Choosing a SaaS sales model by ACV

The sales model argument is usually held as a strategy debate and settled by a founder’s temperament. It should be settled by arithmetic. Average contract value sets the ceiling on how much human attention each deal can absorb, and every model below has a floor where it stops paying for itself.

Here is the whole comparison first, then the reasoning behind each threshold.

The four models compared

Model	ACV floor	Fully loaded cost	Quota	Ramp	Cycle	CAC payback
Self serve	$0	$0 per deal, product and marketing carry it	n/a	n/a	Hours to 14 days	3 to 9 months
Inside sales	$2,000	$110K to $150K per rep	$450K to $700K	2 to 3 months	14 to 30 days	9 to 15 months
AE plus SDR	$15,000	$270K to $370K per pair	$900K to $1.4M	4 to 6 months	45 to 90 days	15 to 24 months
Field and enterprise	$75,000	$280K to $350K per AE plus SE	$1.2M to $2M	6 to 9 months	90 to 270 days	24 to 36 months

Read the ACV floor column as a hard gate. Everything else is adjustable.

The unit economics of a single rep

Start with one number and work outward. A mid market AE on $130,000 on target earnings costs $180,000 to $250,000 fully loaded once you add roughly 28 percent in payroll tax and benefits, a share of the sales manager, and the tool stack: Salesforce or HubSpot seat, Outreach or Salesloft at $100 to $140 a month, Gong at roughly $1,400 a year, plus enablement and travel.

That rep needs to carry four to five times their cost in quota. Call it $900,000. Now divide by ACV.

The $600 row is the one that matters. If you are selling at $600 ACV and you have quota carrying reps, you do not have a sales execution problem. You have a pricing problem, and no amount of coaching, sequencing or CRM hygiene will fix it. Raise the price, change the value metric, or go fully self serve. Our self serve versus sales assisted comparison covers the crossover in more detail.

Where self serve actually pays for itself

Self serve is not free. The cost moves from salary to product engineering, onboarding content, lifecycle email and support. A credible self serve motion needs a genuinely usable trial, in product guidance, billing that handles card failures, and a documentation site that answers questions at 2am.

Budget that at roughly two engineers and one lifecycle marketer, which is $400,000 to $600,000 a year. The difference is that cost is fixed while a sales team’s cost is linear in revenue. That is the whole argument for staying low touch longer than your board wants you to.

Atlassian is the canonical case. It reached hundreds of millions in revenue with no traditional outbound sales force, selling through the website and a partner channel. Calendly did something similar at smaller scale, running for years with essentially no AEs. Both are usually described as sales achievements. They were pricing and product achievements. The sales model was a consequence.

For the mechanics of running a self serve motion and knowing when a human should intervene, self serve sales and when to add humans is the companion piece.

When inside sales starts to make sense

The trigger is not ACV alone. It is ACV above roughly $2,000 combined with evidence that a conversation changes the outcome. Look at your trial to paid rate for accounts that got a human touch versus those that did not. If the gap is under five percentage points, a rep is not adding value and you are buying a comfort blanket.

An inside sales rep at $110,000 to $150,000 fully loaded carrying $450,000 to $700,000 works when deal cycles are 14 to 30 days and the rep can hold 40 to 60 live opportunities. Above 60, quality collapses. That is the practical constraint nobody plans for.

Mid market inside sales teams are where most SaaS companies spend the majority of their lives. The operating detail is in inside sales for mid market SaaS.

Adding SDRs, and when not to

An SDR costs $90,000 to $120,000 fully loaded and should book 10 to 15 qualified meetings a month against a decent list. That is the honest range. Anyone quoting 25 is either counting meetings booked rather than held, or working a list built by an enrichment tool with no qualification.

Add SDRs when two conditions hold together: AEs are spending more than 40 percent of their time prospecting, and pipeline coverage has dropped below 3x quota. If AEs have enough inbound to fill their week, an SDR adds a handoff and a compensation dispute.

The failure mode is worth naming. SDR teams built before the ICP is tight produce meetings that AEs disqualify at 40 to 60 percent, and the SDR gets paid for meetings the AE then wastes an hour on. Tighten the account list before you hire the second SDR.

Field and enterprise: what changes above $75K ACV

Everything slows down and more people join the deal. Above $75,000 ACV you are selling to a committee of five to eleven people, passing a security review, and going through procurement and legal. The SaaS sales cycle benchmarks show enterprise cycles regularly running 180 to 270 days, and that is measured from first meeting, not from first touch.

The staffing changes too. A field AE needs a solutions engineer at roughly 1:2 or 1:3 coverage, which adds $150,000 to $200,000 per SE. Travel, executive sponsorship and proof of concept environments add real cost per deal. Snowflake and Databricks both built field first organisations from the start, because the product required an architecture conversation with a data team and no amount of self serve onboarding was going to replace that.

The contrast with Atlassian is instructive. Same industry, opposite motions, both correct for their product. The difference is whether the buying decision involves a system of record. If it does, humans are in the loop whether you fund them or not. Inside sales versus field sales sets out the cost and coverage differences side by side.

Hybrid motions and how they usually fail

Nearly every company above $20M ARR runs two motions. The theory is clean: small accounts self serve, large accounts get a rep. The practice breaks on three things.

Routing. If the same account can appear in a rep’s pipeline and simultaneously buy on a credit card, you get channel conflict inside a fortnight. Set one threshold, usually seat count or plan tier, publish it internally, and enforce it in the CRM rather than in a Slack channel.

Compensation. Reps discover that self serve accounts they nurtured convert without them and the comp plan does not pay. You either pay an override on accounts in their territory or you watch them stop helping. Pay the override. It is cheaper than the behaviour you get otherwise.

Reporting. Blended CAC across two motions hides the truth. Report them separately or you will keep funding the worse one. This is the single most common reporting mistake we see in combined sales and marketing strategies.

Migration order when you move up market

Moving up market is a sequence, and doing it out of order is how companies end up with expensive reps selling a cheap product.

The SaaS startup sales strategy playbook covers the first three steps for companies doing this from zero rather than migrating.

What we would pick

At under $2,000 ACV: self serve, no debate, and spend the sales budget on onboarding and lifecycle instead. At $2,000 to $15,000: inside sales with heavy inbound, no SDR team until year two. At $15,000 to $75,000: one AE per $900,000 of target, SDRs added only against a coverage gap. Above $75,000: field, with an SE from the first deal, and a plan for a nine month cycle.

The people who should pick differently are companies selling into a market with an entrenched incumbent and a long replacement cycle. There, field sales at a lower ACV than the model suggests can be defensible, because you are buying a land position rather than a payback period. Be honest that you are doing that, and say so on the board slide.

What to do next

Take your current ACV, multiply your fully loaded rep cost by 4.5, and divide. If the deals per month number is above about eight, your model is wrong for your price. Fix the price, not the team. The vocabulary around timing sits in sales cycle length, and the broader set of motions is mapped in the SaaS sales strategies hub.