The Decision Most Companies Get Wrong By Default
Pricing is the single most leveraged decision in most software businesses. A ten percent improvement in pricing produces, roughly, a ten percent improvement in revenue, with no additional engineering cost. The same percentage improvement is rarely achievable through any single feature. And yet pricing is the decision most companies invest the least time in. Many startups set their first prices in an afternoon and barely revisit them for years. Many established companies haven't reviewed their pricing structure in over a decade.
There are reasons for this. Pricing feels political. Changing prices upsets customers and sales teams. The data needed to price well is harder to gather than the data needed to ship features. Most PMs receive no training in pricing. The result is that pricing is left to instinct, accident, or copying competitors, and the leverage is wasted.
This article describes how to think about pricing as a craft rather than an afterthought. We will cover the major pricing approaches, the components of a pricing decision, common mistakes, and how to test pricing without burning customer relationships.
Three Approaches to Pricing
Every pricing decision is, at root, an answer to a single question: what is this product worth, and to whom? Three broad approaches answer this question differently.
Cost-Plus Pricing
Calculate what it costs you to deliver the product, add a margin, and that is the price. Common in commodity goods and certain services. Largely irrelevant to software, where marginal cost is near zero and value can vary wildly across customers. Yet some early-stage companies still anchor on cost-plus thinking, underpricing themselves dramatically.
Competitive Pricing
Look at what competitors charge and price near them, slightly above or below depending on positioning. This is the default in many markets and produces stable but unimaginative pricing. It works fine when your product is comparable to competitors. It dramatically undervalues genuinely differentiated products and dramatically overvalues weak imitations.
Value-Based Pricing
Price based on the value the product delivers to the customer. If your product saves a customer fifty thousand dollars per year, you can capture some portion of that, often ten to thirty percent, as price. The customer is still better off (they keep the rest), and the price is many multiples of what cost-plus would suggest.
Value-based pricing is the right approach for most software products, but it is also the hardest. It requires understanding what specific value you deliver, to whom, and how that value varies across segments. Most companies want to do value-based pricing in theory and end up doing competitive pricing in practice because the value research is hard.
The Components of a Pricing Decision
Pricing is rarely a single number. It is a structure with multiple components, each carrying real weight. Most pricing improvements come from optimising the structure, not just the headline number.
The Pricing Metric
What unit do you charge for? Per seat per month? Per transaction? Per gigabyte stored? Per API call? Per outcome achieved? The choice of pricing metric is one of the most consequential pricing decisions, because it determines how revenue scales with customer success.
The right metric tracks the dimension along which the customer experiences value. If your product's value scales with the size of the team using it, per-seat pricing aligns. If value scales with the volume of transactions processed, transaction pricing aligns. Misalignment between metric and value creates leakage on both sides: customers who get high value pay too little, customers who get low value pay too much and churn.
The Tier Structure
Most software products have multiple tiers. The structure of the tiers shapes how customers self-select and how revenue scales with adoption. Common patterns include three-tier (starter, professional, enterprise), good-better-best, and freemium-plus-paid. Each pattern has implications.
The most common mistake is too many tiers. Five or six tiers create choice paralysis and dilute the message. Three tiers are usually sufficient. The second mistake is non-meaningful differentiation between tiers: customers cannot tell why they should choose one over another. The third is making the highest tier feel unreachable, which caps revenue from customers who would have grown into it.
The Anchor Price
Customers evaluate price by comparison. The anchor (the first price they see, or the highest visible price) shapes their perception of value across all tiers. Many companies have a deliberate high-end tier whose primary purpose is to make the middle tier feel like a reasonable choice. The anchor matters even when nobody buys it.
Add-Ons and Usage Charges
Beyond the base price, what additional charges exist? Premium support? Additional storage? Per-API-call overages? Add-on modules? These add complexity but can capture significant revenue from customers who want more. The risk is making the total cost unpredictable, which damages trust. Many SaaS products lose enterprise deals because the customer cannot predict their annual bill.
Discounting and Negotiation
What discounts do you offer, and to whom? Volume discounts, annual prepay discounts, multi-year discounts, segment-specific discounts? Each discount lever is a real revenue decision, even though it may feel like a sales mechanism. PMs should be involved in these decisions because they shape what customer behaviour you reward.
Choosing the Right Pricing Metric
The choice of pricing metric deserves separate attention because it is so consequential and so often gotten wrong. The right metric satisfies several criteria simultaneously.
- 1. It tracks value. Customers who get more value pay more, customers who get less pay less. The metric scales with the underlying outcome the product enables.
- 2. It is predictable for the customer. Customers can estimate their cost in advance, even as their usage grows. Unpredictable bills produce frustration and churn.
- 3. It is simple to explain. The metric should be understandable in one sentence. Customers should not need a spreadsheet to compute their bill.
- 4. It is hard to game. Customers should not be able to reduce their bill significantly without reducing their value.
- 5. It scales with the business. As the customer's business grows, the price should grow with it, ideally capturing the natural expansion that comes from success.
6. It aligns sales and customer success incentives. Both sides should be rewarded for
customer outcomes, not for extracting more from existing customers without delivering more value.
Pricing Metric Best For Watch Out For
Per seat Tools used heavily by individual Sharing of accounts, single-seat
users in teams workarounds, underpricing
collaborative use
Per active user Products where engagement is the Customers gaming who counts as
value driver active, unpredictable monthly bills
Per transaction Workflow tools where transactions Customers consolidating
are the unit of value transactions to reduce cost
Per consumption Infrastructure, APIs, AI products Bill predictability, customers (storage, calls) throttling usage out of fear, hard
caps required
Flat per organisation High-trust enterprise relationships, Underpricing very large
where seat-counting creates organisations, capping growth friction
Outcome-based Mature products where outcomes Hard to set up, attribution disputes
can be measured cleanly
How to Test Pricing Without Burning Customers
Many companies are paralysed about pricing because they fear that any test will upset existing customers. This concern is real but solvable. There are several ways to test pricing without disrupting current relationships.
Test on New Customers Only
The most common approach. Existing customers are grandfathered into their current pricing. New customers see the new structure. Over time, the new structure becomes the default. This is slow but low-risk and is how most companies update pricing.
Geographic Tests
Roll out new pricing in one country first, observe, and roll out further if it works. The downside is that geographic differences may make the test results non-representative.
Segment-Specific Tests
Different pricing for different customer segments (mid-market vs enterprise, or by industry). This is common and rarely controversial because segments expect different pricing.
Willingness-to-Pay Research
Survey existing and prospective customers about willingness to pay. The Van Westendorp Price Sensitivity Meter is one structured approach: ask customers four questions about price points they would consider too cheap, cheap, expensive, and prohibitive. The intersection of the resulting curves identifies a range of acceptable prices.
Pricing Page Variations
A/B test different pricing presentations on the public pricing page. Track conversion. This tests pricing perception and structure without changing the underlying economics for any customer who actually buys.
Common Pricing Mistakes
Mistake One: Underpricing Through Fear
Most early-stage companies underprice. The founders are afraid to lose deals, so they price low. Customers buy and the company celebrates revenue, but unit economics are poor and the company cannot fund the team it needs to grow. Underpricing is more expensive than overpricing in most cases, because customers acquired at low prices are hard to migrate to higher prices later.
Mistake Two: Round Numbers Without Justification
Prices like nine, nineteen, ninety-nine are everywhere. They exist because of cognitive anchoring, not because they reflect value. Sometimes the right price is twenty-three dollars, not nineteen, and saying so confidently with a clear value story is more powerful than capitulating to round-number conventions.
Mistake Three: Pricing for the Average Customer
If you set a single price that works for the average customer, you under-monetise customers who would pay more and price out customers who would pay less. Tier structures address this by letting customers self-select. A single-price product is leaving money on both ends.
Mistake Four: Confusing Price With Pricing Strategy
Many companies debate the price number while ignoring the structure that produces it. Adjusting the headline price by ten percent is a small lever. Restructuring tiers, changing the pricing metric, or adding strategic add-ons are large levers. Spend time on the structure before debating the number.
Mistake Five: Letting Pricing Decay
Prices set five years ago at one company stage may be wrong now. The product is more valuable, the market has matured, customers have higher willingness to pay. Companies that do not periodically revisit pricing accumulate a slow form of underpricing that becomes structural. Annual pricing reviews are healthy.
Mistake Six: Discount Culture
Many B2B companies have a culture in which sales teams discount every deal. Customers learn this and come to expect it. Within a few years, the list price is irrelevant, every deal is negotiated, and the company has lost pricing power. Once discount culture takes root, it is hard to remove. The fix is discipline at the executive level, plus tier structures that reduce the need for discounting.
Packaging: The Companion to Pricing
Packaging is the question of which features go into which tiers. It is at least as important as the price numbers. A well-packaged product self-segments customers cleanly: small customers find what they need in the entry tier, large customers find what they need in the enterprise tier, and the middle tier is the natural home for most growing customers.
Principles of Good Packaging
1. Differentiating features should follow customer size or maturity. Features that small
teams need belong in the starter tier. Features that only larger organisations need belong in higher tiers. This makes self-segmentation easy.
2. The starter tier should be genuinely useful. If the starter tier is so limited that it is
unusable, customers will not adopt the product at all. The starter must deliver real value while leaving room to grow.
- 3. Each tier should have a clear identity. Customers should be able to say "we are a Pro customer because we have X teams" or "we are an Enterprise customer because we need Y compliance". Without clear identity, customers churn down over time.
- 4. The expansion path should be obvious. Customers who are growing should know exactly when and why to move up. Friction in the upgrade path costs significant revenue.
5. Some features should be add-ons, not tier-gated. Features wanted by some customers
but not all (advanced security, premium support, additional integrations) work better as add-ons than as tier-gated. Tier-gating these forces customers up an entire tier for one feature.
A Process for Reviewing Pricing
If you have inherited pricing or are preparing for a periodic review, here is a sequence that has produced reliable results.
- 1. Gather actual data: average revenue per customer by segment, discount levels, win/loss reasons, churn correlation with price.
- 2. Survey willingness to pay among existing and prospective customers. Look for clusters around natural price points.
- 3. Map current packaging against current customer behaviour. Where do customers cluster? Where do they get stuck? Where do they downgrade?
- 4. Compare to competitor pricing as one data point, not the anchor. Identify where you are over- or under-priced relative to value.
- 5. Develop two or three pricing structures that address the observations. Test the structures on a sample of customers or prospects.
- 6. Choose a structure, communicate the change clearly, and roll it out (typically to new customers first, with grandfathering for existing).
- 7. Measure the impact for at least one quarter before declaring the change successful or unsuccessful. Conversion rates, average deal size, and time-to-close are the primary signals.
A Final Word
Pricing rewards investment more than almost any other product decision. PMs who develop pricing as a competence have a structural advantage in their organisations and produce outsized impact for their companies. The work is more analytical than most product work and less visible than most feature work, which is precisely why so few people do it well and why the leverage remains.
If you do nothing else after reading this article, do this: spend one afternoon understanding your product's current pricing structure in detail. The metric, the tiers, the discount levels, the average revenue per customer by segment. Most PMs find, when they do this exercise, that their intuitions about how the pricing actually performs were wrong. The data is usually surprising. That surprise is the start of becoming useful in pricing decisions.
Key Takeaways
- Pricing is the highest-leverage decision in most software businesses and the most under-invested. A few percent of pricing improvement equals significant revenue.
- Value-based pricing is the right default for software, but it requires real research. Most companies fall back to competitive pricing by default.
- The pricing metric (what you charge per) is more consequential than the price number. Choose a metric that tracks customer value.
- Test pricing on new customers, segments, or geographies rather than disrupting existing relationships. Most price increases produce smaller conversion drops than expected.
- Packaging matters as much as price. Each tier needs clear identity, the starter must be useful, and the expansion path must be obvious.