We’ve written before that SaaS pricing should be simple — if a firm can’t work out what you cost in ten seconds, you’re already losing.
That post was about clarity. This one is about the harder question underneath it: what should you actually charge for?
Because in accounting software, the unit you pick is not a pricing detail. It decides who inside the firm champions you, how you grow inside an account, and whether success for your customer is also success for you.
The four models
Per-seat
Charge per user with access.
Where it works: products used by a defined group of staff, where usage correlates with headcount. Practice management, workflow, document tools.
The accounting-specific problem: firms are capacity-constrained and seasonal. They bring in staff for busy season, they use offshore teams, and partners want everyone to have visibility. Per-seat pricing makes every one of those a cost decision.
You end up with password sharing, artificially low seat counts, and — worst of all — a customer whose adoption is deliberately limited. A product only three of twelve people can open never becomes indispensable, which means it never becomes a control point.
If you go per-seat, price so that adding a user is a trivial decision, and consider free read-only seats. You want more people in the product, not fewer.
Per-client
Charge based on how many of the firm’s clients are managed in your product.
Where it works: anything that operates on the firm’s client base — compliance, onboarding, reporting, client comms.
Why it’s compelling: it scales with the firm’s own revenue. When they grow, you grow. Aligned, defensible, and easy for a partner to understand because it maps onto how they already think.
The problem: sticker shock at signup. A firm with 900 clients does the multiplication before it understands the value and stops there. It also invites gaming — firms will run a subset through your product to keep the number down, which caps your adoption.
Fix it with banding rather than pure linear pricing (1–50, 51–200, 201–500), and make sure the entry band is genuinely reachable for a small practice.
Per-entity
Charge per legal entity, group or file.
Where it works: multi-entity consolidation, group reporting, anything where the entity is genuinely the unit of work.
Why it’s good: it maps precisely to complexity. Complex clients cost the firm more and are worth more to them, so they’re happy to pay more.
The caveat: it only makes sense for a narrow set of products. Applied to anything else it feels arbitrary, and arbitrary pricing erodes trust with an audience that reconciles things for a living.
Flat / tiered
One price per firm, or a small number of tiers with feature or volume limits.
Where it works: early-stage products, self-serve motions, and any market where the buyer is a sole practitioner making a fast decision.
Why it’s underrated: it removes friction entirely. No maths, no forecasting, no fear of the bill growing. For a risk-averse buyer that’s worth real money.
The cost: you leave money on the table with large firms, and you have no natural expansion motion. Expansion has to come from tier upgrades or new products, which you have to engineer deliberately.
The question that decides it
Forget benchmarks. Ask this:
When our customer gets more value, what number goes up?
Charge for that number.
- If value scales with how many people use it → seats
- If value scales with how many clients it touches → clients
- If value scales with structural complexity → entities
- If value doesn’t really scale → flat, and stop pretending otherwise
Pricing on a unit that doesn’t track value is where the pain comes from. Charge per seat for something whose value scales with client count and you’ll watch a firm triple their client base on the same three logins while their value from you 3x’s and your revenue doesn’t move.
The accounting-specific traps
Seasonality. Firms flex headcount around busy season. Per-seat pricing on annual contracts either overcharges them for ten months or forces awkward mid-term changes. Some tolerance for flexing is a genuine competitive advantage.
Offshore and outsourced teams. Many firms run offshore delivery. If your per-seat pricing makes that expensive, you’ve penalised a large and growing segment.
The partner who wants visibility, not usage. Partners want to look at dashboards occasionally. Charging a full seat for that is how you end up with the economic buyer never logging in — and a buyer who doesn’t experience the product is a buyer who cancels it easily.
Client-count sensitivity. Firms are cagey about client numbers. A model that requires them to disclose it at the first conversation adds friction exactly where you can least afford it.
Reselling. If you want firms to resell or bundle you, your model has to leave room for their margin and be explainable to their client. Per-seat is hard to bundle. Per-client is much easier.
Hybrids, carefully
Most mature accounting SaaS ends up hybrid: a platform fee plus a usage dimension. Reasonable, and it solves the sticker-shock problem — small base, scaling component.
But every dimension you add costs comprehension, and comprehension is the thing you can least afford to spend. Two dimensions is manageable. Three is a spreadsheet. If your own team can’t quote a price on a call without opening a calculator, no prospect will ever hold it in their head.
Changing your model
Sometimes you get it wrong and have to move. It’s survivable if you’re straight about it:
- Grandfather generously. Existing customers on the old model, indefinitely if you can afford it. The goodwill is worth more than the recovered revenue, especially in a market this connected — accountants talk, and a pricing change that feels like a stitch-up will be discussed at every conference for a year.
- Tell them early and directly. No burying it in a T&Cs update.
- Make the new model obviously better for someone. If everyone pays more, it’s a price rise, so call it a price rise. Don’t dress it as a restructure.
- Move partners last. They’ve built commercials on your pricing. Blowing that up without warning damages the channel far beyond the affected accounts.
A working starting point
If you’re setting pricing for the first time:
- Identify the value unit. What goes up when they win?
- Three tiers, not seven. Small, standard, and a custom option that lets you have a conversation with larger firms.
- Put your target average revenue at tier one or two, then expand from there.
- Make tier one genuinely usable. A crippled entry tier just teaches firms the product doesn’t work.
- Publish it. Firms will disqualify you for hiding pricing far more often than for being expensive.
- Sanity check it against churn. If firms leave at renewal citing cost, the problem is usually that value never landed — which is a time to value and onboarding problem wearing a pricing costume.
Get the unit right and pricing conversations get easier every year, because your revenue grows for the same reason your customer’s does. Get it wrong and you’ll spend the rest of the company’s life negotiating.
Pricing sits inside positioning, which is where our go-to-market strategy work starts. Related reading: keep it simple, idiot, control points, and why firms churn in the first 90 days.
