With NetSuite, the Renewal Is the Purchase

The first contract was signed at a discount everybody in the room was pleased with. Three years later the renewal arrives, the discount has unwound, three modules have been added since — consolidation, a warehouse capability, an extra sandbox — the user count has grown with the business, and the annual number is now a multiple of what was approved by the board.

Nobody was deceived. Every module was requested by the business and bought willingly. The user growth was real growth. And the first-year discount was always a first-year discount, which is a thing the contract said in a sentence that everybody read and nobody weighted.

This is the specific way NetSuite's cost behaves, and it is not a scandal — it is a structure, it is knowable in advance, and companies that understand it before signing end up paying materially less than companies that discover it at renewal. It sits at the top of the band map in who genuinely operates in each tier of this market, and the tier is bought on a commercial model rather than on a feature list.

The shape of the cost

Nobody can honestly publish a number, because it is negotiated. The structure, though, is stable and worth understanding as four separate things that escalate independently.

A platform fee. The base cost of having the system at all, before anybody uses it and before any capability is added.

User licences, in classes. Full users and lighter classes of access at lower prices. The design of that mix is a real exercise with a real saving in it, and it is usually done once at signature and never revisited even as the population changes.

Modules. This is where the cost lives. Consolidation, advanced revenue management, warehouse and inventory depth, planning, project accounting, additional currencies or subsidiaries — each is a line. The base product is capable and the capability you were shown in the demonstration is frequently not all in the base.

Everything around the licence. Sandboxes, environments, integrations, implementation, and the specialists who change it afterwards. This is a substantial figure and it is the one least represented in the proposal.

The important property of that structure is that three of the four grow without a new decision being made. Users grow with headcount. Modules get added one at a time, each a small approval. And the discount unwinds on a schedule. Only the platform fee is stable, and it is the smallest part.

Why the renewal is where the money is decided

At signature you have leverage: there are alternatives, nothing is running on it, and the vendor wants the logo. That leverage is spent almost entirely on the first-year discount, which is the least valuable thing it can buy.

At the first renewal you have less: the implementation is done, the switching cost exists, and the business is dependent. At the second and third you have very little, and this is precisely when the discount unwinds and the accumulated modules renew together.

So the negotiation to fight at signature is not the first-year price. It is the terms that govern years two to five, and they cost nothing to secure on the day you have leverage.

A capped uplift, stated as a number or an index, for a defined number of years. This is the single most valuable clause in the agreement and it is the one most often left uncapped because the conversation was about the discount.

The price of the modules you will predictably need, fixed now. You know roughly what is coming: another subsidiary, a warehouse capability, more users. A price agreed while you are still choosing is a different price from one quoted when you cannot leave.

The ability to reduce. Almost every agreement lets user counts go up. Getting the right to reduce at renewal, without penalty, is worth a great deal in a market where headcount moves both ways.

A stated multi-year price, not a stated multi-year commitment. These are different, and the second without the first is a commitment to an unknown number.

All of that belongs in the same conversation as the clauses that decide what leaving costs, and none of it can be added afterwards.

What actually drives the number up

Four things, in descending order of how often they surprise people.

Modules added one at a time. Each is a proportionate decision approved on its own merits. The aggregate is never reviewed, because no single approval was large enough to require it. The remedy is not to refuse them; it is to keep one running list of the full annual commitment and put it in front of the same person every quarter.

User count drift. Accounts for people who left, seats at a higher class than the role needs, and users provisioned for a project that ended. This is exactly the licence waste that accumulates in every estate, except that here it is your largest single subscription and the reconciliation is worth doing before every renewal rather than eventually.

The discount unwind. Contractual, disclosed, and routinely omitted from the internal five-year model because the model used year one's figure and grew it by inflation.

Integration and change work. Not a licence cost, and it is part of the annual reality. Specialist day rates for this platform are what they are, and a company with a steady flow of small changes should budget for them as a running cost rather than as projects.

None of that is unique to this product. What is specific is that the product is capable enough that the business keeps asking for more of it, which is the mechanism.

Getting the model right before you sign

The five-year picture is buildable at proposal stage and almost never is. Four columns per year: platform, users, modules, services. Then three simple assumptions written down as assumptions.

Headcount. Model the plan, not today. If the business plan says the company grows, the licence grows with it, and a five-year model built on current headcount is a model of a company that is not going to exist.

Modules. List the capabilities you know you will want by year three and price them now. The list is not speculative — it is usually the things that were cut from scope to make the first-year number work.

The uplift. Whatever the contract permits, not whatever inflation is. If the contract does not cap it, model a number that reflects the absence of a cap, and let that be the argument for capping it.

Build the picture the way a real five-year total cost of ownership is assembled, and compare it against the alternative on the same basis — which means including the alternative's internal owner, upgrade tail and reporting work, because a comparison that models one side honestly and the other optimistically is a sales document with your name on it.

If you are already three years in

The renewal is not a formality and it is more negotiable than it feels, provided you start early.

Start six months out. A renewal negotiated in the final fortnight is a renewal accepted. Six months gives you time to do the user reconciliation, to price an alternative credibly enough to be believed, and to hold a conversation rather than sign a document.

Reconcile the users first. Leavers, over-classed seats, project accounts. This is the one saving that requires no negotiation at all and it changes the baseline every other number is calculated from.

Audit the modules against usage. Something bought for a project two years ago is often still on the contract. Ask the question per module: who used this in the last quarter.

Establish what leaving would cost, quietly, before you talk. Not to threaten with it — to know. A finance director who has an accurate figure for the exit negotiates differently from one who has a fear, and the difference is visible across the table.

Be realistic about the alternative. You do not usually outgrow this product functionally; you outgrow the budget. If consolidation and revenue accounting are genuinely your hardest problems, the alternatives cost real money to replicate, and that comparison is set out honestly in the case both ways round. Pretending otherwise in a negotiation is a bluff that gets called.

The short version

The cost is a platform fee, users in classes, modules, and everything around the licence — and three of those four grow without anybody making a decision. The first-year discount is the thing everybody negotiates and the thing that matters least.

Spend your leverage at signature on the terms that govern years two to five: a capped uplift for a defined period, agreed prices for the modules you will predictably need, the right to reduce users, and a stated multi-year price rather than a multi-year commitment.

If you are already in, start the renewal six months early, reconcile the user list before anything else, audit each module against last quarter's usage, and know your real exit cost before the conversation rather than after it. The product is capable and the capability is why the number grows; that is a structure to plan against, not a grievance — and planning against it is ordinary work of keeping an accurate picture of what the estate costs.