B2B pricing strategy: redesigning prices and tiers without losing customers

Cost-plus

It is the most common method because it is easy to defend in front of a board: you know the cost, you apply a target margin and the price follows. The problem is that the customer does not buy your cost. On products where you deliver more value than average, you leave money on the table. On products where the market already pays less than your cost plus margin, the rep discounts to close and the margin is lost anyway, only without control.

There is a subtler effect. When cost goes up, the method tells you to raise price, exactly when the market is most sensitive. When cost goes down, it tells you to cut price, even if the customer was willing to pay the same.

Copying the competitor

Watching the competitor is necessary. Pricing off the competitor is not. Copying assumes their offer is equivalent to yours and that they know what they are doing. Neither is a given. It also turns every comparison into a price comparison, because you have given up explaining how you are different.

The useful question is not "what do they charge?" but "why does a customer who could buy from them choose us?" If there is no answer, the problem is not price. It is positioning.

The discount each rep decides

This is the most expensive way to set price and the least visible. Every rep has their own judgment, their own history with each customer and their own volume target. If commission is paid on revenue and not on margin, the incentive is clear: close, even at a discount. Over the years, customers of similar size and profile end up paying very different prices, and nobody can explain why.

None of the three methods is bad in itself. What is bad is using them without a structure on top that says what value we sell, to whom and at what defensible price.

The price waterfall was popularized by Michael Marn and Robert Rosiello in Harvard Business Review in the early nineties. The idea is simple and it is still the most useful diagnostic tool we know: start from the list price and subtract, line by line, everything that separates the price list from what the company actually keeps. What is left at the end is the pocket price.

Take a product with a list price of 100. An on-invoice customer discount is applied. Then, off-invoice, come an annual volume rebate, an early-payment discount, freight the company absorbs, a contribution to a promotion and a special term agreed years ago. None of those items looks serious on its own. Added up, the pocket price can end up well below 100, and part of that gap shows up in no sales report because it lives in accounting or in logistics.

How to build it, step by step

1. Pull twelve months of invoices at line level. Deliverable: a table with customer, product, quantity, list price and invoiced price.
2. Add everything that sits off-invoice. Rebates, early payment, freight and contributions, allocated to customer. Deliverable: the pocket price per line.
3. Calculate the dispersion. For each product, the range of pocket prices being collected. Deliverable: the price band per SKU.
4. Cross it with volume. Each customer on a chart: volume on one axis, pocket price on the other. Deliverable: the portfolio map.
5. Attribute the deviation. By rep, by branch and by line item. Deliverable: where the margin goes and whose decision it was.

Step three is the one that surprises people most. The average misleads. What matters is the dispersion. If two similar customers pay very different pocket prices for the same product, there is a leak the market does not explain.

When you plot volume against pocket price, intuition says large customers pay less and small ones pay more. What usually appears is a cloud: large customers who pay well, small customers on large-customer terms and a middle band where nobody knows why each one pays what they pay.

Do not apply this grid mechanically. A small customer with a low price may be a strategic account that is growing, or the entry point to a larger group. What the map gives you is the list of conversations you need to have, ranked by impact.

The good-better-best principle is old: offer three versions of the same solution, with a basic tier that covers the essentials, a middle tier that sells the most and a full tier that anchors the value of the other two. The customer stops asking "how much of a discount can you give me?" and starts asking "which tier do I need?"

Its limit is that almost all the literature comes from software, where tiers are built from features. In a distributor or an engineering firm, tiers are built from different material: service level, lead time, risk absorbed and access to people. That is where the differentiating value sits, and it is where it is given away today.

In professional services the logic is the same with other variables: number of revisions, response time, seniority of the assigned team, reporting and access to a partner. What does not change is the underlying rule: each tier has a clear reason for someone to move up, and that reason is a customer problem, not a feature of yours.

Criteria for deciding whether tiers make sense

• Yes, when the portfolio has different service needs and today everyone gets the same thing at the same price.
• Yes, when reps discount to win customers who actually need less.
• Not yet, when you do not know what it costs to serve each customer. Without that number, tiers are designed blind.
• No, when the product is a pure commodity and the customer only compares price per unit. There, the work is in the discount policy and the cost to serve.

The tool we use to structure this is the offer ladder.

Almost every B2B company provides services it does not charge for and that nobody decided to give away. They surface when you ask logistics, customer service and technical support what they do for each customer that is not on the price list. The list is usually long, and it surprises management.

The decision is not always to charge. Sometimes the right move is to include the service in the top tier to give it substance, or to drop it because it costs more than it returns. What cannot continue is the default: providing it free to whoever asks.

A discount is not bad. A discount without a rule is. A discount policy does three jobs: it protects margin, gives the rep a frame to negotiate with confidence and leaves a trail for every exception so it can be reviewed.

Three principles we always apply:

1. Every discount has a trade. More volume, a term commitment, prepayment, consolidated orders or a tier change. If the customer gives nothing, they get nothing.
2. Rep autonomy is a band, not a judgment call. Inside the band, the rep decides. Outside it, it goes up to whoever owns it.
3. Exceptions expire. Every special term has a review date. If it is not reviewed, it reverts to list.

The width of each band depends on your margin and your market. There is no right number that fits everyone, and be wary of anyone who gives you one without seeing your data.

The policy stands or falls on commission. If the rep is paid on revenue, every discount is free to them. If part of their variable pay depends on margin or pocket price, the policy enforces itself. You do not need to redesign the whole compensation system. You only need the discount to stop being neutral for the person who grants it.

Raising prices without losing customers does not depend on the letter you send. It depends on who gets the increase, how much, in what order and with what conversation behind it. An across-the-board increase is the most comfortable option and the one that works worst: it raises customers who already paid well, falls short on the exceptions and hands every customer the same argument to complain.

Segment the portfolio before you touch anything

The sequence, with a deliverable per step

1. Signed baseline. Pocket price, margin, average order and retention by segment before the change. Deliverable: the starting picture, approved by management.
2. Talk tracks and objections. Written before anyone talks to a customer. Deliverable: a one- or two-page document per segment.
3. Team training. Reps rehearse the objections with their manager. Deliverable: every rep knows what they can and cannot concede.
4. Strategic accounts first. A conversation, not a letter. Deliverable: an agreement or a known position for each account.
5. Written communication to the rest, with reasonable notice so the customer can plan. Deliverable: letter sent and logged in the CRM.
6. Follow-up after a few weeks. Call whoever has not responded and review orders. Deliverable: a list of at-risk customers with an owner for each action.

The talk track

A good price-increase talk track does not apologize and does not hide behind inflation. It explains three things: what has changed, what the customer receives and what options they have. The third protects the relationship most: if the customer can choose between keeping the price on different terms or moving to another tier, the conversation stops being a yes or a no.

The last objection connects to another guide: in many accounts, the person who receives the letter is not the one who decides. The buying center explains how to identify who does.

A price increase gets judged too early and with the wrong metric. If you only look at revenue in the first month, you will see the order pull-forward many customers make before the new price list takes effect, and you will draw false conclusions. If you only look at churn, you will miss the customer who keeps buying but has cut their order.

Read the results against the signed baseline, not against the sales team's impressions. The team will always remember the customer who complained and forget everyone who said nothing. The review fits into the period committee: which segment behaved as expected, which did not and what gets adjusted.

Raising the list price without touching the waterfall. The cause is that the price list is the only thing management controls directly. The result is that reps compensate with more discount and the pocket price barely moves.

Designing tiers without knowing the cost to serve each customer. It happens because cost to serve is spread across logistics, technical support and finance, and nobody adds it up. The tiers look good on paper and lose money in practice.

Writing a discount policy nobody applies. It usually happens when the policy is written without the reps and without changing commission. It is respected for two months and then habit returns.

Sending the increase by letter to everyone at once. It is convenient and looks fair. In reality it treats unequal customers equally and leaves strategic accounts learning about it from a generic email.

Giving in at the first objection. The team has not rehearsed and does not know what it can concede. The first account that pushes gets an exception, and word spreads faster than the letter.

Measuring at thirty days. Order pull-forward and stockpiling distort the first period. A serious reading comes after one full buying cycle of the typical customer.

1. Ask finance for twelve months of line-level invoices, with list price and invoiced price. That is the raw material for the waterfall.
2. Ask logistics and finance which rebates, freight, early-payment discounts and services are granted off-invoice, and to which customers. Note what nobody can answer.
3. Pick one high-volume SKU and calculate the pocket price for its top ten customers. If the dispersion is wide, you already have the case for management.
4. Check how commission is paid. If it depends on revenue alone, write it down as the first decision of the discount policy.
5. List the accounts you cannot afford to lose and who talks to each one. They will be the first conversations when the increase comes.

If you want to do it with method, our pricing and packaging work starts with this diagnostic and is part of commercial architecture. The client decides. We direct. Third parties execute.