Annual sales planning usually starts the same way. A number arrives from the board or the CEO, RevOps builds a model that reaches it, territories get carved to fit, and quotas land on reps sometime in late January.

Every structural choice in that sequence still gets made. Coverage model, role design, territory rules, comp mechanics. They just get made implicitly, in service of a number that was set before anyone examined whether the structure could carry it.

Here is why that order produces a fragile plan, and the seven decisions that should come first.

I. Starting With the Number Means the Structure Gets Designed by Accident

A target is not a plan. It is one line in a plan, and it should be the last line, because it depends on everything above it.

When the number comes first, planning becomes reverse-engineering. If the model comes up short, someone raises the assumed win rate. If it is still short, ramp gets shaved from six months to four, and two heads appear in Q2 that nobody has a hiring pipeline for. That is arithmetic pressure applied to whichever variable is easiest to change in a spreadsheet.

The result is internally consistent and externally false. Reps feel that in week three, and once they believe the plan is fiction, the forecast becomes fiction too.

The fix is sequence. Decide who you sell to, how you cover them, what capacity supports, who does what, how accounts get assigned, what behaviour you pay for, and what you are dropping. Multiply that out and you get a number. It may be lower than the one the board wanted, which is the conversation worth having in September rather than the following October.

II. Seven Decisions That Should Produce the Number

These are ordered because they cascade. Each one constrains the ones below it.

1. Decide Who You Are Actually Selling To

Most ideal customer profiles are inherited. They were written when the company was smaller, and they get repeated in kickoff decks without anyone checking them against outcomes.

Pull the last eight quarters of closed-won and closed-lost. Not the pipeline, the closed data. Sort by segment, industry, size, and entry point, then look at four things:

• Win rate by segment: Where the same effort produces more closed business

• Cycle length by segment: Where cash arrives sooner and forecasting is steadier

• Loss reason mix: Whether you lose on price, capability, or absence of urgency

• Retention by acquisition source: Which wins are still customers two years later

The losses matter more than the wins. Losing to "no decision" means you have not found real urgency. Winning a segment that churns in eighteen months flatters this plan and damages the next.

If the ICP is changing for 2027, settle it first. Coverage, capacity, roles, territories, and comp all inherit it.

2. Decide the Coverage Model Before You Decide Headcount

Coverage is which selling motion touches which customer, and it is where the cost structure of the plan is really set.

Field, inside, partner-led, and self-serve motions carry different costs per dollar of revenue and different ceilings on deal size. The failure is rarely choosing wrong. It is choosing nothing, and letting every account get the same treatment because that is how the team is already organized.

Draw the account universe, then assign a motion to each band. Enterprise gets named field coverage. Mid-market gets inside sellers with a defined account load. The long tail gets partner or product-led motions. Write the rules for where the bands meet now, not in March when two reps are arguing over one logo.

3. Decide What Capacity Honestly Supports

Capacity planning is where most sales plans quietly break, because the arithmetic is done with generous inputs and never revisited.

Start with productive selling days. A year has roughly 260 working days. Remove statutory holidays, vacation, kickoff, training, and quarterly business reviews, and the real selling calendar is meaningfully shorter — and unevenly distributed, since December and August are weak in most B2B markets.

Then ramp. The Bridge Group's 2026 research across 158 B2B companies puts average AE ramp at 6.2 months, the highest in the study's history, attributed to more complex deals and larger buying committees. A rep hired in April is not carrying a full number in 2027.

Then attrition. Some share of the team will leave, and each replacement ramps from zero. A plan with no attrition assumption assumes something that has never once been true.

• Productive selling days: Working days minus holidays, leave, and internal events

• Ramp curve: What a rep produces in month three, six, and nine, from your own data

• Attrition and backfill lag: Your historical rate, plus days from departure to a new hire's start

• Fully-ramped output: The median, not the top performer's number

That last input does the most damage. A capacity model built on what your best rep produced last year is not a capacity model. It is an aspiration with decimal places.

4. Decide the Role Split, Including What AI Changes

Full-cycle sellers own the relationship end to end, which suits complex deals and smaller teams. Specialization — prospecting, closing, and account management as separate roles — raises throughput and lowers cost per activity, at the price of handoffs the customer can feel. Bridge Group's data shows the boundary moving: 38% of AEs now hold sole or blended ownership of renewal and expansion, a real shift back toward consolidated roles.

AI shifts the economics but does not settle the question. The workflows where it currently helps AEs most are post-call summaries, account research, and email personalization — task-level support inside an existing role. The same research rates most AI use cases as hit or miss, so a plan that cuts headcount assuming AI closes the gap is betting ahead of the evidence.

Ask the narrower question instead. Which parts of the role are research and admin that AI can genuinely absorb, and does absorbing them change how many accounts one person can carry? If yes, adjust account load in decision five. If no, say so and move on.

5. Decide Territories Without Burning the Relationships

Territory design gets treated as a January exercise in fairness, and the annual reshuffle usually costs more than the imbalance it corrects. Every reassigned account means a new introduction, a lost history, and a stalled deal. The customer notices, and some use the transition as a reason to reopen a competitive evaluation.

Design for stability. Set assignment by objective criteria, then carve out active opportunities and relationships past a defined threshold. Publish the rules so reps see logic rather than suspecting favouritism.

• Assignment basis: Geography, vertical, segment, or named list, chosen and stated

• Continuity rule: Which live deals and key relationships travel with the rep

• Rebalancing trigger: What imbalance justifies a mid-year change, and who approves it

Territory stability is not sentimentality. Continuity is a forecast input, and broken relationships show up as slipped Q1 deals.

6. Decide What You Are Actually Paying For

Comp plans are the clearest statement of strategy a sales organization makes, because reps read them more carefully than any deck. The common contradiction is asking for expansion revenue while paying almost entirely for new logos. Leadership sets a growth target that depends on the installed base, then builds a plan where the accelerator only triggers on net-new bookings. Reps do what the plan pays for, the expansion number misses, and it gets blamed on execution.

Work backwards from the revenue mix set in decisions one and two. If a third of 2027 growth comes from expansion, comp has to make expansion worth the same effort per dollar as new business. If you want multi-year deals, pay for term. And keep it explainable — a rep who cannot calculate their own commission on a whiteboard will not chase it.

7. Decide What You Are Going to Stop Doing

Every annual plan adds. New segment, new product line, new partner motion, new tooling, new reporting cadence. Almost none subtract, and the weight lands on the same reps.

Make the subtraction explicit and write it down. Which segment are you no longer pursuing? Which reports stop being produced? Which low-yield partner relationship ends? If nothing comes off, the new work gets absorbed by cutting selling time, and the capacity model in decision three is wrong the day the plan launches.

III. Every Decision Needs an Owner and a Lock Date

Decisions without a named owner and a deadline drift until someone makes them by default. This is the accountability grid.

Sequencing matters more than the exact dates. These decisions cascade, so changing an upstream one after the downstream work is done forces a rebuild. A coverage model revised in December invalidates the capacity model, the territory carve, and the quota allocation built on it — weeks of RevOps work, redone in the least productive fortnight of the year.

When something upstream genuinely must change late, name the rework openly and reschedule the steps below it. The costly version is the quiet one, where the change gets patched into the model and nobody recalculates what it broke.

IV. Work the Calendar Backwards From January 1

The target is simple to state and rare to hit: on the first selling day of January, every rep has a territory, a quota, a comp plan, and an account list. Work backwards from that.

• August: Assemble the data — win/loss by segment, ramp curves, attrition history, actual selling days.

• September: Lock decisions one through three. ICP, coverage, capacity. The load-bearing month.

• October: Lock roles and territories. Begin comp modelling in parallel, since it needs finance review time.

• November: Finalize comp, approve the stop-doing list, set the number. Executive approval lands here, not in December.

• Early December: Produce territory maps, account lists, quota letters, comp agreements. Brief managers first.

• Mid-December: Managers hold one-to-one conversations with every rep before the break.

• January: Kickoff reinforces a plan the team already knows. It does not introduce one.

Most organizations run about six weeks behind this schedule, and the compression always lands in the same place: comp plans that arrive in February. The cost is not administrative. Reps spend January prospecting against last year's incentives, and that month of pipeline is aimed at the wrong targets.

V. A Plan Reps Hear About in February Has Already Lost a Month

Communication is a planning step, not a follow-up to one. Frontline managers go first, in November, and they need more than slides. They need the reasoning, the trade-offs considered and rejected, and honest answers to the questions their reps will certainly ask. A manager who cannot defend the territory logic will side with their team against the plan.

Reps come next, individually, in December. A quota increase delivered in a group setting produces silence in the room and complaints afterward. One-to-one, with the reasoning attached, it produces a conversation. Every rep should leave able to state their number, their accounts, and how they get paid.

The wider organization comes last, at kickoff. By then the plan should be old news to the people executing it. Kickoff is for energy and shared context, not the first reading of a document that governs someone's income.

The Real Risk Is a Number With Nothing Underneath It

A plan built backwards from a target does not fail loudly in January. It fails slowly, and the diagnosis arrives too late to act on.

• Bad assumptions surface as missed quarters. By the time Q2 confirms the model was wrong, half the year is gone.

• Late comp plans cost real selling weeks. January prospecting aimed at last year's incentives is January mostly wasted.

• Reshuffled territories show up as slipped deals. Broken relationships are a forecast problem, not just a morale one.

• Unsubtracted work eats the capacity model. Every addition without a removal quietly shrinks selling time.

Seven decisions, made in order, with owners and dates. That is the whole method. It needs no bigger planning team and no better model, only structural choices made early enough that the number becomes a conclusion rather than an assignment.

Set the number first and you will spend the year defending it. Make the seven decisions first and you will spend the year executing against something the structure can actually carry.

Ready to Build a 2027 Plan Your Team Can Actually Execute?

Most planning cycles run out of time before they run out of decisions, and the ones that get skipped are structural. Coverage, roles, territories, and comp are harder to fix in March than to design in September.

At Change Connect, we specialize in sales transformation and go-to-market design — sequencing planning decisions, building honest capacity models, and aligning compensation with the strategy it is meant to serve. Book a Strategic Audit with Change Connect today and start 2027 with a plan your reps believe in.

Frequently Asked Questions

When should annual sales planning for 2027 start? Data gathering should begin in August, with ICP, coverage, and capacity decisions locked by the end of September. Working backwards from a January 1 launch, comp plans need executive approval in November so documents can be produced and delivered to reps in December.

What is the right order for sales planning decisions? ICP first, then coverage model, then capacity, then role design, then territories, then compensation, then the stop-doing list. Each decision constrains the ones after it, so changing an upstream choice late forces the downstream work to be rebuilt.

How do you build a realistic sales capacity model? Start with productive selling days after holidays, leave, and internal events. Apply your own historical ramp curve rather than a generic one, subtract expected attrition and backfill lag, and use median fully-ramped rep output rather than top-performer output.

How long does it take a new sales rep to ramp? The Bridge Group's 2026 research across 158 B2B companies reports average AE ramp of 6.2 months, the highest in the study's history, driven by more complex deals and larger buying committees. Use your own historical data where you have it, and treat mid-year hires as partial-year contributors.

When should reps be told about their 2027 quota and territory? Frontline managers should be briefed in November and reps in December, one-to-one, before the holiday break. Plans communicated at a February kickoff have already cost the organization a month of correctly aimed prospecting.