IT consulting

Most of the consulting we are asked for is one of four questions: should we build or buy, is this architecture going to hold, is this supplier’s estimate reasonable, and where should we actually start with AI. Each of those is answerable in weeks rather than months, and each is worth answering before a much larger commitment is made. The engagement is fixed fee and the deliverable is yours — including when the recommendation is to do nothing, or to use somebody other than us.

• 01

  Fixed fee and a fixed date. You know the cost of the answer before you commission it.

• 02

  A written recommendation with the reasoning and the rejected options, so it survives the meeting it is presented in and can be handed to a board.

• 03

  Costed options rather than a single direction, including the option of doing nothing and what that costs over three years.

• 04

  Independent of the build. We will recommend a product, another supplier or your in-house team where that is the right answer, and we have.

• 05

  Delivered to your team, not only to the sponsor — the people who will execute it get the reasoning, not a summary.

1. 01

   Frame the question

   Half a day agreeing what decision this has to support and what would change it. A surprising number of consulting engagements answer a question nobody needed answered.

2. 02

   Evidence

   Interviews, code and architecture review, cost data and whatever the systems themselves can be made to tell us. We prefer measurement to opinion where measurement is available.

3. 03

   Options and costs

   Three or four genuine options, each with an estimate, a risk profile and a first step. Options that are obviously wrong are excluded rather than included to flatter the recommendation.

4. 04

   Recommend and hand over

   A written report, a session with the team who would execute it, and a first ninety days specific enough to start on Monday.

• Written report: findings, options, costs, risks and a clear recommendation.
• Architecture assessment with the specific things that will not scale, and when.
• Estimate ranges for each option, with the assumptions behind them stated.
• A ninety-day plan detailed enough to begin without further consulting.
• Presentation to the executive sponsor and a working session with the delivery team.
• Everything in an editable format — it is your document, and you will want to reuse it.

• 01Will you recommend yourselves for the build?

  Only where we are genuinely the right choice, and the report says what would make another option better. We have recommended a product instead of a build, another supplier with a specialism we lack, and in two cases doing nothing for a year. A consulting engagement that can only ever conclude in a build contract is a sales process.

• 02How long does an assessment take?

  A focused architecture or build-versus-buy question is two to three weeks. A broader technology strategy across an estate is four to six. Anything longer than that is usually a symptom of an unclear question, which is what the first half-day is for.

• 03Where should we start with AI?

  With whichever process is high volume, low judgement and already measured — because you will need a baseline to prove anything. The common mistake is starting with the most visible use case rather than the most measurable one, which produces a pilot nobody can evaluate and a programme that stalls at the second budget review.

• DGT

  Digital transformation

  Multi-year programmes that digitise a business and modernise what it already runs — sequenced so each stage pays for the next.

  /services/digital-transformation
• CLD

  Cloud and Azure solutions

  Cloud-native applications, migrations and the infrastructure underneath them — with the bill and the recovery plan designed in.

  /services/cloud-azure-solutions
• AI

  AI and generative AI solutions

  Assistants, chatbots, voice agents and generative features that are wired into your actual systems and measured against a baseline.

  /services/ai-and-generative-ai