Good-better-best is the default and usually one tier too many. Tiers should follow segments you have observed, not fill a pricing page.
Good-better-best is the default SaaS tier packaging structure, and it is usually one tier too many for a solo product. Three tiers require three segments you can name and describe differently. Most early SaaS has one.
Tiers exist to capture different willingness to pay, not to look established.
A pricing page with three columns implies you know three things about your market. Before twenty customers you know approximately none of them, so the tiers get invented — a cheap one to catch the hesitant, an expensive one to look serious, and a middle one to steer people toward.
That structure costs you in two ways. It splits your feedback across segments too small to read, and it makes every future pricing change three changes.
One price is also easier to defend in conversation, which matters when most of your early sales happen in a direct exchange. Start there and let the tiers emerge from evidence — the pricing framework covers which method should set that first number.
The split should follow a difference you have already observed, not one you expect. Three splits that work, in rough order of how well they hold up.
Usage volume. Some customers process ten of something a month and some process four hundred. This is the cleanest split because the buyer can self-select honestly and the difference is visible before they pay. It follows directly from your value metric.
Number of people involved. Solo versus small team. Only use this if your product genuinely changes with more people — otherwise it is a tax on collaboration and it pushes teams into sharing one login.
A capability only one segment needs. API access, exports, integrations, white-labelling. This works when the capability is genuinely irrelevant to the lower tier rather than artificially withheld.
| Split | Works when | Fails when |
|---|---|---|
| Usage volume | Volume varies widely and predicts value | Buyers cannot estimate their own volume |
| Capability | The feature is irrelevant to the lower tier | You crippled something obviously core |
| Seats | Value genuinely grows with team size | Most customers are individuals |
| Support level | Rarely for solo products | You are the support, at every tier |
A low entry tier looks like it widens the funnel. In practice, at solo scale, it usually selects for the customers who cost you the most.
The people who choose the cheapest option tend to have the smallest budgets, the least urgent problem, and — because the price is low relative to their expectations — often the highest support expectations. They also churn fastest, which distorts the churn number you are trying to read.
If you want a lower entry point, a limited free trial usually serves the purpose better than a permanent cheap tier — see freemium versus free trial for why. If you've already decided to run freemium anyway, freemium versus paid covers how to keep the free tier from cannibalising the paid one.
The highest tier does two jobs. It captures customers who genuinely need more, and it makes the middle tier look reasonable by comparison.
The second job only works if the top tier is credible. A tier priced at ten times the middle with no clear justification reads as a stunt and undermines the whole page. Two to three times the middle tier, with a visible reason, is the range that holds up.
A one-time lifetime option is a different instrument — it trades future revenue for cash now, and it belongs in a launch or funding-gap decision rather than in your standing packaging.
Four things that matter more than the tier count.
Test for whether your tiers are real: cover the prices and ask whether you can still tell who each tier is for. If the only difference is the number, you have one product at three prices and customers will notice.
Restructuring packaging is harder than changing a price, because existing customers have to be mapped onto a new set of options. Grandfather for twelve months and apply the new structure to new customers first — the price increase playbook covers the sequence, which is the same here.
If your problem is that revenue is flat rather than that pricing is wrong, packaging may not be the constraint at all. Breaking the MRR plateau covers the wider diagnosis.
One below roughly twenty customers, two between twenty and a hundred, and three only when you can describe the person who buys each in a sentence. Tiers exist to capture different willingness to pay, not to make the pricing page look established.
On a difference you have already observed. Usage volume is the cleanest split because buyers can self-select honestly. A capability split works when the feature is genuinely irrelevant to the lower tier. Seat-based splits usually fail when most customers are individuals.
Usually not at solo scale. The cheapest tier tends to select for the smallest budgets, the least urgent problems and the highest support expectations, and those customers churn fastest — which also distorts the churn rate you are trying to read.
Two to three times the middle tier, with a visible reason. The top tier both captures customers who need more and anchors the middle one, but the anchoring only works if the tier is credible. A tenfold jump with no justification reads as a stunt.
Apply the new structure to new customers first and grandfather existing ones for twelve months. Restructuring is harder than a price change because customers must be mapped onto a new set of options, so give more notice rather than less.
Tell Marcus your price, your customer count and how usage varies. You get a specific answer on how many tiers you should run.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime