What do these guides cover?
Each guide settles one practical question that comes up when a firm decides to do something useful with AI rather than buy another licence. Some are about deciding: which job to rebuild, what the lost hours cost, how a consultancy sets its fee. Others are about doing it well: how long it really takes, why rollouts stall, and how to measure the return once the new process is live. They are written for the person who has to sign off the decision and answer for it afterwards.
Why do AI rollouts fail?
The seven ways AI rollouts stall in professional services firms, the warning signs to look for in your own, and how to rescue one that has gone quiet.
How do you calculate unbillable hours?
A step-by-step method for estimating the unbillable hours your fee earners lose each week and their annual cost, with a worked example and the audit logic.
What should a professional services firm automate first?
A ranked shortlist of the jobs most professional services firms should rebuild first, why each earns its place, and the ones to leave until later.
How long does AI implementation take?
What decides how long it takes to put AI to work in a firm, where the weeks actually go, what slows projects down and what a 30-day process rebuild involves.
How do you choose a process to automate?
Six criteria for judging any process as a candidate for automation, a simple weighted scoring method, the disqualifiers to check, and a worked comparison.
How much does an AI consultant cost, and how is the fee set?
The pricing models AI consultancies use, what each means for the buyer, what drives the fee for a process rebuild, and the questions to ask before you sign.
How do you measure the return on AI?
How to measure whether AI has paid off: set a baseline, track the right measures, turn saved hours into value honestly, and avoid the figures that flatter.
Why do AI tools not improve productivity on their own?
Why firms buy Copilot or ChatGPT licences and nothing changes: the difference between a tool and a process, and what it takes to rebuild the job itself.
Can we use AI on client data under UK GDPR?
A plain guide to UK GDPR, the Data Protection Act 2018 and client confidentiality when your firm uses AI on client work. Not legal advice.
How do you get a team to adopt AI?
Why licences sit unused and what gets fee earners using AI every week: pick one job, change the process, name an owner and measure use.
How do you write an AI policy for a firm?
A practical AI policy structure for a 20 to 200 person UK firm: approved tools, data rules, client work, review, disclosure and ownership.
What questions should you ask an AI consultant?
Sixteen questions to put to any AI consultant before you sign, what a good answer sounds like, and the warning signs in a weak one.
How do you document a process before automating it?
A step by step method for mapping a job as it really runs: inputs, templates, handoffs, checks and exceptions, with a template you can copy.
How do you prepare templates for automation?
How to get your templates ready to be filled automatically: pick the master, mark the fields, separate fixed from variable text and set the rules.
How do you review AI output before it goes to a client?
A reviewer's checklist for AI drafted client work: facts, figures, sources, advice, confidentiality and tone, plus how to build review into the process.