Steve Holloway
ADVISORY

Someone has to say
whether it's worth building.

Fractional and interim CTO work, and technology due diligence. The judgement comes from doing this inside a bank, as an adviser to CIOs, and on the operations board of a group. The tooling handles the analysis. It doesn't handle the thinking.

Book a conversation

Free, 30 minutes, no obligation.

steve@steveholloway.co.uk
FRACTIONAL AND INTERIM CTO

I sit on your leadership team and say the unpopular thing.

Across these engagements I've rebuilt a team, replaced a platform, killed a programme that should never have started, and told a board something it didn't want to hear.

I take post-acquisition integration and turnaround mandates.

What you're buying is the judgement about which of those your business needs — and the standing to make it stick.

Before you sign with a software supplier, I'll read their proposal or contract and tell you what it commits you to, what it leaves out, and what I'd ask them to change. You get an independent second opinion while you can still act on it.

I take no commission, referral fee or margin from any supplier I recommend. Nobody but you pays me for that advice.

An engagement can usually start within two weeks of agreeing it.

The first 90 days, as they usually run.

WEEKS 1 TO 6

I assess the people, systems, suppliers and risks, and agree a plan and priorities with your CEO and board.

FROM ABOUT DAY 45

Delivery of the first changes begins.

EVERY MONTH

I send the board a written report, and I attend board meetings with your IT lead.

Anything outside the agreed scope is a change. It goes ahead only once its scope, its effect on the timeline and its price have been agreed in writing by both sides, and no work on it starts before then.

TECHNOLOGY DUE DILIGENCE

I read the estate instead of sampling it.

Traditional diligence samples, because there's a fixed number of analyst days — so you look at a slice and infer the rest. I automate the document review, the architecture read and the data work, which changes how much of the estate fits inside the same window.

You don't get the report faster. You get a report that covers more of what you're buying.

What it covers.

Every review accounts for the same twelve areas:

Architecture and scalability · Infrastructure and cloud · Internal business systems · End-user computing and the workplace · Data, reporting and management information · Information security · Business continuity and disaster recovery · Technology organisation and delivery capability · Technical debt · Product and roadmap · Technology spend · Technology strategy and operating model

One to three of them are agreed with you before signing, from your investment thesis, as the critical areas to go deeper on.

Three depths, three standard windows.

RED FLAG

Every area screened, and the critical areas examined. You get an investment-committee deck and a spreadsheet of findings.Standard window: 10 working days, or 4 expedited.

STANDARD

The same depth, with a management report deck and a 100-day plan added.Standard window: 15 working days, or 7 expedited.

FULL

The critical areas examined in depth: a session with whoever holds the knowledge, and the evidence checked. A second set of areas is examined too. You get a report as well as both decks, and a read-out session.Standard window: 25 working days, or 12 expedited.

A window runs from when the review is ready to start: the data room open, the people I may contact named, and the questionnaire sent.

Who owns what is built, on exit.

Each company owns what we build for it once it has paid for it, and keeps it if the company is sold. Our general toolkit stays ours, and each company gets a permanent licence to it that goes with the build.

If a fund wants one build used across its portfolio, we agree that at the start: the fund owns the build and licenses it to each company, on terms that survive the sale of any one of them.

How a result is measured.

Before anything changes, I measure how the process performs. After the change, I measure it again, the same way. A result is the difference between the two, and nothing else counts as one.

WHAT STAYS HUMAN

I automate the analysis. I don't automate the judgement.

Software does the routine analysis, so the review covers more of what you're buying. A person makes every decision that matters: no agent can sign a contract, take on a client or close a project.

AN ILLUSTRATIVE EXAMPLE

What a diagnostic leaves you with.

This example is made up. It describes no client and claims no result. It shows what you would have in your hands at the end: what was found, the options ranked, and roughly what each would cost you.

THE QUESTIONOrders take three days to reach the warehouse. Is it the system, or the way we work?
WHAT WAS FOUNDMost of the delay is two steps where someone types an order again by hand. The order system can do both itself. Nobody switched it on.
OPTION 1 / RECOMMENDEDSwitch on the order system's own transfer to the warehouse, and retrain the two people who do it now. £5,000 to £10,000. About six weeks.
OPTION 2Connect the two systems with a new integration. £20,000 to £35,000. About three months. Worth it only if option 1 leaves a gap.
OPTION 3 / NOT RECOMMENDEDReplace the order system. £80,000 to £150,000, and most of a year. The system was never the cause of the delay.
WHAT YOU KEEPThe findings, the evidence behind each one and the reasoning for the ranking. They are yours, whoever you choose to do the work.

A straight conversation first.

Thirty minutes on the decision you're actually facing. No slides.

Book a conversation

Free, 30 minutes, no obligation.

steve@steveholloway.co.uk