Raising prices on customers you already have

You have not changed your prices in two years. You know it. The spreadsheet where you modelled it is still open in a tab somewhere. Every month you do not send the letter you take a small cut you never agreed to, and you take it quietly enough that nobody has to have a difficult conversation.

Stats NZ put annual inflation at 3.7 percent in the 12 months to the September 2026 quarter, and at 3.0 percent in the year to September 2025. Compound the two and prices across the economy rose about 6.8 percent over those two years. If the number on your invoices has not moved in that time, you have handed every existing customer a real-terms discount of about 6 percent. Nobody decided that. It happened because deciding was uncomfortable.

This is not the "you're too expensive" objection, which I covered in why "you're too expensive" is rarely about price, and it is not the discount you concede inside a live deal, which is what a discount teaches your buyer. Both of those happen before the customer has paid you anything. This one is harder. They already pay a number, they have budgeted for that number, and you are asking them to pay a different one for something that looks identical from where they sit.

The delay is almost never about the customers

Ask a founder why the increase has not gone out and you get a reason about timing. Push once and you get the real answer, which is nearly always a single account. One customer, usually an early one, usually on a price set when you badly needed the logo, often the one who takes reference calls for you. The reluctance is not about a price rise. It is about that one conversation.

That is worth saying plainly, because an entire book of business ends up held at the wrong price to protect one relationship. That trade rarely survives being written down, so write it down.

Work out the number before you work out the words

There are three inputs: what has happened to your cost to serve, what comparable products now charge, and what your newest customers already pay you. That last one is the most useful and the least used. If your last five deals closed at a number your three-year customers would find startling, you do not have a pricing problem, you have a two-tier book and a growing liability.

Set one new price and one date. Resist building it per customer. A per-customer number reads as negotiable, and anything negotiable invites every account to open a negotiation.

The order you tell people in

Tell your team first, with the reasoning and answers to the questions they will be asked. Then your largest and best-informed customers, individually, by phone or in person, before anything goes out in bulk. Then everyone else in writing, on the same day.

Getting that backwards is the common failure, and it is expensive. A founder sends one email at seven in the morning and spends the next fortnight on calls from exactly the accounts who deserved a heads-up. Those accounts are usually not objecting to the price. They are objecting to finding out the same way as everybody else.

What do you do with the account that threatens to leave?

Expect one, sometimes two. The first job is to work out whether you are hearing a threat or a decision. A threat asks you to respond; a decision tells you something and has already been made. Treat the first as a conversation and the second as information.

For a threat, the useful question is what they would need to see in order to stay at the new number. If the answer is a fair ask about scope, support or timing, meet it on scope, support or timing. Do not meet it on price. If the answer is only the old price, with nothing attached, you have just learned what that relationship actually is, and it is not a partnership.

Where you do have room is time. A three or six month runway to the new number costs you very little and lets a customer who genuinely has to re-budget do it without losing face. What it must never become is a quiet permanent exception, because exceptions leak. Someone talks, and then you are explaining to a second customer why the first one is special.

Put it in writing once

Make it one page: the new price, the date it takes effect, what is not changing, what has improved since they last had a price change, and the name of the person to call. No apology, and no lecture about inflation. The tone that lands is the tone of a business that has thought this through, not one that is sorry it came up.

A price rise is not a finance task that falls out of a spreadsheet at year end. It reaches into your positioning, your packaging and what your pipeline is worth. That is why it keeps getting postponed, and why it is worth doing properly rather than quickly.

So: what does your newest customer pay for the same thing your longest-standing customer is paying for, and could you say both numbers out loud, in the same room, to either of them?

Sources. Inflation: Stats NZ, "Annual inflation at 3.7 percent in September 2026", which states that "New Zealand's consumers price index (CPI) increased 3.7 percent in the 12 months to the September 2026 quarter"; and Stats NZ, "Annual inflation at 3.0 percent in September 2025", which states that the index "increased 3.0 percent in the 12 months to the September 2025 quarter". The two-year figure is my own arithmetic from those two published rates and nothing else: 1.030 multiplied by 1.037 is 1.0681, so 6.81 percent cumulatively. The real-terms discount is the same sum inverted: 1 divided by 1.0681 is 0.9362, a cut of 6.38 percent, which I have rounded down to "about 6 percent" in the text. Both are economy-wide consumer price measures, not a software price index and not a measure of any one company's cost to serve, so they describe the size of the drift rather than what any particular business should charge. I deliberately have not used the producers price index here: Stats NZ reports that the output PPI rose 1.6 percent in the June 2026 quarter, but an output index measures what producers charge rather than what it costs them to deliver, so quoting it as a cost-to-serve figure would be wrong. I have also not estimated a New Zealand software price index, because none is published and inventing one from these two numbers would not mean anything. The three inputs, the telling order, the threat-versus-decision test, the runway-not-exception rule and the one-page letter all come from my own work with New Zealand B2B tech founders rather than from published data.

Nick Burns is a fractional CRO for New Zealand B2B tech companies, at home and expanding internationally. He co-founded Emendo and sold it to McKesson, then the 14th-largest company on the Fortune 500, and has since helped 60+ tech companies grow sales.