I treated pricing like a final exam I could postpone until the product was “ready.” That meant I had a working app, a landing page, and no clue whether the price made sense for the person buying it. Not ideal.

There is no magic number hiding in a competitor's pricing page. Your price has to work for the buyer, the value they get, and the business you are trying to keep alive. That sounds obvious, but it helped me stop copying whatever a bigger SaaS company was charging.

Start with the cost of staying the same

Ask what the customer does today. Maybe they spend two hours a week chasing approvals, miss a booking because a calendar is stale, or hire a contractor to clean up data every month. The cost might be money, time, mistakes, or the mental load of remembering one more thing.

Do not assume all that value belongs to you. Customers still have budgets, other priorities, and switching costs. But if your product reliably gives back six hours a month, saves a missed deadline, or removes a recurring mess, you have a useful anchor for the conversation.

Ask a question like “What happens if this stays exactly the same for six months?” That often reveals the cost more clearly than “How much would you pay?” People are bad at inventing a price for a hypothetical, but they know what the existing workaround costs them.

Do the unglamorous business math

Add up your own costs: hosting, payment fees, support time, AI usage, backups, contractors, and the boring admin stuff. Then estimate how many customers you can support without making every Tuesday miserable. If the product is usage-heavy, include a realistic busy month, not only the average.

Now decide what a healthy customer looks like. A $9 plan can be great if support is light and usage stays predictable. A $99 plan can be a bargain if it replaces a weekly manual workflow. Price does not tell customers whether the product is “nice.” It communicates who it is for and how much help they should expect.

A founder and small business owner discuss a product offer together

Sell the outcome before you add tiers

Founders love making three columns because the template has three columns. Then we invent limits that only make sense to ourselves. Before creating Starter, Growth, Pro, Studio, and Mega Studio, try one plan with a clear outcome and a clear boundary.

For example: “$24 per month for one shop, unlimited reminders, and the first 500 tracked invoices.” That sentence lets a buyer tell whether it fits. A plan name like “Momentum” doesn't.

You can add tiers when customers have real differences in needs: one-person use versus team approvals, basic reports versus multiple locations, or self-serve versus hands-on setup. Don't gate an essential success step to push someone into a higher plan unless the extra value is real.

Have pricing conversations without the weird pitch voice

Show the price while the customer can still ask questions. Ask what they compare it to and what part feels expensive or confusing. If they say “that seems fine,” follow with “what are you comparing it against?” The answer might be another app, the cost of a part-time assistant, or absolutely nothing.

You can test a paid pilot, a monthly subscription, annual billing, or a one-time setup fee. Change one thing at a time, keep notes, and honor the offer you already made. Don't pretend a discount expires in two hours because your conversion dashboard looks lonely.

When someone refuses to pay, be curious. Was the pain low? Did the product miss the problem? Is their budget already spoken for? Is the buyer different from the user? A no can be about any of those. Discounting before you know the reason teaches you very little.

Keep AI and variable costs in view

If your app calls a model, calculate the cost for different usage patterns. A normal customer might send ten short requests; a power user might upload long docs all afternoon. Include retries and tool calls. Consider reasonable limits, caching, smaller models for simple work, and a friendly heads-up before a customer hits a boundary.

Do not make the customer guess whether a button will trigger a costly action. Clear usage language is part of trust. And if your provider bill jumps, check cost per successful task, not just cost per request. A lower-cost model that fails often may cost more after retries and cleanup.

A useful pricing check-in

Write down these numbers once a month: revenue per customer, direct service cost, support time, refunds or cancellations, and how often a customer reaches the intended outcome. Then talk to a few customers. Numbers tell you what changed; conversations help you understand why.

The right price is not the one that feels least awkward to say out loud. It's the one that lets the customer get value and lets you keep delivering it. Start simple, be transparent, and adjust when you have evidence.