SMB SaaS Pricing: Examples & Companies

What is it

SMB SaaS Pricing is pricing plans aimed at small and medium businesses — typically self-serve, with team features and modest per-user fees.

It is by far the broadest segment in the corpus: 263 of the in-corpus companies address SMBs. The defining trait is the buyer. An SMB purchase is made by a team lead with a small budget who pays by card and onboards without procurement, so the plan is published, self-serve, per-seat, and lightly administered — shared billing and basic roles, but not the SSO/SCIM/audit machinery an enterprise demands.

On most pricing pages, “SMB pricing” is literally the middle tier — the “Team” or “Business” plan that sits between the individual plan and the enterprise “contact sales” tier. It clusters in a tight $15–$20/seat band, all published and card-checkout, and spans every product category in the corpus — from creative video tools like Synthesia and HeyGen to workspace tools like Notion AI to the developer APIs small engineering teams adopt without ever talking to a salesperson.

How it works

SMB pricing is defined less by a formula than by a set of levers a vendor tunes to keep the plan self-serve and low-friction. The mechanics fall into four dimensions:

The most common SMB structure is a flat per-seat fee. The total bill is straightforward:

Unit math: Team bill = seats × seat_price. A five-person team on Linear Business pays 5 × $16 = $80/month; the same team on Canva Business pays 5 × $20 = $100/month.

A second structure, increasingly common as AI features get metered, is seat-plus-usage — a per-seat base with an AI meter layered on top. Notion AI bundles AI into its $20 Business seat but meters Custom Agents at $10 per 1,000 Notion credits; Gamma bundles a monthly credit pool into each seat and tops up at 1,500 credits for $6. Descript runs the same shape at the SMB end — its $50/person Business seat bundles a fixed allowance of media hours and AI credits, with paid overage.

The team-tier markup — roughly 2× the individual seat — rarely buys a materially different product. It funds shared admin and reflects the budget shift from a personal card to a team expense. Cursor Teams at $40 versus Pro at $20 is the canonical example; Codeium’s Windsurf Teams at $35/user versus Pro at $15 is another.

Companies using this

263 companies in the current corpus offer SMB-oriented plans — nearly the entire corpus, spanning every product category. Representative published team tiers include Linear ($16/user), Canva ($20/user), GitHub Copilot ($19/user), Cursor ($40/user), Notion AI ($20/seat), and creative tools like Synthesia, HeyGen, and Descript — alongside the developer APIs SMBs adopt self-serve, from Groq to Together AI and Replicate. The table below lists each, with pricing model, billing units, and last-verified date.

Patterns observed

Across the 263 SMB companies, a handful of pricing patterns recur so consistently they read as segment conventions:

• SMB is the published middle tier. For most companies, “SMB pricing” is literally the Team or Business plan on the page — self-serve, per-seat, monthly. Linear Business ($16), Canva Business ($20), and GitHub Copilot Business ($19) all sit in a tight $15-$20 band, one step below their custom-quoted enterprise tier.

• The team price runs ~2× the individual seat. The markup funds shared admin and reflects the shift from a personal card to a team budget rather than a real product difference — the feature delta is admin, not capability. It holds across categories: developer tools, workspace apps, and creative suites all price the team seat at roughly double the solo plan.

• Light admin, not compliance. SMB plans add roles, shared billing, and centralized billing through Stripe, but stop short of SSO/SCIM/audit — those gate the enterprise tier. Gamma draws this line explicitly: its Team plan ($240/seat annual) has centralized billing while SSO and SOC 2 documentation live only on Business and above.

• AI features are increasingly metered on top of the seat. As raw AI cost bit into margins, vendors moved from bundled AI to the seat-plus-usage shape described above. Metering the AI keeps the headline SMB seat price low while heavy users pay for what they consume, so the sticker stays competitive even as compute costs rise.

• SMB is where PLG expansion happens. A single PLG user converting a few colleagues onto a Team plan is the canonical expansion motion. Nearly every self-serve tool in the corpus — Cursor, Linear, Notion AI — is designed for exactly this bottom-up spread.

Counterexamples & variants

The clean counterexample is the enterprise-first vendor with no real self-serve SMB tier. Harvey — the legal-AI vendor — carries only midmarket and enterprise segments and sells exclusively top-down through sales-led quotes with no public rate card. Its product and price assume an institution, not a small team, so the SMB band simply isn’t served.

A subtler variant is the nominal-but-gated SMB tier. Glean lists an SMB segment, but its motion is sales-led: there is no published price, seats are quoted per-user via Enterprise Flex, and a small team can’t self-serve. On paper it addresses SMBs; in practice a five-person shop cannot buy it the way it buys Linear or Canva. This is a common trap — a vendor claims the segment without building the self-serve motion that actually reaches it.

The most important genuine variant is usage-based SMB. Developer-tool SMBs frequently bypass per-seat plans entirely in favour of a usage-priced API, because a small engineering team’s true cost tracks volume, not headcount. Groq meters tokens (Llama 3.3 70B at $0.59/$0.79 per million in/out); Together AI meters tokens and dedicated GPU time; Replicate bills per compute-second and per image (FLUX 1.1 Pro at $0.04/image). For these teams, pure-usage pricing outcompetes per-seat at low seat counts — a two-engineer team may spend $30/month on tokens where a per-seat plan would charge for licenses it barely uses.

Finally, the SMB tier is where repricing backlash is loudest. Canva’s September 2024 conversion of Canva for Teams from a flat “$300/year for the first 5 people” bundle to ~$100/user/year with a 3-seat minimum tripled the bill for small teams and triggered one of the year’s loudest SaaS pricing backlashes — a reminder that SMB buyers are price-sensitive and vocal in a way enterprise buyers, cushioned by custom contracts, are not.

What this means for buyers vs vendors

For buyers

The Team or Business tier is the SMB sweet spot — self-serve, published, and quick to adopt — but model the per-seat cost across your whole team before you commit, and watch for the jump to enterprise the moment you need SSO or a security review. If your workload is bursty or engineering-heavy, price out a usage-priced API (Groq, Replicate) against per-seat plans: at low seat counts, usage often wins. And read the fine print on metered AI features — a $20 seat with a Custom-Agent meter can bill well past its headline. Our choosing the right usage metric guide helps you decide which unit actually reflects your cost, and the Cursor pricing calculator and Vercel pricing calculator let you model a real team bill before you buy.

For vendors

Make the Team tier self-serve and published — SMBs won’t wait for a quote, and every extra step of friction leaks conversions. Price it roughly 1.5-2× the individual seat to fund shared admin, keep the admin light, and reserve SSO, SCIM, and compliance for the enterprise tier so you keep a reason to escalate. If your AI costs are volatile, layer a usage meter on top of the seat rather than raising the headline price, the way Notion AI and Descript do — but be transparent about it, because SMB buyers are price-sensitive and repricing them badly (see Canva) invites a public backlash. Our usage invoicing & billing cycles guide covers the billing mechanics of running a self-serve seat-plus-usage plan at scale.

Explore this theme in the knowledge graph →