How to set a pricing strategy for tech startup growth
A founder I mentor asked me the same question three times in one call: "What should we charge?" Each time I gave the same answer and each time he changed the subject. That is the most common pricing problem I see in tech startups — not a lack of frameworks, but a refusal to decide. Pricing feels like a final answer when it should be a moving dial.
Here is the thing: your price is the fastest lever you have on growth. Not your landing page. Not another feature. If you sell a €200/month product to 400 customers, a 15% price increase adds roughly €144,000 in annual revenue without a single new signup. A 15% increase in traffic to your funnel is far harder to buy. Yet most founders tweak the homepage copy ten times before they touch the number on the pricing page.
Key takeaways
- A 10–15% price increase usually beats a 10–15% traffic increase in terms of revenue, and costs far less to test.
- Before product-market fit, price low to learn, not low to win. The goal is signal, not margin.
- The 5 C's — company, customers, competitors, costs, channel — are a checklist, not a formula. Two of them usually dominate.
- Charm pricing (.99, .95) is weak in B2B SaaS and strong for self-serve consumer products. Context decides.
- Any pricing decision you cannot reverse in a week is a decision you will regret making in a hurry.
Price before product-market fit? Do it differently
Most pricing advice quietly assumes you already have customers who renew. If you are pre-PMF, that advice will hurt you.
Pre-PMF pricing has one job: make people react. A price that produces no objections is a price that produces no information. I once advised a small dev-tool startup that charged €9/month for a product used by teams of 20+ engineers. Every single trial converted. The founder read this as validation. It was the opposite — nobody cared enough to push back, so she learned nothing about where the ceiling sat. We tripled the price to €29 and conversions dropped to about half, but the customers who stayed were engaged, referred colleagues, and one of them asked for an enterprise contract within six weeks. That single change told her more about her market than four months of free-trial data.
Three practical rules for pre-PMF pricing
- Charge something. Free users pre-PMF are noise. Even €15/month filters for people with an actual problem.
- Pick a price you can defend in one sentence — "it costs less than the coffee budget of one engineer per week" is enough.
- Plan the increase in advance. Tell early customers their price is locked forever. It buys goodwill cheaply and gives you room to move up later.
The switch point comes when you have retention data. Once the same cohort renews two or three times without hand-holding, you are post-PMF and pricing becomes an optimisation problem: segments, tiers, expansion revenue. Before that, it is a learning problem. Treating a learning problem like an optimisation problem burns runway.
What are the 5 C's of pricing?
The 5 C's of pricing are company, customers, competitors, costs, and channel. Each answers a separate question: what does your positioning allow you to charge, what will buyers actually pay, what alternatives surround you, what is your floor, and what cut does the sales path take?
Most founders work through all five and treat them as equal. They are not. In my experience, two of them dominate at any given moment, and figuring out which two is the whole exercise.
| C | What it constrains | Weight |
|---|---|---|
| Company | Positioning, brand, roadmap ambition | Heavy if you are entering as a premium player |
| Customers | Willingness to pay, budget owner, urgency | Heavy, nearly always |
| Competitors | Anchors and reference points | Light, unless you are a near-substitute |
| Costs | Your floor and margin structure | Light for software, heavy for usage-based infra |
| Channel | App-store cuts, reseller margins, procurement discounts | Heavy if you sell through marketplaces |
Notice the asymmetry. Founders obsess over competitors because competitor pricing pages are public. Customers are not public, so they get less attention than they deserve. Real talk: a price set by copying a competitor's page is a price set by someone else's strategy.
What is the 80/20 rule for startups?
The 80/20 rule for startups is the observation that a small slice of your customers, features, or segments generates most of your revenue — and the practical instruction to find that slice and price for it specifically. In a pricing context, it usually means the top tier of accounts pays for the whole company.
I have seen this hold with uncomfortable precision. On a project two years ago, we segmented 340 paying accounts and found that 22 accounts produced 71% of revenue. Those 22 had very different needs from the other 318 — multi-seat management, audit logs, a procurement process, an actual human to email. The other 318 wanted a login and a credit card field. We had been selling both groups the same plan at the same price.
How to apply it without overcomplicating things
- Export revenue by account, sort descending, and mark where you cross 70–80%.
- Look at what that top group has in common — size, use case, integration needs, industry.
- Build a plan that fits only them, and price it against the value of the problem, not the cost of the servers.
- Leave the lower tier alone. It funds your acquisition engine, it is not your margin engine.
The mistake is trying to serve both groups with one plan "to keep it simple." Complexity lives in your sales motion anyway; pretending otherwise just means your biggest customers subsidise your smallest ones without knowing it.
Does the .99 trick actually work?
The .99 trick works in specific contexts and mostly fails in others. It is a consumer-psychology device: prices ending in .99 or .95 read as belonging to the tier below in a fast, low-attention purchase. In B2B software sold to a budget holder, it usually reads as odd or unserious.
Where I have seen it pay off: self-serve products with a card-on-file checkout and a price under roughly €50. Where I have seen it backfire: a €990/month plan listed as €989. Every finance person who approves that invoice will wonder whether the number was generated by an algorithm.
My short rule of thumb on price endings
Under €50, self-serve, no procurement: charm pricing is worth testing. Above €200, B2B, invoice-based: round numbers signal confidence and make the internal approval conversation easier. Between the two, run an A/B test on your own traffic instead of trusting anyone's rule, including mine.
One caveat that gets lost: charm pricing is nearly impossible to combine with annual prepay discounts, seat-based scaling, or usage bands without producing prices like €1,187.04. If your model is inherently computed, rounding is not a trick — it is basic cleanup.
What are the 7 pricing strategies worth knowing?
The classic seven are cost-plus, value-based, competitive, penetration, skimming, bundle, and freemium. You will recognise all of them. What matters is knowing which one your business can actually execute.
- Cost-plus — add a margin to your cost base. Simple, safe, and almost always leaves money on the table in software.
- Value-based — price against the outcome the customer gets. The most defensible option and the hardest to research.
- Competitive — anchor near rivals. Fine as a sanity check, dangerous as a strategy.
- Penetration — go cheap to grab share fast, then raise. Works if switching costs are high; a trap if your product is easy to leave.
- Skimming — launch high and drop. Rarely viable for startups with no brand equity.
- Bundle — combine features or services so the total is harder to compare. Surprisingly effective in practice.
- Freemium — free tier as an acquisition channel. Only works when the free tier has a real ceiling and the paid tier has a real reason to exist.
Here is where I'll take a side: for a tech startup with any kind of differentiated product, value-based pricing is the right default and I will argue for it every time. Penetration pricing is the one I see justified most often and regretted most often — because the customers you acquire on price leave on price.
Pricing on value when you have no data yet
You can run value-based pricing before you have analytics. Ask five customers what the problem costs them per month in hours, rework, or lost deals. If they say "about a day of an engineer's time a week," you have a floor and a ceiling in one sentence. Then price at a fraction of that figure — a fifth to a third is a reasonable starting band — and watch what happens to churn and to close rates. That is your real benchmark, and it came from a conversation, not a report.
The decision you are actually avoiding
Founders ask me about pricing strategy when the real question is simpler and more uncomfortable: are we willing to lose some customers on purpose? Because that is what a price increase is. You will lose a few. The ones you keep will fund the next stage of the company.
Set a date. Put a number on the page. Give yourself permission to be wrong by 20% and fix it in a quarter. The founders I have watched grow fastest were not the ones with the cleverest pricing model — they were the ones who changed the number four times a year and had the data to know why. Everything else is theory you can read anywhere.
