Chapter 1
Why Pricing Is the Most Underrated Growth Lever
Startups spend months on product features and weeks on marketing. On pricing they often spend a few hours. This is a strategic mistake. McKinsey studied the growth levers for SaaS companies: a 1% price increase raises profit by 11.1%. A 1% rise in sales volume raises it by 3.3%. A 1% cut in CAC raises it by 3.75%. A 1% cut in churn raises it by 6.71%. Pricing is the most powerful lever at minimal operating cost.
“We don't know what price to set” is the normal state for an early-stage startup. But that is no excuse for a random price. The right approach: start with a value-based pricing hypothesis. Test it. Iterate. Most startups set a price once and are afraid to change it. This is a mistake, because pricing must evolve along with the product.
Expensive vs. cheap: the intuition of most founders says “set it lower to attract more customers.” The data says the opposite. A low price attracts price-sensitive customers with high churn. It limits perceived value. It kills the margin for Customer Success and Sales. A high price attracts customers who truly value the product. It gives you resources for better service. It improves retention. Patrick McKenzie (Stripe): “Raise your prices. No, higher. Higher than that.”
Cost-plus pricing (bad): price = costs + margin. Why it is bad: it does not reflect the value to the customer. It ignores competitors. It does not scale as the product's value grows. Competitive pricing (neutral): look at competitors and price similarly. A neutral approach, but not optimal. Value-based pricing (good): price = a percentage of the value you create for the customer. It is harder but significantly more profitable.
Chapter 2
Three SaaS Pricing Models
There are three main models: Flat Rate, Per Seat, and Usage-Based. Each has its own advantages and suits different types of products.
Flat Rate (a fixed price): one plan, one price, all features. Example: Basecamp, $99/month for unlimited users. Pros: simplicity for the customer, predictable MRR, easy to sell. Cons: it does not scale as the value to the customer grows, and small companies pay as much as large ones. When it fits: products with limited seat needs, and when retention matters more than expansion revenue.
Per Seat (per user): price × number of users. Example: Slack, $7.25/user/month. Pros: easy to understand, revenue grows as the customer hires, and expansion revenue is built in. Cons: it encourages customers to limit sharing, and it is hard for large team deployments. When it fits: collaboration tools, and products where each user means separate value.
Usage-Based Pricing (UBP): payment for usage (API calls, transactions, storage). Examples: AWS (per GB), Stripe (2.9% + $0.30 per transaction), Twilio (per SMS). Pros: revenue is proportional to value, the barrier to entry is low, and the customer pays more when they get more. Cons: unpredictable MRR, complex billing infrastructure, and customer anxiety about the bill. When it fits: infrastructure products, products where usage is a clear proxy for value, and AI/LLM products (per token).
Hybrid models (the best for most): a base fee plus usage. Example: a $49/month base plus $0.01 per API call. It gives predictability plus room for expansion. Seat tiers plus feature gating. Example: Starter $29/month (3 users), Pro $79/month (10 users plus advanced features), Enterprise custom. This is the most common structure among successful SaaS companies.
Good/Better/Best (3-tier pricing): it almost always wins. The middle plan always sells best (the anchoring effect). The expensive plan makes the middle one look reasonable. The cheap plan lowers the barrier to entry. Name plans by persona or outcome, not by features: “Starter,” “Growth,” “Enterprise” are better than “Basic,” “Pro,” “Premium.”
Chapter 3
Freemium vs. Free Trial vs. Demo-Only
This decision affects the whole go-to-market. There is no universally right answer. It depends on the product and the market.
Freemium: a free plan forever with limited functionality. Classic examples: Slack, Dropbox, Spotify, Notion. When it works: network-effect products (value grows with every user), a large horizontal market (millions of potential users), and viral product-led growth potential. When it does not work: expensive CAC, B2B enterprise with a long sales cycle, and niche markets where freemium does not give the scale you need. The freemium trap: if your slogan is “free,” then an upgrade is a loss for the customer, not a gain. Freemium → paid conversion is 1–5% (Spotify: 2.6%. Dropbox: 4%). That is fine only at huge volume.
Free Trial: full access for a limited period (14 or 30 days). Time-limited (14 days) vs. action-limited (the first 100 events). Time-limited creates urgency better. Action-limited shows value better before the paywall. When it works: a product with tangible value that can be felt within 14 days, and B2B with a moderate CAC. When it does not work: if 14 days is not enough to feel the value, or if setup takes most of the trial. Recommendation: make onboarding so fast that the key value is reached in the first 24 hours of the trial.
Demo-only (sales-led): there is no self-serve trial. You have to talk to sales. Examples: Salesforce, most enterprise software. When it works: ACV > $10,000/year, complex setup and customization, and lead qualification is necessary. When it does not work: if your customers want to try before they talk (developer tools, consumer SaaS).
The trend of recent years: Product-Led Growth (PLG) plus Sales-Assisted. Start with a free trial for bottom-up (individual users). A PQL (Product Qualified Lead) algorithm identifies who is ready to upgrade. Sales contacts only warm leads that have shown usage signals. Atlassian, Figma, and Notion Enterprise use this hybrid approach.
Chapter 4
How to Find the Right Price: Experiments
The right price cannot be calculated. It has to be found through experiments and data.
Van Westendorp Price Sensitivity Meter: 4 questions worth asking your users or potential customers: “At what price would the product seem expensive, but you would still consider buying it?” “At what price would the product seem too expensive, and you would not buy it?” “At what price would the product seem cheap, a good buy?” “At what price would the product seem so cheap that you would doubt its quality?” The intersection of the answers gives an acceptable price range. It is a qualitative method to start with.
A/B testing prices: show different prices to different segments. Measure conversion rate, ACV, and churn. This works for high-volume self-serve. For B2B enterprise with a small number of deals, it is harder.
A real tactic: cohort pricing. Offer new users plan A or plan B (different prices). Existing users keep their grandfathered rate. This gives you data without conflict with existing customers.
Willingness to Pay interviews: ask 5–10 potential customers directly: “If we launched tomorrow, how much would you pay?” Most founders are afraid of this question. But the answers are gold. “Nothing” is valuable information too. “$500/month” means you may be badly underpriced.
Signs that the price is too low: no price objections on sales calls. Low churn (customers see the value). Good expansion revenue. Customers say “this is very affordable.” If there is not a single price objection, you are underpriced. The target is 20–30% price objections. That is normal. Signs that the price is too high: low conversion despite a good product. Many prospects drop off exactly at the pricing stage. Customers ask for a discount in the first meeting.
When to raise prices: the product has grown significantly. Churn is low (customers are happy). You have a queue of sales conversations. You have more inbound than you can handle. For existing customers, communicate in advance and give a lock-in period (grandfather for 12 months). For new customers, just raise the price.
Chapter 5
Anchoring, Packaging, and the Psychology of Price
The psychology of pricing is a well-studied field. These principles work regardless of industry.
Anchoring: the first number the customer sees becomes the anchor for all later comparisons. If you show 3 plans, show the expensive one first (left to right). The customer sees $299 → $99 → $29, and $99 looks reasonable. If they had seen $29 first, $99 would look expensive. Annual vs. monthly: “$99/month or $890/year.” The annual anchor makes monthly look expensive.
Decoy pricing: three plans where the middle one is deliberately designed to be the one people pick. A classic example: Small: 5 users, 10 projects, $29/month. Medium: 20 users, unlimited projects, $79/month. Large: 50 users, unlimited projects, $89/month. Large is the decoy: who would pick 50 users for $89 when 20 users cost $79? Almost no one. But the presence of Large makes Medium look like exceptional value. Medium sells significantly better in this structure.
Charm pricing ($99 vs. $100) is well studied: a price ending in .99 or .97 is perceived as significantly lower. $97 vs. $100, with the same difference ($3), is perceived as bigger. B2B enterprise: less effective (procurement buyers are less emotional). B2C and SMB: it works.
Annual billing: offer a 15–20% discount for annual. This improves cash flow, LTV, and churn (annual contracts churn significantly less often). How to nudge toward annual: a “Most popular” badge on annual. Show the monthly equivalent: “$74/month when paid annually.” Onboarding specifically for annual: an extra onboarding call.
Upgrade paths: make upgrading friction-free. Self-serve upgrade: the customer can upgrade in 2 clicks. Proactive upgrade prompts: when the customer approaches a limit, it is a natural moment for an upsell. In-app prompts: “You've used 80% of your monthly quota. Upgrade for unlimited?” Success-based upsell: “You generated 100 leads this month. The Pro plan gives advanced analytics for those leads.”
Download this guide in EPUB and PDF
