When Your Software Vendor Raises the Price

An email arrives in November. The platform your business runs on is increasing by forty percent from January, and there is a paragraph about continued investment in the product.

You have three weeks to decide, four years of data inside it, and eleven people who know how to use it. Everybody understands what happens next.

The increase is priced off your switching cost

This is not opportunism, it is arithmetic, and understanding that is the first step to handling it well.

A vendor knows roughly what it would cost you to leave: the migration, the retraining, the integrations that would break, the months of reduced output. They price the increase comfortably below that number. As long as staying is cheaper than going, the increase holds, and they can repeat it next year.

Which means the size of the increase you receive is a direct measurement of how trapped you are. It is unflattering information and it is genuinely useful.

What actually locks you in

Rarely the software. Four things, in descending order of how much they cost to unwind.

The data, in a shape only that system understands. Exports frequently produce a technically complete file that loses the relationships between records, which is where the meaning lives.

The integrations. Every other system that reads from or writes to it. Each one is a small project to rebuild and they are rarely documented anywhere.

The habits. Eleven people who are fast in this system and would be slow in another for six months. This is usually the largest real cost and it never appears in a comparison spreadsheet.

The undocumented logic. The custom fields, the workflow somebody configured in 2022, the report the bank asks for. Nobody knows the full list, which is why migrations overrun.

What to do in the three weeks you have

You will probably pay it. Do three things anyway, because they change next year.

Ask for the multi year rate. Two or three years at a fixed price is frequently available and rarely offered. Vendors value predictable revenue and the discount for asking is real.

Ask what you are paying for that you do not use. Seats for people who left, a tier bought for one feature, modules nobody opens. Most businesses find something, and it is the least confrontational form of negotiation available.

Get the increase in writing with a cap. A commitment that future rises are limited to a stated percentage is worth more than a one off discount and vendors give it more readily.

Then start the work that means the next email is a smaller problem.

Reduce the lock in without leaving

The choice is not stay or migrate. There is a middle position and it is where the value is.

Export monthly, automatically, to somewhere you control. Not as a migration plan. So that leaving becomes a project rather than an archaeology exercise, and so you have your own history if the relationship ends badly.

Write down the integrations. One page listing what connects, in which direction, and what breaks if it stops. An afternoon, and it is the document that makes any future decision possible.

Document the custom logic. The fields, the rules, the reports somebody built. This is the part that always gets discovered halfway through a migration.

Keep the core of your process outside the platform. Wherever the important logic can live in something you own, put it there. Then the platform becomes a place data is stored rather than the thing your business is.

Where AI changes the arithmetic

Migration used to be brutally expensive for one specific reason: matching two systems that describe the same world differently is judgement work, at volume, and judgement meant people.

Mapping ten thousand records where the old system has one address field and the new one has four. Deciding which of three similar customer records is the real one. Working out that a status called Pending in one platform corresponds to two different states in the other. Reading four years of free text notes and turning them into structured fields.

That was the bulk of the cost and the bulk of the risk, and it is exactly what AI now does well. A migration that was a six figure project with a real chance of failure is, for a business of your size, frequently a matter of weeks.

Two consequences. Your genuine switching cost is lower than it was three years ago, which changes what you should accept. And the effort of keeping an exportable, documented position is far smaller than it used to be, because the documentation itself can be assembled from the system rather than written by hand.

You may still choose to stay. Choosing to stay is a completely different commercial position from having to.

When leaving is genuinely the right call

Three situations where paying the increase is the more expensive option, and they are worth recognising because the default is always to stay.

The product has stopped improving. Rising price against a static roadmap is a company harvesting an installed base rather than building. That trajectory does not reverse, and every year you stay makes the eventual move larger.

You are on the wrong shape of product. Businesses grow into platforms built for a different size or sector, and end up paying enterprise pricing for something they use like a spreadsheet. The increase is a good prompt to ask whether this was ever the right fit.

The lock in is about to get worse. If you are about to add two more integrations or another forty users, moving costs less today than it will next year. The best time to leave is always earlier than it feels.

Do not decide this in three weeks

The deadline is the vendor’s, and it exists precisely to prevent a considered decision.

Pay the increase if you must, then run the evaluation properly over the following quarter with no deadline attached. A migration decided under time pressure in December is how businesses end up on their second bad platform.

The question to ask before you buy anything next

Every future supplier gets one question, asked in the first conversation, and how they answer tells you more than the demonstration.

If we left in three years, what exactly would we take with us, and in what format.

A good supplier answers immediately and specifically. A poor one talks about how nobody leaves. That reaction is the most useful thing you will learn in the sales process, and it costs nothing to ask.

AI Optimize does the data mapping and the record matching that used to make migration unaffordable, and builds so that the important logic sits somewhere you own. That work sits under Custom Software.