Three tiers are enough for most B2B offers, and the middle one is what most buyers will choose: it deserves the most work. A tier is not split by features but by customer profile and moment of maturity. The test that settles it: a prospect should know within thirty seconds which one is for them, without calling you.
When a company calls us to review its prices, the problem is rarely the amount. It is the split: how the offer is cut, named and presented. Price is only the last line of a story that plays out earlier, and we develop that point in value before price.
This article covers the mechanics of the split itself. How many tiers, what goes in each, and how to check that the architecture holds.
Three tiers, and why three
Three is the format that works for most B2B offers, for a reason that has nothing to do with aesthetics.
A prospect does not judge a price in the abstract, they judge it by comparison. With a single tier they have no reference and compare against their expectations, which is to say against nothing. With three, they place themselves: they see what they do not need to pay for and what they would gain by going up.
Beyond four, choosing becomes work. A prospect facing six columns does not choose, they postpone, and postponement is the worst commercial outcome there is. A fourth tier is only justified when an enterprise segment has genuinely distinct requirements, of compliance or service, that do not belong in the others.
Split by profile, not by feature
This is the most widespread and most expensive design error. Features get listed, spread across columns, and the prospect is asked to evaluate your product before knowing whether it concerns them.
A good split starts from the customer. Each tier matches a profile and a moment of maturity: the person starting alone, the team getting organised, the organisation that has to govern and integrate. Features follow that logic instead of dictating it.
The test is direct: a prospect should be able to say within thirty seconds which one is for them, without calling you. If they have to compare line by line, your split is technical rather than commercial.
That assumes knowing who your profiles are, which points back to the work described in why an ICP that is too broad kills growth. A badly split offer is often the symptom of an unsettled segment.
The middle tier sets your average revenue
It is the one most people will choose, and it is almost always the one designed fastest, squeezed between two others that are more fun to write.
Design it from your typical customer, the one you serve best and who stays. It should contain exactly what that customer needs, no more and no less. Adding things to make it attractive devalues the tier above; removing things to push people upwards produces frustration and discount requests.
One sign the middle is miscalibrated: a large share of your customers take the low tier and then ask for exceptions. That means the middle is too expensive or badly composed for them, and your grid does not match your actual base.
What triggers the move to the next tier
A tiered architecture only pays if moving up is natural. Otherwise you have three separate offers, and every customer stays where they entered.
The trigger must be an objective, legible fact: a number of users, a volume handled, a governance need that appears with size. It is observed rather than negotiated, and the customer can anticipate it.
A vague trigger, of the “for advanced needs” kind, moves nobody up. It turns every upgrade into a commercial negotiation, which costs time and usually ends in a discount. The unit you bill on largely determines this mechanism, a subject we cover in usage-based pricing.
The three splits that do not work
The low tier that serves no purpose. Built to display an entry price, it is deliberately crippled to the point of being unusable. The prospect understands, and concludes you are trying to push them up. You lose trust before the conversation.
The top tier with no real content. An “enterprise, contact us” column containing nothing different beyond the price. It reassures nobody and it deprives your grid of its anchoring effect, which is precisely its job.
The exhaustive list. Twenty-five comparison rows where only three matter. It creates an impression of transparency and produces the opposite: the prospect cannot find the line that concerns them and gives up.
Publishing prices, or not
The question comes up in every project, and the answer depends less on your market than on how legible your offer is.
A prospect who cannot place themselves before talking to you assumes it is expensive, and some will never call. You will never see those lost deals, which makes the decision comfortable and wrong.
The position that works in most cases: publish at least an order of magnitude on the first two tiers, and keep bespoke work on request. The prospect places themselves, they know whether they are in your market, and your sales team stops qualifying companies that could never have paid.
Checking the architecture holds
Three checks, before publishing anything.
Show the grid to five existing customers and ask which tier they would take today. If they do not pick the one they have, your split does not match your base.
Then look at the real distribution of your current customers. If more than half sit on the low tier, your middle is miscalibrated. If nobody is on the top one, it has no reason to exist in that form.
Finally, count the exceptions granted over the last twelve months. Each exception is a place where the grid did not match reality, and their distribution sketches the split you should have made.
Where to start
Take your last twenty customers and place them in your current grid. Then look at how many obtained an exception, and on what.
In most companies we open up, exceptions cluster on two or three identical points. Those points are your new architecture, and it is dictated by your market rather than by a workshop.
Frequently asked questions
- How many tiers should you offer?
- Three in most cases, four if an enterprise segment has genuinely distinct requirements. Below three you lose the comparison effect that helps a prospect place themselves. Above four, choosing becomes work and the prospect postpones, which is the worst outcome.
- What should tiers be split on?
- On customer profile and moment of maturity, never on a feature list. A feature split forces the prospect to evaluate your product before knowing whether it concerns them. A profile split lets them recognise themselves, then verify.
- How do you design the middle tier?
- By starting from your typical customer, the one you serve best and who stays. The middle tier should contain exactly what they need, no more and no less. It is the one most people will pick, so it sets your average revenue, and yet it is the one designed fastest.
- Should you publish your prices?
- Publish at least an order of magnitude, yes, unless your offer is genuinely bespoke. A prospect who cannot place themselves before talking to you assumes it is expensive, and some will never call. Bespoke work can stay on request, provided the tiers below it carry numbers.