Most MSPs are competing on price by accident
Nobody sets out to be the cheap option. But if the customer cannot tell the difference, price is the only thing left to decide on.
Ask an MSP owner whether they compete on price and almost all of them say no. Ask how many of their last ten deals came down to the number, and the answer is usually most of them.
That gap is not self-deception. It is what happens when a business differentiates in ways the customer cannot see. Your engineering is better. Your response times are better. Your people stay longer. All true, all invisible from the outside — and a customer comparing three proposals that describe the same services in the same language will decide on the one number that differs.
Undifferentiated is not a marketing problem. It is a pricing problem wearing a disguise.
Two proposals, one visible difference.
Scope, hours, licences, total
The proposal describes what will be done. Every competitor’s proposal describes roughly the same thing, because the work genuinely is similar. Quality lives in delivery, which the customer cannot assess until after they have chosen.
The only comparable field is the price.
What changes, and what that is worth
The proposal describes the outcome in the customer’s own numbers — hours released, cost avoided, capacity gained — and the price sits against that. The comparison is no longer with a competitor but with doing nothing.
Price becomes a proportion of value.
If the customer cannot tell the difference, price is the only thing left to decide on.
This matters far more in an AI practice than it did in managed services, and the reason is arithmetic. Managed services have a reference price — per user, per device, per month. Customers know roughly what it should cost, and so does everyone bidding.
AI work has no reference price. Nobody in the customer’s business knows what a process redesign and an agent deployment ought to cost, which means the price is set entirely by the frame you put around it. Framed as a project, it is compared to other projects. Framed against a business case, it is compared to the value of the change.
Two MSPs can quote the same work at very different numbers and both be accepted, because they are answering different questions.
Four habits that quietly set the frame.
None of them looks like a pricing decision at the time.
Quoting before you have diagnosed
A number produced quickly to be helpful anchors the whole conversation on cost, before anyone has established what the change is worth.
Describing the work, not the outcome
A proposal listing days, workshops and deliverables invites a comparison of days, workshops and deliverables. It is a shopping list, and shopping lists get shopped.
Pricing from your cost base
Day rate times days plus margin produces a defensible number that has no relationship to what the outcome is worth to the customer. It is arithmetic, not pricing.
Talking to the wrong person
The IT manager has a budget and a comparison set. The CFO has a problem and a return threshold. The same proposal is expensive to one and cheap to the other.
Establish the value before you name the number.
The sequence matters more than the technique. Find the process that is costing them, quantify it in their own figures, agree that number with them, and only then put a price beside it. A customer who has agreed that a process costs them £400,000 a year does not experience a £60,000 engagement as expensive — they experience it as a proportion.
That is not a negotiating trick. It is simply doing the commercial work first, which is the same discipline every other part of this business rests on: win the business case before you win the build.
And if you cannot quantify it, that is the finding. An opportunity whose value nobody can express is an opportunity that will be decided on price no matter how you write the proposal.
Four things worth remembering.
- Nobody chooses to compete on price. It is what is left when the customer cannot see a difference.
- AI work has no reference price, so the frame you put around it sets the number almost entirely.
- Quoting early, describing the work, pricing from your cost base and talking to the wrong person all set that frame without anyone deciding to.
- Quantify the cost of the problem in the customer’s own figures before you name a price. If it cannot be quantified, that is the finding.
How many of your last ten deals came down to the number?
A Growth Review is a structured conversation with you and your leadership team: where your AI practice is constrained, whether your offers are priced against value or against your cost base, and what changing that is worth. Ninety minutes, no cost, and an honest answer — including when the answer is that you do not need us.
Win the business case before you win the build.