SaaS Pricing Tiers: A Founder's Guide

Most bootstrapped SaaS needs exactly three pricing tiers: a cheap entry plan, a core plan where most of your revenue will live, and a higher plan that anchors value and captures larger accounts. Design the middle tier first, build the other two around it, and tie every tier to a value metric the customer can predict. Two tiers leave money on the table. Five tiers create decision paralysis.

I have priced products badly enough to respect this. The temptation as a builder is to list features and slap a number on each plan. That produces a pricing page that describes your software instead of selling an outcome. The fix is structural, and it is the same for almost every early SaaS.

This post covers the three-tier skeleton, why you price the middle first, how to choose a value metric, what the anchor tier is really for, and how to test the price before you commit to it. If you have not yet found a buyer to price for, start there; then choose between a free trial and freemium and check it against your break-even MRR. Then track every pricing change on a founder dashboard. For B2B, pre-selling validates the price before you build, and productizing a service gives you real pricing data first. And if you sell to developers, developer-tool pricing plays by its own rules. And before you can charge in dollars at all, forming a US LLC as a non-resident is the entity step that makes the pricing real.

Key takeaways

• Most bootstrapped SaaS needs exactly three tiers: an entry plan, a core plan, and an anchor — not two, not five.
• Design the middle (core) tier first, then build the other two around it.
• Tie every tier to a value metric the customer can predict.
• The anchor (top) tier mostly exists to make the middle tier the obvious choice.
• Pre-sell to validate the price before committing, and raise prices deliberately as value grows.

Why this matters for solo founders

Pricing is the most powerful number you control. A change to your price flows straight to the bottom line with no extra work, no new feature, no new customer. Patrick Campbell’s team at ProfitWell spent years publishing data on this, and the recurring finding is blunt: founders underinvest in pricing relative to its impact, often spending a few hours on it total.

For a solo founder, pricing also decides what kind of business you are running. A $5/month plan and a $200/month plan demand different customers, different support loads, and different acquisition channels. You are not just setting a number. You are choosing who you serve.

How many pricing tiers should a SaaS have?

Three is the default for a reason. It gives buyers a clear comparison without overwhelming them, and it lets you serve three real segments: the price-sensitive entry buyer, the core customer, and the larger account willing to pay for scale or support.

Two tiers force a binary choice and usually undersell. Without a higher anchor, your top price looks expensive instead of reasonable. Five tiers do the opposite damage. Each added option raises the cognitive cost of deciding, and a buyer who cannot decide does not buy. The psychology is well documented: more choices past a small number reduce conversion, not increase it.

There are exceptions. A pure usage-based product may have one plan and a meter. A product selling only to enterprise may have “contact us” and nothing else. But for the common case, a bootstrapped self-serve SaaS, three named tiers is the shape that works. Start there and deviate only with a reason.

Price the middle tier first

Founders usually design pricing left to right, starting with the cheap plan. Reverse it. Design the middle tier first, because that is where most of your revenue will come from and where most of your customers will land.

The middle tier is your real product. It should contain everything a serious customer needs, priced at the number you actually want most people to pay. Once that anchor is set, the other two define themselves. The entry tier is the middle with the most valuable things removed, priced low enough to remove the risk of trying. The top tier is the middle plus the things larger accounts need: more usage, more seats, priority support, security and compliance features.

A concrete way to set the middle number: estimate the monthly value your product creates for a typical customer, then price at a fraction of it. If your tool saves a customer several hours a month, price against the cost of those hours, not against your hosting bill. Cost-plus pricing, where you mark up your own costs, is the most common bootstrapper mistake. Your costs are irrelevant to the buyer. The value to them is the only anchor that matters.

What is a value metric, and how do you pick one?

A value metric is the unit your price scales on: per seat, per project, per thousand API calls, per contact, per gigabyte. Choosing it well is more important than choosing the dollar amounts, because it determines whether your revenue grows as your customers get more value.

The test of a good value metric, drawn from the usage-based pricing work that OpenView and Kyle Poyar have published openly, is threefold. It should align with the value the customer receives, so they pay more only as they get more. It should be predictable, so a buyer can estimate their bill before committing. And it should be hard to game, so customers cannot extract heavy value while staying on a low tier.

Per-seat pricing is predictable and easy to understand, which is why so much B2B SaaS uses it. Its weakness is that it can punish adoption: if every new user costs money, teams ration access, which slows the word-of-mouth growth you want. Usage-based pricing aligns better with value but is less predictable, which scares some buyers. Many products land on a hybrid: a base platform fee plus a usage component, capturing the predictability of seats and the value alignment of usage.

Pick the metric a customer would name if you asked “what do you get more of as this becomes more valuable to you.” That is almost always the right axis to price on.

What should each pricing tier actually contain?

Gate on capacity and stakes, never on basic usability. The entry tier should solve one complete problem for one person. The core tier adds volume, collaboration, and integrations. The anchor tier adds administrative control, compliance, and support guarantees. Anything a customer needs in order to trust you belongs in every tier.

Gate on this	Never gate this	Why the line sits there
Seats, projects, workspaces	Data export	Holding data hostage produces refunds, chargebacks, and public complaints
Usage volume: API calls, contacts, storage	Two-factor authentication and encryption	Charging for basic account security reads as a hostage fee, not a feature
Integrations with paid business tools	Bug fixes and core reliability	A broken cheap plan is a broken product, and reviewers do not distinguish
History and data-retention window	The first complete win	An entry tier that cannot deliver one whole outcome is a demo with a price
Support response-time guarantees	Basic email support	Silence toward paying customers costs more in reputation than the tier earns
Roles, audit logs, admin controls	Onboarding help in the first week	Early churn is far more expensive than the support hour you saved

Single sign-on is the genuinely contested case. Putting SSO in an enterprise tier is defensible, because the buyers who require it also have budget and procurement processes that cost you real time. Putting two-factor authentication behind a paywall is not defensible, and the internet reliably notices.

Apply the upgrade-trigger test to every boundary. For each tier, name the specific event that makes a customer outgrow it: the fifth teammate, the tenth project, the first month they exceed 10,000 API calls, the day their security review lands. If you cannot name that event in one sentence, your tiers are cosmetic and buyers will sit on the cheapest one indefinitely.

The strongest signal that the boundaries are right is that customers upgrade without you asking. If every upgrade requires a sales conversation, the tiers are not tracking the value the customer receives, they are tracking your hope.

The anchor tier exists to sell the middle tier

The top tier does a job most founders miss. Its first purpose is not to be bought often. It is to make the middle tier look like the reasonable choice.

This is price anchoring, and it is one of the most reliable effects in pricing psychology. A $199 plan next to a $49 plan makes the $49 feel modest. Remove the $199 and the $49 becomes your ceiling, and ceilings feel expensive. The expensive option reframes the middle option as sensible, even for buyers who never seriously consider the top plan.

So design the anchor deliberately. It should be a real plan that some larger customers genuinely want, not a fake number. But its day job is to shift the comparison in favor of the tier you actually want most people to choose. Companies like Basecamp have at times gone the other direction with radical simplicity, a single flat price, and it works for them because their brand and distribution carry it. For most early founders without that brand pull, the three-tier anchor structure converts better.

Pricing-page mechanics that reduce decision friction

The tiers are the strategy. The page is where the strategy succeeds or fails. A few mechanics consistently help.

Mark one tier as recommended. Buyers want to be told where to look. Highlighting your middle tier as “most popular” guides the eye and reduces the work of deciding. Make the comparison scannable: a short feature list per tier, not a 40-row matrix that buries the differences. Lead each tier with the outcome it unlocks, then the features, then the price.

Show the price. Hiding everything behind “contact sales” is right for genuine enterprise deals and wrong for self-serve SaaS, where a hidden price reads as “expensive and slow.” Offer annual billing at a discount, because it improves your cash flow and your retention at once, and label the annual saving clearly. Keep the number of decisions on the page small. Every extra toggle, add-on, and asterisk is friction between the buyer and the purchase.

How much does it cost to accept payments for a SaaS?

Budget roughly 3 to 6 percent of revenue for payments, and more if you sell internationally. Card processing is the visible part. Currency conversion, subscription billing tooling, sales-tax compliance, failed payments, and disputes are the parts founders leave out of the model until the first payout lands smaller than expected.

Line item	Published rate at time of writing	Cost on one $49 charge
Card processing, US domestic	2.9% + $0.30	$1.72
International card surcharge	about +1.5%	+$0.74
Currency conversion	about +1%	+$0.49
Subscription billing tooling	about 0.5% of recurring revenue	+$0.25
Sales-tax and VAT automation	about 0.5% per transaction	+$0.25
Dispute fee	around $15, flat, per dispute	The fee plus the reversed charge

Rates move and vary by country, so check Stripe’s published pricing for your own market before you build the model rather than trusting a table in a blog post. The shape is what matters: a domestic-only US customer costs you about 3.5 percent all in, and an international customer paying in another currency can cost closer to 6 percent.

A merchant of record changes the arithmetic and the legal position. Services like Paddle charge more, commonly around 5 percent plus a fixed per-transaction fee, and in exchange become the legal seller of your software. They register for and remit VAT, GST, and US sales tax in dozens of jurisdictions on your behalf. For a solo founder selling globally, and particularly for one running US company rails from outside the US, paying two extra points to not maintain tax registrations you cannot realistically administer alone is frequently the correct trade.

The pricing implication is direct: set your number against what you keep, not against what the customer pays. The gap between 3 and 6 percent on a $49 plan is about $1.47 a month per customer, which is roughly $1,700 a year across 100 subscribers. That figure is illustrative arithmetic rather than a benchmark, but the discipline behind it is not optional.

One line item hides inside the others. Cards expire, get replaced, and get declined, and every failed renewal is revenue you already earned and are about to lose. Retries and dunning emails recover part of it, and that recovery is one of the few revenue gains available to you without acquiring a single new customer.

How pre-selling validates the price before you commit

You do not have to guess the price into existence. You can test it before the product is finished, which is the safest way to set it.

Put up a page with the three tiers and real numbers, and take a commitment against it: a deposit, a discounted annual pre-pay, or a letter of intent at a named price. What people do when a real number is in front of them tells you far more than what they say in a survey. Surveyed willingness to pay is notoriously inflated, because saying yes to a hypothetical costs nothing.

For B2B specifically, a pre-sell conversation doubles as price research. Quote a number and watch the reaction. A buyer who agrees instantly told you the price is too low. A buyer who pushes back hard told you it is above their value perception, or that you are talking to the wrong buyer. The friction in that conversation is the signal. The cost of running this test is a landing page on Cloudflare Pages, a domain for about $10, and Stripe in test mode. Under $15 to avoid mispricing a product for a year.

When and how to raise your prices

The first price is a hypothesis, and the most common direction it needs to move is up. Bootstrapped founders almost universally start too low, because pricing from a position of low confidence feels safer at a small number. The result is a product that stays underpriced for years and a founder who has to acquire twice as many customers to reach the same revenue.

The signals that your price is too low are specific. Customers say yes without hesitating. Nobody ever pushes back on the cost. Your buyers turn out larger or more serious than you expected. Support is light relative to revenue. Each of these means you have room above your current number, and the room is pure margin.

The safe way to raise prices is on new customers first. Leave existing customers on their current plan, or grandfather them for a long, clearly communicated period, and raise the price for everyone who signs up after a chosen date. You learn the effect of the new price on conversion without breaking trust with the people who bought early. If conversion holds, the increase flows straight to the bottom line. If it drops sharply, you have found your ceiling cheaply and can adjust.

Test increases in real increments, not timid ones. Moving from $19 to $21 teaches you almost nothing. Moving from $19 to $29 produces a signal you can actually read. Patrick McKenzie’s widely shared advice on this is blunt and correct: charge more, and then charge more than that. The discomfort of the higher number is not evidence that it is wrong. It is the normal feeling of capturing the value you create.

A worked example: building a three-tier ladder from zero

Here is the whole method run end to end on one hypothetical product. Every number below is illustrative, chosen to show the arithmetic. None of them are results from a product I operate, and you should replace all of them with your own inputs.

The product: a reporting tool for operations managers at small logistics firms.

Step 1, estimate the value. The tool removes about four hours a month of manual spreadsheet work. At a fully loaded cost of roughly $50 an hour for that role, the monthly value created is about $200.

Step 2, take a fraction of it. Software tooling that captures 10 to 25 percent of the value it creates is a defensible starting band, which puts this product between $20 and $50 a month. Pick $49 for the core tier, because it sits at the top of the band and leaves room to discount rather than room to regret.

Step 3, build outward from the middle.

Tier	Price	What it contains	Who it is for
Entry	$19	One user, one workspace, 30 days of history	A single manager testing it on one site
Core	$49	Five seats, unlimited workspaces, integrations, 12 months of history	The real product, and where most revenue lives
Anchor	$149	Roles and audit log, SSO, priority support, unlimited history	A multi-site operation with a security reviewer

Step 4, check the margin. On the $49 plan, subtract about $1.72 for card processing and, on this hypothetical stack, about $1.50 of infrastructure per active customer. Gross margin lands near $45.80, or roughly 93 percent, which is the shape a subscription business needs in order to survive its own support costs.

Step 5, connect it to survival. If your fixed business costs plus your living costs total $2,500 a month, then at 93 percent gross margin you need about $2,690 of MRR, which is roughly 55 customers on the $49 plan. That number is the honest goal your pricing page is working toward, and it is the reason the break-even MRR calculation belongs beside the pricing decision rather than after it.

Step 6, pressure-test the anchor. At $149, the buyer needs roughly three times the value, so the anchor tier has to plausibly save twelve hours a month or protect something expensive. If it cannot, buyers read it as a decoy, and a decoy that gets noticed damages trust in the whole page.

The Three-Tier Skeleton worksheet

Fill this in before you publish a pricing page.

1. Value metric: the one unit your price scales on, and why a customer would name it.
2. Middle tier: the price you want most customers to pay, and the value estimate behind it.
3. Entry tier: the middle minus its most valuable parts, priced to remove the risk of trying.
4. Anchor tier: the middle plus what larger accounts need, priced to make the middle look reasonable.
5. Annual discount: the saving you offer for paying yearly.
6. The pre-sell test: the commitment you will collect against these numbers before building further.

What are the most common SaaS pricing mistakes?

Five mistakes cost bootstrapped founders the most, and none of them are about picking the wrong dollar amount. Pricing against gross revenue, shipping plans your billing system cannot version, over-deep annual discounts, coupons with no expiry, and naive regional pricing are all structural choices made at launch that become expensive to unwind.

Pricing against gross revenue instead of net. Payment processing, currency conversion, and sales tax take somewhere between 3 and 6 percent before you see a cent. A margin model built on the sticker price is wrong from day one, and the error compounds with every international customer you add.

Shipping plans your billing system cannot version. Grandfathering existing customers is the standard way to raise prices without breaking trust, and it only works if your billing setup can hold multiple live versions of the same plan. Founders who hard-code a single price object discover this the week they want to raise prices, and the fix is a manual data migration performed on live subscriptions.

Annual discounts that are too deep. Two months free, roughly 15 to 20 percent, is the convention because it buys a year of committed cash without gutting the plan’s value. Half off annual trains every buyer to wait for the annual deal and turns your headline monthly price into fiction.

Coupon codes with no expiry. A discount code issued once and never retired escapes onto deal sites and stays there for years. Set an expiry date and a redemption cap on every code you create, without exception, because a coupon is a price change wearing a costume.

Naive purchasing-power pricing. Regional pricing is a legitimate strategy and a real one for founders selling into markets with very different incomes. Done carelessly it is arbitraged within days: a VPN, a foreign card, and your discounted regional price becomes your global price. If you offer it, verify against something harder to spoof than an IP address, and accept that some leakage is the cost of the strategy.

What I would do differently

The argument against this structure is worth taking seriously. Rigid three-tier pricing can be wrong for genuinely novel products where the buyer has no reference price, and for usage-heavy infrastructure where a meter beats named plans. Pricing is contextual, and a framework is a starting point, not a law.

The deeper mistake, and one I have made, is treating pricing as a one-time decision. You set it, ship it, and never touch it again because changing prices feels risky. That is backwards. Pricing is the experiment you should run most often, because it is the cheapest change with the largest effect. Raise the price on new customers and watch conversion. Test a new value metric on a cohort. The founders who compound treat the price as a dial they are always adjusting, not a number carved at launch.

If I were starting a new SaaS today, I would set the three-tier skeleton, pre-sell against it to pressure-test the numbers, and then plan to revisit the price every quarter. The first price is never the right price. It is the first hypothesis.

Want the system, not just the article?

The Bootstrapped Founder Operating System includes the Three-Tier Skeleton as a fillable pricing worksheet, a value-metric picker, and a pre-sell pricing-page template. Stop guessing at the number. Launch price: $29. Get the workbook →