Your sales org structure determines how information flows, how deals get worked, and how fast you can scale. Pick the wrong model and you'll fight the structure every quarter. Pick the right one and the structure does half the management work for you.

There are three dominant models: the island, the assembly line, and the pod. Most companies end up using a hybrid, but understanding the pure models helps you design the hybrid that fits. I've operated in all three across companies ranging from $2M ARR startups to $500M+ enterprises. Each works. Each breaks. The difference is context.

The Island Model

In the island model, each rep owns the entire sales cycle. They prospect, they demo, they negotiate, they close. No SDR handoff. No SE support (or very limited). No account management transition. One person, start to finish.

When islands work

  • Early stage (under 10 reps): You don't have the volume to specialize. An SDR team of 2 supporting 4 AEs is just overhead and handoff friction.
  • Relationship-heavy sales: Industries where the buyer wants one point of contact throughout. Wealth management, professional services, consulting sales.
  • Founder-led transition: Your first 2-5 sales hires should be islands. They need to do everything the founder was doing, because you're testing whether the process can transfer.
  • High ACV with low volume: If each rep works 20-30 active opportunities per quarter, there isn't enough activity to justify splitting the funnel.

When islands break

Above 10-15 reps, the island model creates two problems. First, prospecting suffers because closing is more rewarding. A rep with 3 deals in negotiation and a prospecting target won't prospect. Second, you can't build a repeatable process because every island develops their own. Rep A qualifies differently than Rep B. When Rep A leaves, their process leaves with them.

The telltale sign: pipeline creation drops as the team grows. You're adding reps but total pipeline doesn't scale proportionally. That's the island model hitting its ceiling.

The Assembly Line Model

The assembly line separates the sales process into specialized stages. SDRs generate qualified meetings. AEs run discovery through close. CSMs handle onboarding and expansion. Each person does one thing well.

This is the dominant model at scale. Aaron Ross documented it in Predictable Revenue, and most B2B SaaS companies between $10M and $200M ARR run some version of it. There's a reason: it works for high-velocity, repeatable sales motions where efficiency matters more than relationship continuity.

Standard assembly line roles

Role Owns KPIs Typical Ratio
SDR (Outbound) Prospecting, qualifying Meetings booked, pipeline created 2-3 SDRs per AE
SDR (Inbound) Lead qualification, routing Speed to lead, qualification rate 1 per 200-400 inbound leads/mo
AE (SMB) Discovery through close Bookings, win rate, cycle time $600K-$1M quota
AE (Enterprise) Discovery through close Bookings, deal size, multi-year % $1M-$2M quota
SE Technical validation POC win rate, time to value 1 SE per 2-3 AEs
CSM Onboarding, retention, expansion NRR, churn rate, expansion % 1 per $1-2M ARR managed

When assembly lines work

  • High deal volume: 50+ new opportunities per month across the team
  • Predictable sales cycle: consistent stages that map to repeatable activities
  • Sufficient headcount: 15+ total sellers (below that, the handoff overhead outweighs the specialization gains)
  • Measurable pipeline: you can track conversion rates between stages because the stages are discrete

When assembly lines break

Handoffs. Every handoff between roles is a potential failure point. The SDR-to-AE handoff is the most studied and the most broken. A 2024 Pavilion study found that only 62% of SDR-sourced meetings convert to qualified opportunities. That means 38% of the meetings your SDR team books are wasted AE time. Most of that waste comes from misaligned qualification criteria, poor handoff notes, and SDRs optimizing for meeting count rather than meeting quality.

The second failure point: the AE-to-CSM handoff. The customer builds trust with the AE during a 3-month sales cycle, then gets handed to a stranger on day one of being a customer. Companies that don't invest in formal handoff processes (documented account context, introductions during the sales cycle, shared onboarding calls) see 20-30% higher churn in the first 6 months.

The Pod Model

The pod model groups cross-functional roles into small, autonomous teams. A typical pod includes 2 SDRs, 3-4 AEs, 1 SE, and 1 CSM, all working the same accounts or territory. The pod operates like a mini-business within the larger org.

Pod structure in practice

At a company with 40 sellers, you might have 5 pods of 8 people each. Each pod owns a geographic territory, a vertical, or a company-size segment. The pod has its own pipeline, its own quota, and its own internal workflows. Handoffs happen within the pod, so the SDR knows the AE personally, the AE knows the CSM personally, and context doesn't get lost in a CRM transfer.

When pods work

  • Mid-market to enterprise selling: where deal complexity requires collaboration, not just handoffs
  • Vertical selling: pods organized by industry develop deep expertise that islands and assembly lines can't match
  • Companies with 20-100 sellers: big enough for specialization, small enough for pods to feel like teams
  • High-churn environments: keeping the CSM in the same pod as the AE who sold the deal creates accountability

When pods break

Pods create internal competition that can become toxic. Pod A lands a whale account that pushes them to 130% attainment while Pod B struggles in a weaker territory. If comp is pod-based, B's top performers start lobbying to switch pods. If comp is individual, the pod structure doesn't create the collaboration you designed it for.

Pods also make hiring harder. You can't just hire an AE; you need to hire an AE for a specific pod with specific territory knowledge. And if a pod loses its best AE, the entire pod's performance drops because they were load-bearing.

Span of Control: The Numbers

Span of control is the number of direct reports per manager. Get it wrong and you're either paying for managers who don't add value (too narrow) or burning out managers who can't coach (too wide).

Manager Role Ideal Span Maximum Before Quality Drops
SDR Manager 8-10 12
SMB AE Manager 7-8 10
Mid-Market AE Manager 6-8 9
Enterprise AE Manager 4-6 7
Director (manages managers) 4-6 8
VP Sales 5-7 direct (mix of directors + IC leaders) 9

SDR managers can handle wider spans because SDR work is more repetitive and coachable in group settings. Enterprise AE managers need narrower spans because each deal requires individual coaching, multi-threading strategy, and executive engagement. A manager running deal strategy for 8 enterprise accounts is already spread thin. At 12, they're not coaching, they're just tracking.

Layers vs. Flat: When to Add Management

Every management layer adds cost and communication overhead. It also adds coaching capacity and career paths. The tradeoff depends on your stage.

Under 15 sellers: stay flat

VP Sales manages everyone directly, maybe with 1-2 player-coaches who handle some coaching duties. Adding a Director of Sales when you have 10 reps creates a layer that slows decisions without adding enough coaching capacity to justify the $180K+ salary.

15-30 sellers: add frontline managers

This is where you need team leads or frontline managers. 2-3 managers, each with 6-8 reps, reporting to the VP Sales. The VP shifts from deal coaching to manager coaching, pipeline review, and strategy.

30-75 sellers: add a director layer

Directors manage managers. Each director owns a segment (SMB, Mid-Market, Enterprise) or a territory. The VP Sales now manages 3-5 directors plus potentially RevOps and enablement leads. This is where the VP role becomes a true executive position, removed from individual deals by two layers.

75+ sellers: add a CRO or SVP

At this scale, you need someone above the VPs. A CRO who manages VP Sales, VP CS, VP Partnerships, and RevOps. The org chart has 4-5 layers from CRO to SDR. Communication becomes a real challenge. The quarterly all-hands and weekly leadership sync become critical infrastructure, not nice-to-haves.

Territory Design: The Hidden Structure Decision

Territory design matters as much as org model. A perfect pod structure with bad territories still fails. Territories need to balance opportunity, effort, and fairness.

Common territory models

  • Geographic: works when physical proximity matters (field sales, local market knowledge). Increasingly less relevant in remote-first selling.
  • Named accounts: works for enterprise, where relationship depth matters more than volume. Assign 30-50 target accounts per rep.
  • Vertical: works when industry expertise drives win rates. Healthcare, financial services, and government are the verticals where specialization pays off most.
  • Revenue-based: SMB vs. mid-market vs. enterprise, segmented by company size or deal value. The most common model and the easiest to balance.

The biggest territory mistake: over-rotating on potential rather than accessible pipeline. A territory with 500 accounts and no existing pipeline is worse than a territory with 200 accounts and 20 existing conversations. New territory assignments need pipeline seeding (3-6 months of SDR or marketing effort) before quota kicks in at full weight.

Choosing Your Model: A Decision Framework

Use these questions to narrow your choice:

  1. Fewer than 10 sellers? Start with islands. Don't add complexity you can't use.
  2. Deal cycle under 60 days with 50+ deals/month? Assembly line. Volume justifies specialization.
  3. Deal cycle over 90 days with high ACV ($100K+)? Pods or a hybrid. Complex deals need collaboration, not handoffs.
  4. Selling into regulated verticals? Vertical pods. Industry expertise compounds over time.
  5. High churn, expansion-driven revenue? Pods with integrated CSMs. Keep the post-sale team close to the pre-sale context.

Most companies between $20M and $100M ARR end up with a hybrid: assembly line for SMB (high volume, low touch) and pods for mid-market and enterprise (high touch, longer cycle). The assembly line feeds volume, the pods close complexity. Two models, one org, unified under a VP Sales who understands when each approach applies.

Scaling Traps to Avoid

Trap 1: Promoting your best rep to manager

Top sellers and top managers require different skills. The best AE on your team might be a terrible coach. Promote based on coaching instinct, team building, and process thinking, not quota attainment. If you must promote a top seller, give them a 3-month trial with a clear path back to IC if it doesn't work.

Trap 2: Restructuring every year

Org changes reset relationships, territories, and pipeline. One restructuring per 18-24 months is acceptable. More than that signals that leadership is guessing rather than designing. Every restructuring costs you a quarter of productivity while the team adjusts.

Trap 3: Copying another company's structure

Salesforce's org structure works for Salesforce. It won't work for your 40-person startup selling to mid-market. Match your structure to your stage, your market, and your deal complexity. Not to what you read in a case study about a company 100x your size.