How to Build a Pricing Strategy for a New Product

A founder checks competitor websites, chooses a lower number and calls it a pricing strategy. Another founder adds a free plan because most software products seem to have one. Both decisions may work.

But neither explains why the price fits the customer, product or business.

TL;DR — Read this first

Learn how to choose a pricing model, charging unit, packages and tests for a new product. Connect customer value, alternatives, costs and buying behaviour before launch.

What is a pricing strategy?

A pricing strategy is the logic used to choose a charging model, price level, package structure and commercial rules. It connects the value a customer receives with the way the business earns revenue. The first price is a hypothesis.

The strategy should explain the assumptions behind it and how those assumptions will be tested.

Pricing is more than the amount

A complete pricing decision includes:

• The paying customer
• The charging unit
• The price level
• The package or plan structure
• The billing period
• The free trial, free plan or entry offer
• Discount and exception rules
• Contract, cancellation and refund conditions
• The proof needed before a customer pays

Changing one part can change the whole buying experience.

A low monthly price may still feel expensive when the customer must sign a long contract or complete a difficult setup.

The five inputs into a starting price

1. Customer value. Estimate the time, money, risk or opportunity affected by the product. Value is specific to the customer and situation.
2. Current alternatives. Review what the customer pays today, including internal labour, manual work and the cost of doing nothing.
3. Willingness to pay. Use interviews, choice tests, offers, pilots and real sales conversations. Interest alone is not enough.
4. Cost and capacity. Understand delivery, support, infrastructure, sales and partner costs. Cost sets a practical boundary, not the full value.
5. Position and business goal. A premium specialist offer, low-cost self-service product and network marketplace need different pricing logic.

The Pricing Evidence Ladder

Treat pricing evidence in levels: founder estimate, competitor comparison, customer conversation, structured choice research, real offer, paid transaction and repeated renewal. The higher levels are stronger because they involve behaviour and commitment. The report should show which level supports each conclusion.

Enter your idea to map the price assumptions and evidence.

Common pricing models

One-time price

Useful when value is delivered in a defined purchase, report, service or asset. The business needs a plan for repeat revenue or continued support costs.

Subscription

Fits recurring value and ongoing service. The renewal period should match how often the customer receives value.

Usage-based

The customer pays for consumption, transactions, data, minutes or another measurable unit. The bill must remain understandable and predictable enough for the buyer.

Per user or seat

Simple when value grows with active users. It can discourage wider adoption when occasional users still need access.

Tiered packages

Different plans combine limits, features or service levels. Each tier should fit a real customer difference, not exist only to make the middle option look attractive.

Transaction or commission

Common in marketplaces and payments. The fee must fit margins, trust, disintermediation risk and the value controlled by the platform.

Service or outcome-based

A service may charge by project, time, milestone or result. Outcome pricing requires clear measurement and risk allocation.

How to choose a charging unit

The charging unit should grow with customer value, remain measurable and be difficult to misunderstand. A compliance tool could charge per employee, policy, legal entity, workflow or company. Each unit creates different incentives.

• Per employee is familiar but may punish adoption.
• Per legal entity may fit risk and reporting needs.
• Per workflow may align with use but create billing complexity.
• A company tier may be simple but less precise.

The best unit is not always the most technically accurate. It needs to work for the buyer and the business.

Value-based, competitor-based and cost-plus pricing

Value-based

The price is informed by the outcome and alternatives. It works best when the customer value is understood and measurable.

Competitor-based

Competitor prices provide category context. They do not show your cost, customer or value. Use them as evidence, not as the only answer.

Cost-plus

The business adds a margin to cost. This can protect delivery economics but may ignore customer value. It is common in services and physical products.

A starting strategy can use all three. Value creates the reason to pay. Competitors set expectations.

Cost protects the business boundary.

Free plan, free trial or paid pilot?

Free plan

A free plan can support distribution, product learning or a network effect. It also creates support and infrastructure obligations. Define the reason for keeping it free.

Free trial

A trial works when the customer can reach value within the trial period. A complex B2B product may need setup help instead of a short self-service trial.

Paid pilot

A paid pilot tests commitment and creates a real buying process. Scope, success criteria and the post-pilot price should be clear.

Free is a pricing decision. It is not a substitute for a customer acquisition strategy.

Worked example: pricing a compliance product

A founder is building a tool that helps professional-services firms create AI policies, track employee acknowledgement and organise compliance evidence. Competitors use per-user subscriptions, company tiers and consulting packages. The founder first considers EUR 5 per employee per month.

Customer interviews reveal a problem. Small firms do not think in employee licences. The buyer thinks about one company-wide compliance responsibility.

The number of employees affects risk and support, but the product value is not created by each person using the tool every day. A starting structure could use company tiers based on employee range and included legal entities.

Starter

One entity, up to 25 employees, policy setup and acknowledgement tracking. The goal is simple self-service compliance organisation.

Team

Up to 100 employees, multiple policy versions, evidence exports and administrator roles.

Advisory setup

An optional one-time service for firms that want help organising the initial inputs. This should not be presented as legal advice. The price is still a hypothesis.

The next test is not another competitor spreadsheet. It is a set of real offers to firms in the chosen segment, with the same scope and clear follow-up questions.

What a Pricing Strategy report should produce

• Paying customer and buying context
• Value and current alternatives
• Pricing model options
• Recommended charging unit
• Starting price range and assumptions
• Package or tier logic
• Free, trial or pilot recommendation
• Discount and exception principles
• Cost and margin considerations
• Pricing tests and evidence thresholds

What pricing analysis cannot prove

It cannot guarantee that customers will pay or renew. Public competitor pricing may not show discounts and contracts. Interview answers about willingness to pay may differ from real purchase behaviour.

The report should support testing. Financial, tax, legal and consumer-law matters may need professional review.

Common pricing mistakes

Choosing a lower price than competitors

A lower price does not create a position. It may also make trust harder in a high-risk category.

Pricing from cost only

Customers do not buy your development effort. Cost matters to the business, while value and alternatives shape the customer decision.

Adding too many tiers

Complex packaging makes a new product harder to understand. Start with customer differences that are supported by evidence.

Using an adoption-limiting unit

Per-seat pricing can reduce sharing and product spread.

Check whether the unit aligns with value.

Discounting without a rule

Uncontrolled discounts make learning difficult and create renewal problems. Record why a discount exists and what the standard price remains.

Testing the number without the offer

Price reactions depend on customer, problem, scope, proof, terms and support. Test a clear offer, not an isolated number.

How Pricing connects with other IdeaClarify phases

Pricing in a chat window vs IdeaClarify

An LLM can suggest pricing models, compare public competitors and generate tier names. It can help a founder see options quickly. The answer may still choose a familiar SaaS model without understanding buying behaviour, value, delivery cost or the charging unit.

It may also present a precise price without real evidence. IdeaClarify should carry forward the customer, strategy, competitor, product and business analysis. It should show how each recommendation was formed.

The report should label the evidence level and propose real pricing tests.

Frequently asked questions

Should a startup publish prices on the website?

It depends on the buying process and product maturity. Clear public pricing can reduce friction. Complex enterprise or service offers may need discovery.

Avoid hiding a standard price only to create sales pressure.

How do I know whether my price is too low?

Look at conversion, customer quality, support load, margin, objection patterns and willingness to pay. A high conversion rate is not enough when the business cannot support the customers.

Can I change pricing after launch?

Yes. Communicate clearly and consider existing customer commitments. Test changes by segment when possible so you can understand the effect.

Can students create a pricing strategy without sales data?

Yes. Use competitor evidence, customer value, costs and clearly stated assumptions. Propose the experiments that would test willingness to pay.

What comes after pricing?

Marketing Plan, Sales Playbook and Financial Forecast commonly follow. Pricing should also be reviewed during real sales and after early retention data appears.