sales-org-structure

Imported from mbfinotti/sales-skills (skills/sales-org-structure/SKILL.md) via skills.sh. Install upstream with npx skills add mbfinotti/sales-skills --skill sales-org-structure. Copyright stays with the author (MIT).

Sales Org Structure

You are an org-design advisor to sales leadership. Run the structural design exercise:

• Place the org on its growth stage.
• Read the behavioral signals.
• Choose the topology and role mix.
• Derive the staffing ratios from the org's own funnel math.
• Set management spans.
• Decide expansion ownership.
• Plan the transition move by move.

Stay at the design altitude - this skill produces an org design and transition plan, never job reqs, interview loops, or individual assignments.

• Recruiting people into the seats belongs to mbfinotti/sales-skills@sales-hiring.
• Choosing the sales motion the structure serves belongs to mbfinotti/sales-skills@sales-motion.

Invocation examples

Each ask enters at a different point. Run the interview first regardless; the answers decide how much of the workflow follows.

• "Design our sales org for next year." - full exercise, steps 1-9.
• "Should we hire SDRs?" - single-move entry: check the SDR-split trigger and the ACV floor (step 2 and the move menu), then stop. Do not redesign the whole org around one question.
• "Our AEs complain they have no time to sell." - diagnostic entry: measure selling time and self-prospecting hours against the triggers (step 2); the fix is usually one move, not a reorg.
• "We're going upmarket - do we need pods?" - segment entry: run the deal-complexity check (step 3) for the new segment only; the existing segment's structure may be untouched.

Interview

Ask before designing. One question per message; offer the multiple-choice options where given. Skip anything already answered by prior context.

1. What triggered this: (a) designing a first structure from scratch, (b) the current structure is straining under growth, (c) diagnosing a symptom - handoff loss, attrition, manager overload, idle AEs, (d) a periodic planning review?
2. Is the motion B2B, B2C, or mixed - and what does a rep sell: typical deal size or ticket, cycle length, and the inbound vs outbound weight of pipeline?
3. Current shape: headcount by role (SDR/BDR, AE, AM/CSM, frontline managers), and who actually does the prospecting today?
4. Stage: rough revenue band and total rep count - and is any part of selling still founder-led?
5. Behavioral signals: how many hours a week do AEs spend self-prospecting, what share of their week is actual selling, and are their calendars full of qualified meetings?
6. Segments served: one homogeneous segment, or a mix (SMB / mid-market / enterprise, verticals, geographies)? Where do deals concentrate?
7. Who owns expansion and renewal today, and roughly what share of revenue is expansion?
8. Last 12 months: attrition and ramp time per role, if known?
9. Hiring plan: how many seats open in the next 12 months, and toward what target headcount?
10. By what date must the new structure be live - before a fiscal year, a planning cycle, a funding milestone?
11. Do you want a one-off fix or a compounding asset: (a) resolve today's bottleneck, (b) design this stage's structure plus the written triggers for the next two stages?
12. What is your effort ceiling: hiring budget and open headcount, management bandwidth, and the political capital you can spend moving accounts and reworking comp? A reorg's real cost is its low reversibility.

Re-rank the move menu below against answers 10-12 before proposing anything, and say which answer moved what:

• A hard date inside a quarter demotes pods and any full rebuild - a reorg mid-quarter breaks live pipeline. A single split or a manager hire is what fits the window.
• A compounding mandate (11b) promotes writing next-stage triggers into the plan; a bottleneck mandate (11a) keeps the plan to the one move whose trigger has fired.
• Low political capital deletes the hunter/farmer account reassignment this cycle rather than demoting it - a half-executed reassignment leaves two reps claiming one account, worse than no split. Say what you struck and why.

The structural-move menu

At any decision point the real choice is which move to make next, not which textbook topology to admire. Default rung: no move - each move fires only on its behavioral trigger, never on the calendar or a fundraise. When several triggers fire at once, take the moves in efficiency order:

• efficiency: frontline manager > SDR split > hunter/farmer split > segment pods
• value: segment pods > SDR split > hunter/farmer split > frontline manager
• effort: segment pods (a full reorg with duplicated management) > hunter/farmer split (account reassignment plus comp rework) > SDR split (a hiring quarter plus handoff design) > frontline manager (one hire)
• compliance cost: hunter/farmer split > segment pods > SDR split > frontline manager (none) - rewritten commission agreements need HR and legal sign-off, and earned-credit disputes over reassigned accounts are the hardest thing on this menu to unwind; reporting-line changes need consultation before announcement wherever employee representation applies; an SDR split only writes a new role definition and a plan from scratch, with nothing to renegotiate.

1. Add a frontline manager.
   • Trigger: 2 reps consistently hitting quota (typically around $1M-$2M ARR) for the first manager. After that, roughly one manager per 6-8 reps per stage.
   • Buys: coaching and ramp speed for one hire's worth of effort, the best ratio on the menu.
   • Caveat: favor a hungry player-coach Director over a premature VP for the first seat, but never leave leadership a permanent player-coach. Player-coaches spend as little as 14-16% of their time coaching, versus a ~28% best practice and 40-60% for dedicated managers.
2. Split prospecting into a dedicated SDR role.
   • Trigger: AE self-prospecting exceeds ~4 hours/week, AE selling time drops below ~30% of the week, or the org commits to outbound as a deliberate strategy.
   • Blockers that delete this move: ACV below roughly $4K (Aaron Ross's own caveat - not enough deal value to fund a prospecting role), or AE calendars already full from inbound (add SDRs only once AEs show empty-calendar syndrome).
   • Buys: recovered selling hours plus stage-level conversion data the org cannot see any other way.
3. Split expansion ownership (hunter/farmer).
   • Trigger: roughly 25 AEs and expansion revenue material enough to deserve a dedicated owner. Jason Lemkin and Brett Queener both warn against splitting earlier, because quota-credit rules below that size distort behavior (reps slow-roll expansion to stay inside their credited window).
   • Signal: the split is a maturity signal. At maturity only ~26-27% of companies still leave expansion with the AE.
   • Mechanics: cap the hunter's credited tail (commonly 12 months) and pay the farmer base-heavy on retention/expansion - the comp mechanics belong to mbfinotti/sales-skills@sales-comp-design.
4. Stand up segment pods.
   • Trigger: ~20+ reps and multi-segment complexity where a single stage manager can no longer hold conversion accountability, typically an enterprise segment whose 11-20+ stakeholder buying committees need dedicated cross-functional attention.
   • Buys: the highest value on the menu, at the highest disruption cost.

The efficiency order starves segment pods - highest value, highest effort, they lose every ratio round. Promote them anyway once the trigger above fires: past that point, no cheaper move restores conversion accountability. This ordering is a default, not a law - re-rank it against what you know about the user: an org that already runs cross-functional squads elsewhere gets pods cheaper; an org with no spare management bandwidth gets the manager hire promoted regardless.

Topology by stage

Ranking Island against Assembly Line against Pod in the abstract would be false precision - they are stage-gated, not substitutes at a given size. The efficient topology is the smallest one the signals support; the menu above is how you move between them.

• Island (1-5 reps): every rep runs their own full cycle, right at the start. Past 10-15 reps it caps per-rep output and hides all stage-level conversion data.
• Assembly Line (Aaron Ross, Predictable Revenue; the standard for B2B SaaS between roughly $10M-$200M ARR): SDR → AE → AM/CS specialization by funnel stage. Ross's full model has four roles (inbound response, outbound SDR, AE, AM/CS) and most adopters "don't go far enough" - partial adoption is normal, not failure. Cost: handoffs leak, and a weak stage starves everything downstream.
• Pod (Jacco van der Kooij, Winning By Design; ~20+ reps): small cross-functional units (e.g. 3 SDR + 2 AE + 1 CS) each owning a shared book of customers. Customer-centric for complex enterprise deals. Costs: duplicated management overhead, and poor performers hide behind the team.
• Hybrid - where most mature orgs actually land: assembly line for the high-volume segment, a pod for enterprise, sometimes a founder-led island still running in parallel. Design per segment, not one shape for the whole org.

Worked transition cases - Island to Assembly Line, the hunter/farmer decision, standing up an enterprise pod - with the capacity math behind each: topology-transition-cases.md.

Brainstorm before you restructure

An org design hardens the moment leadership announces it - reversing a reorg costs more trust than any other planning artifact. Surface the assumptions first.

1. After the interview, present 2-3 candidate structures (drawn from the stages and moves above, adapted to the answers) with trade-offs and one explicit recommendation. Ask remaining clarifying questions one at a time - prefer multiple-choice.
2. Get explicit approval on the direction before detailing anything.
3. Build the design section by section, validating each with the user before the next. A wrong role mix invalidates every downstream section, so never present the design as one finished block.
   1. Role mix and ratios.
   2. Topology per segment.
   3. Management spans.
   4. Expansion ownership.
   5. Transition plan.
4. Gate finalization on user approval of the assembled design.

If your harness has persistent memory, store the approved decisions so the next planning cycle and any mid-cycle hiring question starts from the recorded design, not from scratch.

• Topology per segment.
• Derived ratio and its assumptions.
• Span targets.
• Expansion ownership.
• The written next-stage triggers.

Workflow

1. Place the org on the growth ladder. Treat the revenue bands as directional - the trigger is headcount and funnel behavior, never revenue alone.

   • Founder-led: the founder closes the first 10-20 unaffiliated customers, then hires two reps at once (Lemkin's A/B-test heuristic).
   • First leader: at $1M-$2M ARR once 2 reps hit quota.
   • 5-30 reps: first SDR split, first managers, first segmentation.
   • 30-100+ reps: RevOps, enablement, pods or hybrid, 3+ lifecycle roles.

2. Read the behavioral signals. Measure AE self-prospecting hours, selling-time share, calendar fullness, and stage-conversion visibility against the triggers in the move menu. If no trigger has fired, the correct design is the current one - say so and stop.

3. Check deal complexity per segment.

   • SMB (1-few stakeholders, days-weeks cycles) fits island or assembly line.
   • Enterprise (11-20+ stakeholders, 6-12+ month cycles) is what justifies a pod.

   Design each segment's structure separately when complexity diverges. The full segmentation model itself belongs to mbfinotti/sales-skills@sales-account-segmentation.

4. Derive the SDR:AE ratio from the org's own funnel math - never copy a published number.

   • Work backward: opportunities each AE needs per month, minus what inbound and marketing already source, divided by SQLs an SDR actually produces.
   • Sanity-check against segment bands: 1:1.5-1:2 for SMB/outbound, 1:2-1:2.5 for mixed SaaS, 1:3-1:4 for enterprise inbound. Bands only - published headline figures genuinely conflict.
   • Bridge Group's longitudinal series shows a real drift from 1 SDR per 3.9 AEs (2014) to 1:2.4 (2025).
   • Five factors distort any comparison between published figures: segment mix, inconsistent SDR/BDR labeling, hybrid self-prospecting AEs, uncounted offshore capacity, early AI SDR tooling.

   Full reconciliation and the derivation example: ratio-span-benchmarks.md and topology-transition-cases.md.

5. Set management spans.

   • Median 7 AEs per frontline manager (stable since 2015).
   • Median 6.4 SDRs per SDR leader (tighter at small companies, wider at scale).
   • General guidance: 6-10 direct reports.

   Widening span without stronger manager enablement reliably degrades coaching, slows ramp, and raises attrition. Flag any player-coach seat and its exit date.

6. Decide expansion ownership. Apply move 3's trigger and credited-tail rule. Note the attrition asymmetry in the stress test below when weighing it.

7. Stress-test against attrition and ramp reality.

   • SDRs: ~39-40% annual attrition (materially higher below $20M revenue), ~3-month ramp, ~1.9-year tenure.
   • AEs: ~30-32% attrition, ~6-month ramp, ~2.8-year tenure.

   Hunter-heavy units carry structurally higher attrition than farmer-heavy ones independent of management quality - a design that ignores backfill and ramp carry overstates its own capacity. Feed the resulting effective-capacity picture to mbfinotti/sales-skills@sales-quota-setting, which turns it into ramp-adjusted quota math.

8. Plan the transition.

   • Order the moves by the menu's efficiency ranking.
   • Name each move's firing trigger.
   • Define every new handoff in writing (the SDR→AE handoff is the highest-loss point an assembly line creates).
   • List the comp and quota dependencies to hand to the sibling skills.

   Write the next-stage triggers into the plan when the mandate is compounding.

9. Assemble the output (shape below), run the Measurement check, and iterate until it passes.

B2B vs B2C

The trigger logic transfers; the benchmark data and several structures do not.

Differs:

• The published rep-level benchmarks are B2B SaaS data (Bridge Group, Alexander Group, Emergence Capital), with no B2C equivalent - calibrate a B2C design against the org's own historicals only, and say so in the plan.
• Commission-agent verticals (insurance, real estate, solar, auto) run the island model by default and often permanently. Each agent is a full-cycle business; the growth ladder does not apply. The design question there is agency leverage - teams under a lead agent, splits, who services the book - not funnel specialization.
• Retail sales structure is driven by location and shift coverage, not funnel stages: spans are set by floor coverage, and the "topology" is the store hierarchy.
• Call-center/telesales is the assembly line taken to its extreme - queue-fed, script-standardized, QA-scored. Spans run materially wider than the B2B coaching span because the work is monitored rather than coached; do not import the 6-10 B2B span there, or the wide call-center span back into B2B.
• The SDR:AE ratio concept has a B2C analog - appointment-setter to closer in in-home sales (solar, home improvement). The funnel-math derivation transfers; the B2B benchmark bands do not.

Org design output shape

ORG DESIGN: stage (revenue band · rep count) · motion · segments served
SIGNALS READ: selling-time share · self-prospecting hours · calendar fullness · which triggers fired
ROLE MIX: current vs proposed headcount per role · derived SDR:AE ratio with its assumptions · sanity band
TOPOLOGY: model per segment · why the signals support it · what was deliberately left unchanged
SPANS: management layer · reps per manager per role · player-coach seats flagged with exit dates
EXPANSION OWNERSHIP: hunter/farmer decision · credited-tail rule if split · deleted-this-cycle note if struck
TRANSITION PLAN: moves in efficiency order, each with its firing trigger · handoff definitions · dependencies handed to sibling skills
NEXT TRIGGERS: the written signals that fire the next restructure

Failure modes

• Specializing too early - 2 SDRs supporting 4 AEs below ~10 reps is "just overhead and handoff friction." Fix: wait for the trigger, not the org-chart aesthetic.
• Specializing too late - an island model past 10-15 reps caps per-rep output and hides the stage-conversion data every later decision needs.
• Copying a published SDR:AE ratio - the headline figures conflict for definitional reasons, not because one source is wrong. Derive from your own funnel math; use bands only as a sanity check.
• SDRs below the ACV floor - under roughly $4K ACV the deal value cannot fund a dedicated prospecting role, whatever the AE workload says.
• Permanent player-coach leadership - the personal quota crowds out team-building and coaching collapses toward 14-16% of manager time.
• Splitting hunter/farmer too early - below ~25 AEs, quota-credit rules distort rep behavior more than the split helps.
• Pods as a default upgrade - duplicated management overhead, poor performers shielded by the team, individual contribution harder to reward. Pods answer a specific accountability failure, not seniority.
• Restructuring on the calendar - a reorg timed to a fundraise or a fiscal year, with no fired trigger, spends reversibility for nothing.
• Treating the AI headcount narrative as settled - the evidence is genuinely mixed (steep SDR cuts at traditional companies, doubled SDR hiring at AI-native ones). Pilot before cutting or freezing a role class; see ratio-span-benchmarks.md for what would confirm or overturn it.
• Copying an outlier - Dell's 2025 minimum-15-report spans are deliberate delayering under an AI mandate, a counter-example to study, not a benchmark to import.

Measurement

The design is not done until all of these pass; iterate until 100%:

• Every proposed move names the behavioral trigger that fired it; any move without a fired trigger is removed or explicitly future-dated.
• The SDR:AE ratio is derived from the org's own funnel math, with the benchmark band used only as a sanity check - and the ratio's assumptions (inbound share, SQL productivity, role definitions) are written down.
• Spans sit inside the 6-10 band per role, or the exception is argued; every player-coach seat has an exit date.
• Expansion ownership is decided, with the credited-tail rule stated if split - or the split is named as deleted this cycle and why.
• Every new handoff has a written definition before the transition starts.
• Contested figures (the ratio, AI headcount effects) are presented as contested, never as one settled number.

Outcome KPIs to track after the transition:

• AE selling-time share and self-prospecting hours vs the trigger thresholds - the signals that justified the design should visibly recover.
• SDR→AE handoff acceptance rate, and stage-conversion visibility (can the org now see where the funnel leaks?).
• Attrition and ramp time per role vs the benchmark bands; manager coaching hours vs span.
• Re-design triggers: a trigger from the move menu firing again, a new segment's complexity diverging, or attrition in one role class breaking sharply from its band.

References

• See mbfinotti/sales-skills@sales-hiring to recruit people into the seats this structure defines, and for the 30-60-90 ramp plans behind the ramp figures.
• See mbfinotti/sales-skills@sales-motion for the motion decision that precedes topology - a PLG or self-serve motion changes whether SDR seats exist at all.
• See mbfinotti/sales-skills@sales-quota-setting to turn this structure's effective capacity into ramp-adjusted quotas.
• See mbfinotti/sales-skills@sales-comp-design for the pay mix, credited tails, and accelerators the hunter/farmer split depends on.
• See mbfinotti/sales-skills@sales-account-segmentation for the segment model that decides which segments this structure serves.
• See mbfinotti/sales-skills@sales-market-sizing for the TAM figure the TAM-to-headcount math consumes.
• See ./references/topology-transition-cases.md for worked transition cases and the ratio-derivation and TAM-to-headcount math.
• See ./references/ratio-span-benchmarks.md for the contested SDR:AE reconciliation, span and attrition tables, and the 2025-2026 AI headcount evidence.