Two ways to price a software product, and I will tell you early which one I favour for most B2B companies. The first is one price for one package: simple to explain, simple to quote. The second is tiers, usually three, with names like Starter, Business and Enterprise, each adding features, limits or service. The second is more work and is usually worth it. Here is the reasoning, including where I think it fails.
What a single price leaves on the table
Imagine, illustratively, 100 prospects for a team-collaboration tool, priced per user per month. They are not alike. Say 40 of them are small teams who would pay up to ₹900. Another 40 are growing companies who would go up to ₹1,600. The last 20 are large organisations that value admin controls and support and would pay up to ₹3,000.
With one price you must choose among three bad options. At ₹900 everyone can buy, giving 100 × 900 = ₹90,000 of monthly revenue per user, but you leave money on the table with 60 of them. At ₹1,600, the small teams walk away and you get 60 buyers: ₹96,000. At ₹3,000 only the 20 large firms remain: ₹60,000. The best single price earns ₹96,000.
With three tiers at ₹900, ₹1,600 and ₹3,000, and each group buying the one meant for it, revenue is 40 × 900 + 40 × 1,600 + 20 × 3,000 = ₹1,60,000. That is two-thirds higher. This is the whole theory of tiering: different customers value the product differently, and tiers let you charge each close to what it is worth to them.
Reality is less tidy
Real buyers do not sort themselves perfectly. Some who could pay ₹3,000 will notice that the ₹1,600 tier does nearly everything they need. Suppose 10 of the 20 large firms trade down. Revenue becomes 40 × 900 + 50 × 1,600 + 10 × 3,000 = ₹1,46,000. Still better than the ₹96,000 single price, but the benefit shrank by almost a third. Trade-down is the central risk of tiering, and the quality of the design is measured by how much of it you suffer.
What belongs in a tier, and what does not
The fences between tiers should follow things that rise with the customer’s size or seriousness, not things that merely annoy people. Good fences:
- Usage limits: number of users, records, API calls, projects.
- Controls that large organisations need, such as single sign-on, audit logs, role-based permissions and data residency options.
- Service levels: response time, a named account manager, uptime commitment.
Weak fences are core features withheld just to push people upwards. A buyer who feels the cheap tier was deliberately crippled resents the vendor. My view: put the features that make the product work in every tier, and charge for scale, control and service.
The middle tier does the work
Behavioural research on choice suggests that when people see three options, many drift to the middle one, and that a visibly expensive top option can make the middle look sensible. This effect is real but not guaranteed, and I would not build a company on it. Use it lightly: make the middle tier the one you genuinely want most buyers on, and give it the best value per rupee. The top tier is for those who truly need it, and its price should reflect that, not only act as a decoy.
Where I would stay with one price
Tiering is not always right. A very early product with ten customers and no data on who values what should often keep one price, learn, and add tiers later. A product where customers are nearly identical, such as a niche tool sold to one type of firm, may not benefit. And an enterprise sale where every deal is negotiated has an effective price per customer already; a public three-tier page there may be only a starting anchor for sales.
There is a cost side as well. Three tiers mean three sets of entitlements to test, document, support and bill. Billing systems break on odd plan changes, such as a mid-month upgrade, and sales teams invent exceptions. If your finance and engineering team cannot yet bill reliably, the extra revenue will be eaten by mistakes.
Making the call
Before choosing, ask whether your existing customers differ meaningfully in size, in use and in what they say they would pay. Look at the spread: if your largest customer uses twenty times as much as your smallest and pays the same, you already have your answer. Then design the tiers around the differences, test with a few customers before publishing, and watch the share choosing each tier in the first quarter. If almost everyone lands on the bottom, the fences are in the wrong place; if almost everyone lands on the top, you are probably underpricing the middle.
Reprice carefully. Existing customers on a single price will compare themselves with the new tiers, so offer them a path, such as a grace period at their old price, rather than a surprise.
Figures and names in the examples are illustrative.
Leave a Reply