Consultancy

Advice we have to live with

The decisions that are expensive to get wrong. Which systems to keep, how to secure what you already have, whether to build or buy, and where AI genuinely saves time rather than quietly adding risk. We build and run what we recommend, so we cannot hand you a strategy and walk away from it.

See where we help →

No product to push

The recommendation is allowed to be leave it alone, and often that is the right one.

From people who run things

Advice from the team that hosts it, patches it and gets the call when it breaks.

Written down and priced

You keep the findings and the plan, whether we deliver any of it or not.

The problem

Almost everything is possible now. That is the problem.

A few years ago most technology decisions were limited by what could realistically be built. That constraint has largely gone. AI can draft, summarise, classify and answer, the demos are genuinely impressive, and a great deal of money is now being spent on things nobody actually asked for.

So the useful questions have moved. Not can we, but should we, at what cost, with whose data, and what happens on the morning it gets something wrong. That is most of what we get asked, usually by someone who has been sold a solution and wants a second opinion before they sign.

We answer as though it were our own money, because on Road XS, Pinga and Passport Proven it is. Every call we make here is one we have already had to make for ourselves, and then live with.

Asked should we put an AI chatbot on our website

Only once the answers exist somewhere it can read. A chatbot on a thin website will invent things confidently, and you own whatever it says. Fix the content first and the bot becomes a genuine shortcut.

Verdict · Not yet

The kind of question we answer most weeks

Capability

Four disciplines, by one team.

Technology strategy

Where the money should go over the next two years, and what to stop paying for. We look at what you run, what it costs, what it is quietly holding together, and whether the roadmap you have been sold matches the business you actually are. Build, buy or leave alone, with the reasoning written down so you can argue with it.

Technology roadmapsBuild or buy reviewsSystems and integrationSupplier and contract reviewBudget planningSecond opinions

Cyber security and compliance

The unglamorous work that decides whether a bad morning turns into a bad year. Access, backups that are tested rather than assumed, patching on a schedule, and the paperwork that proves it happened. We start with the basics almost everyone is missing, long before anyone sells you a product with a dashboard. For obvious reasons we do not publish our full methodology.

Cyber EssentialsRisk assessmentWebsite and application securityAccess and identityBackup and recoveryIncident response

AI adoption

Where AI earns its place, and where it adds risk you will not see until later. In practice that means a handful of specific tasks, tools chosen on more than a demo, a policy people can actually follow, and enough training that the people doing the work know what to trust. Three things automated properly beats a transformation nobody uses.

AI readiness reviewUse case selectionCopilot and assistantsData and privacyAI policyPilots and rollout

Governance and training

Whatever we recommend has to survive contact with the people who will use it. Documentation somebody will genuinely read, training that is not a slide deck, and a clear answer to who owns each system once the person who set it up has left. This is the part that decides whether any of the rest holds.

Staff trainingPolicy and documentationData protection and GDPRSupplier managementBusiness continuityHandover and ownership

Context / S&P Global Market Intelligence, 2025

42%

of companies abandoned most of their AI initiatives in 2025, up from 17% the year before. On average, 46% of proofs of concept never made it into production at all. Very little of that was the technology failing. It was the wrong problem picked, data that was not ready, or nobody owning the thing once the pilot finished. Those are all decisions taken before a line of code is written, which is the conversation we would rather have with you first.

Talk it through with us →

How it runs

A conversation, not a slide deck.

Listen

What you are trying to achieve this year, and what is currently in the way of it.

Assess

What you run, what it costs, where the risk sits, and what depends on one person remembering something.

Recommend

A written plan with priorities and prices, including the option to do nothing. Yours to keep either way.

Support

We help you deliver it, or brief whoever does. Advice that stops at the document rarely survives.

Revisit

Technology moves. We come back, look at what actually happened, and change the plan rather than defend it.

Straight answers

Things we will probably talk you out of.

We have to live with the advice we give, because we also build, host and support what we recommend.

An AI project without a problem
Starting with the technology and hunting for a use case burns a budget and a lot of goodwill. Start with a task people already dislike doing.
Replacing a system people simply need training on
A surprising number of failed systems are perfectly good systems that nobody was ever properly shown how to use.
Buying security products before the basics
Multi factor authentication, tested backups and knowing exactly who has access will stop more than a tool nobody has time to watch.
A strategy document nobody will read
If it does not change what happens next month, it was an expensive way to feel organised.
A long contract for a discount
Three years is a long time in technology, and the saving rarely covers the cost of being stuck with the wrong thing.

Insights

Technology moves fast. We write down what actually matters.

Read everything →