What pricing models do AI consultancies use?
Most offers in this market fall into a handful of models. Each puts risk in a different place, and that matters more than the headline figure.
- Time and materials. The buyer pays for time. It is simple and flexible, but the total is unknown until the work ends, and the buyer carries the cost of every delay, change and misjudged estimate. The provider has no reason to finish sooner.
- Retainers. A regular fee for ongoing access to advice or a set allowance of time. Useful for a long relationship, but it can pay for availability rather than results, and unused time rarely rolls over.
- Fixed fee for a defined result. One price, agreed before work starts, for a specified outcome. The provider carries the risk of overrun, so the scoping has to be done properly at the start.
- Fixed fee for a document. An assessment, roadmap or feasibility study for an agreed sum. The price is fixed, but what is delivered is advice about what the firm could do, not a change to how it works.
- Licence-led pricing. Consulting bundled with, or recovered through, software subscriptions. The cost continues after the project, and the recommendation may lean towards the product being sold.
The trade-off between charging for time and charging for a result is explored in fixed fee or daily rates.
What should a buyer compare between quotes?
Comparing figures alone is misleading, because quotes rarely buy the same thing. Line them up on these points instead:
- What exists at the end. A working process the team uses on live work, or a document describing one?
- Who carries overrun. If the work takes longer, who pays?
- What is included. Mapping, building, testing, training and support after launch, or only some of them?
- What continues afterwards. Licences, hosting, support contracts or further phases.
- What counts as done. Delivered to the firm, or live and in daily use?
Comparing AI consultancies goes further, and questions to ask an AI consultant gives wording you can use.
How does Aldbry set its fee?
Aldbry's model has two parts: one fixed implementation fee, agreed in writing before work starts, then a subscription once the system is live. Changes outside the agreed scope are quoted in days before any work starts. The implementation fee is set by the process being rebuilt, and three things drive it:
- How many steps the process has. A job with a short, fixed sequence is simpler to rebuild than one with many stages, branches and handoffs between people.
- How many documents and templates feed it. One template and a single source of figures is a smaller job than several templates, a precedent library and inputs from across the firm.
- What it has to connect to. A process that works within documents and email is simpler than one that reads from a practice management system, a case management system and shared spreadsheets.
The implementation fee covers the build itself: scoping, the system built on your own systems and documents, a parallel run against the current process, and go-live once your team signs it off. From then on the subscription covers hosting, support, monitoring, a monthly performance report and a quarterly review, with no per-user fees. Timing is set at scoping. The stages are set out in how we work.
Why is the fee scoped after the audit?
Because the fee depends on the process, and the process has to be known before it can be priced. The audit estimates where your team's admin hours go and names the one job to tackle first. Once that job is identified, it can be scoped against the three drivers above and the implementation fee and subscription agreed. Pricing before that point would mean guessing, and a guessed fixed fee is either padded or wrong.
How do you judge whether the fee is worth paying?
Set it against what the process costs today. Work out the hours the job takes across the firm each week and what those hours are worth at your charge-out rate over a working year; calculating unbillable hours explains the method. Then compare that recurring cost with the implementation fee and the ongoing subscription. Be realistic about how much of the freed time will go to chargeable or otherwise valuable work, and plan to measure it afterwards, as set out in measuring return on AI.
A useful test for any quote from any provider: if the work goes badly, what do you still have, and what have you paid? On time and materials, you have paid for the time either way. With a fixed fee and a sign-off before go-live, you know the cost in advance and nothing goes live until your own team is satisfied with it.