Costlens
← Blog

How much SaaS should a 20-person startup spend?

CostLens · Apr 17, 2026 · 5 min read

Founders ask this question at board meetings. Finance leads ask it when they pull together the first proper SaaS spreadsheet. Nobody has a confident answer because the information is scattered across vendor reports that each cite slightly different numbers.

Here's what the research says, with sources. No hand-waving.

The headline: about $8,000 per employee per year

For companies with up to 20 employees, software spend lands at roughly $8,000 per full-time employee per year. That's from Cledara's 2025 Software Spend Report, based on the transactional data of the companies on their platform.

For a 20-person startup, that puts the annual SaaS budget around $160,000 — or $13,000/month.

Most founders guess lower when asked. The number feels wrong because it's spread across 40+ individual charges, most under $500/month, each defensible on its own. The full bill only becomes visible when someone pulls the list.

Spend per head drops sharply as the team grows

The same Cledara report shows an interesting pattern:

  • 0–20 employees: ~$8,000 per employee
  • 50–100 employees: ~$2,583 per employee
  • 100–200 employees: ~$1,741 per employee

That's a 4.6× drop in per-head software spend between a 20-person team and a 150-person team.

The explanation isn't that bigger teams use less software. It's that bigger teams have procurement functions, volume discounts, and enough scale to consolidate overlapping tools. A 20-person startup pays retail, one seat at a time, often on per-seat pricing that was designed for larger buyers.

Nothing about this is unfair. It's the shape of the market. But it means small teams can't out-optimize their spend with pure negotiation — the structural disadvantage is real. The single biggest win is cutting overlap and unused licenses, not squeezing each vendor.

You probably have more tools than you think

The Zylo 2025 SaaS Management Index puts the average for small companies (1–500 employees) at 152 applications. BetterCloud's 2024 State of SaaSOps lands at 106 apps across all company sizes (down from 112 the previous year — the first year-over-year decline in a decade).

For a 20-person team, 152 apps means roughly seven subscriptions per person. Some of those are team-level (one Notion workspace, one Slack, one Linear), but many are individual (an engineer's Linear seat, a designer's Figma plan, a marketer's Ahrefs).

Most teams consistently underestimate their tool count by 20–40% before they actually list everything. The gap is usually in these categories:

  • Domain registrations nobody thinks of as SaaS
  • Developer tools on individual engineer cards
  • Personal AI tools (ChatGPT Plus, Claude Pro, Cursor) paid on personal cards and expensed
  • Grandfathered accounts from past projects

About a third to half of your licenses are wasted

Two independent sources converge on the same pattern:

Small teams tend to do slightly better than this average — Productiv reports 70% adoption at SMBs, which means about 30% waste. But even the best case means nearly a third of your $160k is going to licenses nobody opened last month.

On that budget, 30% unused works out to roughly $48,000/year of pure waste. That's the biggest lever in the whole picture.

Funding events break the pattern

Cledara's post-funding spend research tracked what happens after a startup raises a round:

  • Software spend jumps 21% on average
  • Tool count grows 13% in the first six months

If you're planning a fundraise, expect your SaaS line item to spike right after. If you already raised recently and your budget feels heavier than it did six months ago, that's normal — and so is the accompanying ghost-sub accumulation that rarely gets audited until the next downturn.

What the $160k number means in practice

Three ways to read it:

  1. You're below $160k. Lean team, good cost discipline, or small footprint. Healthy baseline — focus on keeping it there as you scale, not on cutting further.
  2. You're around $160k. You're roughly average. The 30–40% waste figure applies: about $50k of that is probably recoverable through duplicate consolidation and unused-license cleanup.
  3. You're above $240k (+50% over benchmark). Audit immediately. At 20 people, this number is almost certainly driven by either overlapping tools across teams, grandfathered subs from the previous stage, or one specific category running out of control (usually infrastructure or AI tooling).

What the number doesn't tell you

Benchmarks are averages across companies with very different shapes. A 20-person fintech handling regulated data will legitimately spend more than a 20-person agency doing design work. "Healthy for your company" might be 30% above or below the benchmark.

What matters more than hitting a specific dollar figure is knowing what's in your SaaS stack and who owns each subscription. Teams with a current, owned ledger find the waste during their quarterly review. Teams without one find it during their next down round.

What to do with this

Three concrete actions:

  1. Pull a complete list of what you pay for. If you haven't done this in 90 days, your list is wrong. Start with the method in how to track SaaS renewals.
  2. Run the audit described in how to find unused SaaS subscriptions. Most teams recover $500–2,000/mo on the first pass.
  3. Check your number against the $8k/FTE benchmark. If you're more than 50% over, the gap is usually two or three specific tools, not death by a thousand cuts.

The benchmark is useful to know but not worth optimizing against. The actionable version of the question isn't "how much should a 20-person startup spend" — it's "what's in our stack, and who owns each piece of it." Get the second question right and the first one answers itself.

Sources