I raised prices on my own SaaS in 2021 and lost 4 customers in three weeks. I'd spent months building the case, ran the numbers twice, and still got the timing wrong. What I learned from that mess is the same thing I'd tell anyone asking how to price a SaaS product for growth: the number matters less than the mechanism around it. Get the mechanism right and you can adjust the number for years. Get it wrong and no clever spreadsheet saves you.
Key Takeaways
- Pricing is a growth lever, not a one-time launch decision. Treat it as a recurring process, not a project.
- The 5 C's (Cost, Customers, Competitors, Compatibility, Channel) form a fast first-pass filter before you touch a spreadsheet.
- Most SaaS companies leave real money on the table because they price on cost-plus logic instead of willingness to pay.
- Usage-based and hybrid models grow faster than flat seats, but they punish you if your billing infrastructure is weak.
- Grandfathering existing customers is a comfort move, not a strategy. It has a cost, and you should name it before you offer it.
- Measure LTV/CAC per pricing tier, not in aggregate. Blended metrics hide which tier is actually profitable.
How to price a SaaS product for growth (without guessing)
Here's the thing about SaaS pricing that nobody says out loud at conferences: most founders are pricing on vibes. They pick a number that feels defensible next to competitors, round it down, and move on. I did exactly that for eighteen months.
Then a mentor asked me one question that reshuffled everything: "Who is the customer you'd be happiest to lose?" I had no answer. That's when I realized I wasn't pricing for growth. I was pricing to avoid awkward sales calls.
Why pricing moves the needle more than most product work
Patrick Campbell, who founded ProfitWell before selling it to Paddle in 2022, spent years publishing data on this. His repeated finding: a 1% improvement in pricing typically produces a larger profit impact than a 1% improvement in acquisition or retention. That math isn't intuition. It's compounding — price flows straight to margin, while acquisition spend leaks through CAC.
Which is why I now treat pricing as a quarterly exercise, not an annual review. Every ninety days I ask three things: Are we converting the right customers? Are we losing the right customers? Is a tier quietly subsidizing another one?
The failure mode I see over and over
Companies price once, then never revisit. They grow headcount, infrastructure costs double, and the price stays frozen because "customers would be upset." Meanwhile every new feature ships into the same flat fee.
Result: revenue grows linearly, costs grow exponentially. That gap is where good products die.
What are the 5 C's of pricing?
The 5 C's of pricing are Cost, Customers, Competitors, Compatibility, and Channel. They're a classic framework used in pricing literature to force you through five lenses before committing to a number. None of them is sufficient alone. Together they catch the mistakes that a spreadsheet won't.
I've watched teams skip straight to "Competitors" — matching the nearest rival's price — and then wonder why they're stuck at a sub-$10k ACV. Here's how each C actually plays out in SaaS.
Cost
Not your historical cost. Your marginal cost to serve the next customer. If you're shipping an AI feature that hits OpenAI on every request, that matters enormously. I once priced a tier at $49/month and only realized after three months that heavy users were costing me $31/month in API calls alone. That tier existed to lose money.
Customers
What are they willing to pay, and — harder question — what are they comparing you against? A $200/month tool feels cheap next to a $4,000/month analyst salary. It feels expensive next to a $30/month Notion add-on. Positioning drives the ceiling, not features.
Competitors
Useful as an anchor, dangerous as a compass. If you copy a competitor's price without copying their cost structure, you inherit their margin problems. I've seen three-person teams price identically to ten-year incumbents and burn out inside a year.
Compatibility
Does this price fit how the customer already buys software? Enterprise procurement wants annual contracts and PO numbers. A solo dev wants a credit card form and instant access. Mismatch here kills conversions regardless of the number.
Channel
Where the sale happens changes the price. Self-serve implies a lower number with higher volume. Sales-led lets you charge more but demands a rep who can defend the value. Mixing both across the same product usually means one channel is subsidizing the other without anyone noticing.
SaaS pricing models that actually drive growth
There's no single best model. There's a best model for your product's natural unit of value. That unit might be seats, API calls, workflows, gigabytes, or something weirder like "completed invoices."
| Model | Best fit | Growth lever | Main risk |
|---|---|---|---|
| Flat-rate | Early-stage, simple product | Fast signups | Ceiling on expansion revenue |
| Per-seat | Collaboration tools | Viral within orgs | Discourages sharing, seat compression |
| Usage-based | APIs, infrastructure, AI | Aligns with customer success | Revenue spikes and dips, billing complexity |
| Tiered / feature-gated | Broad B2B SaaS | Upsell path | Features leak down to cheaper tiers |
| Hybrid (platform + usage) | Modern infra + AI tools | Predictable base + upside | Requires real billing infrastructure |
My current setup is a four-tier hybrid: a cheap entry tier for volume, two middle tiers where 80% of revenue lives, and a custom enterprise tier that I quote manually. The middle tier is where the business runs. It took me two years to admit the entry tier was a marketing expense, not a product line.
When usage-based actually pays off
Usage-based pricing works when your cost scales with usage and your customer's success scales with usage. Both conditions must be true. If your cost is flat but usage-based pricing lets customers rack up huge bills on light usage, you're gambling on their goodwill. They'll churn.
When hybrid beats pure
Hybrid models — a platform fee plus metered usage — are the fastest-growing in B2B right now for a boring reason: they let finance departments forecast. Flat usage makes CFOs nervous. A predictable floor plus variable ceiling is easier to sign. I moved to hybrid in 2023 and my enterprise close rate roughly doubled, though I'll admit the sample is small.
How to actually set the number
Stop looking at competitors' pricing pages as if they hold the answer. They don't know your cost structure either.
Step 1: find your willingness-to-pay anchor
Ask five existing customers one question: "What would you be doing if our product didn't exist?" The answer tells you the alternative cost. If they say "we'd hire a part-time contractor for $2,500/month," your $200 tier is underpriced by an order of magnitude.
Step 2: run a price sensitivity conversation, not a survey
Van Westendorp surveys are fine for FMCG. For B2B SaaS they're noisy — the person filling it out isn't usually the budget holder. What works better for me: a 30-minute call where I show three price points and ask "which of these would make you stop and think?" The flinch point is where the real information lives.
Step 3: test on new customers, not existing ones
If you change the price tomorrow, apply it to new signups only for the first 90 days. Track conversion rate by tier, and — this is the part people skip — track LTV/CAC per tier, not in aggregate. My blended LTV/CAC looked healthy at 4.2 for over a year. Per-tier, my entry tier was at 0.9. It was bleeding.
Step 4: instrument expansion revenue from day one
Expansion — upgrades, add-ons, seat growth inside accounts — is where growth compounds. If your billing system can't tell you expansion MRR by tier and by cohort, fix that before you touch price.
Transitioning between pricing models without burning your base
This is the part every guide mangles. Changing from flat to usage-based, or from three tiers to five, is not a pricing problem. It's a customer-communication problem wearing a pricing costume.
Three things I now do every time:
- Announce 60-90 days out. Not two weeks. Atlassian's own public guidance on price increases lands in the same window, and for good reason.
- Grandfather generously but with a time box. "Your current rate stays for 12 months" is a gift that has an expiry. "You stay at your old price forever" is a liability you'll be explaining to a future board.
- Give the discount as a migration credit, not a permanent price. It's cleaner on the books and cheaper in the long run.
I broke rule two the first time. Grandfathered a customer indefinitely, then had to raise prices again two years later. That conversation went far worse than if I'd set an expiry at the start.
Tools and calculators worth your time
You don't need much. A spreadsheet, willingness-to-pay conversations, and a billing system that can meter usage accurately.
Billing tools I've used and would use again:
- Stripe Billing — fine up to a few hundred customers, great for hybrid metering
- Chargebee or Paddle — better fit once you have procurement, tax, and dunning complexity
- Metronome — if you're doing serious usage-based and finance wants reconciliation reports
For calculators, ignore anything that asks for one number and spits out a price. They're toys. The only calculator that matters is one that compares LTV per tier against CAC per tier against marginal cost per tier. Build that in a spreadsheet in an afternoon.
The thing I wish I knew earlier
The first time I raised prices, I thought I was making a financial decision. I wasn't. I was making a positioning decision and pretending it was math.
That's the honest answer to how to price a SaaS product for growth. Every pricing change is a message about who you're for and who you're not for. Most of the anxiety around the number is really anxiety about that message.
So the question I'll leave you with isn't "what should I charge?" It's this: if you doubled your price tomorrow, which customers would leave — and would you actually miss them?
