gitlab.com/handbook/total-rewards/compensation/ — GitLab's public compensation handbook and location-factor methodology
• https://www.wtwco.com/en-us/solutions/services/compensation-survey — WTW compensation survey methodology and benchmarking data
• https://www.mercer.com/solutions/total-rewards/ — Mercer total rewards and global compensation planning research
• https://www.worldatwork.org/ — WorldatWork sales compensation programs and practices research
• https://hbr.org/2015/04/motivating-salespeople-what-really-works — Harvard Business Review on sales incentive design and accelerators
• https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights — McKinsey growth, marketing and sales insights including compensation research
• https://www.bain.com/insights/topics/customer-strategy-and-marketing/ — Bain go-to-market and incentive design research
• https://www.gartner.com/en/sales — Gartner sales practice research including quota and compensation benchmarks
• https://www.pave.com/ — Pave real-time compensation benchmarking and location-tier data
• https://carta.com/data/ — Carta compensation datasets structured by location tier

Related on PULSE

• How do I build a true bottom-up revenue forecast by region?
• Who should own territory and quota reassignment — the regional manager or the CRO?
• How do I design compensation for a company running two different sales motions?
• How do I audit whether a comp redesign actually improved deal quality?
• When should a company formalize its sales compensation structure?
• What are CRO compensation benchmarks by company stage?