Choose on credentials and rapport
Three firms present, all are credible, and the decision comes down to who felt like the better fit. The structural differences that will decide how the engagement goes were never on the agenda, because nobody asked.
The selection
How to choose a fractional CIO
Most selection processes evaluate the pitch. Three questions about the arrangement will sort the field faster, and they work on us as well as on anybody else.
The approach
Evaluate the arrangement, not the pitch
The instinct is to evaluate the person. Their background, their references, whether the conversation felt right. Those things matter, and they are also precisely what every firm in this market optimises for, which makes them weak discriminators. Nobody turns up to a first meeting with a bad story.
The arrangement is much harder to dress up. Who employs the person, what you own when it ends, and who does the work as opposed to who is selling it. Those have factual answers, they can be checked afterwards, and they determine more about how the engagement goes than any amount of rapport.
They are also uncomfortable to ask, which is most of their value. A firm's reaction to being asked a direct question about its own incentives tells you how it will react when you ask a direct question about a recommendation it has made. That is the behaviour you are actually buying.
Everything below is written to be used on this firm as much as on anybody else. A selection guide whose author passes every one of its own tests is an advert with a checklist typeface, and it deserves to be read as one.
The fork
Two ways this usually goes
Choose on incentives and ownership
Ask the three questions in the first call. Most of the field answers at least one of them badly, which turns a long comparison into a short one and leaves rapport to decide between candidates who have already passed.
The three questions
What separates firms fastest
Each of these has an answer we could give that should cost us the work. That is what makes them worth asking rather than merely worth publishing.
Do you sell, resell or take referral fees from anything you might recommend?
This is the one that decides whether their advice can be trusted on the largest decisions you will bring them, because most technology decisions end in a purchase from somebody.
A plain no, with the business model explained: fees come from you and nowhere else, and they can say so without qualification.
Partnership tiers, preferred vendors, or a hesitation followed by an explanation of why it is not really a conflict. The hesitation is the answer.
What do we own at the end, and can we use it if we never work with you again?
Leadership work produces artefacts: a roadmap, a risk register, an inventory, documentation. If those leave when the firm does, you rented a feeling of progress.
Everything, outright, in formats you can open without them. Ideally it lives in your systems from the first week rather than being handed over at the end.
It lives in their portal, their template, or their methodology, and access ends with the engagement.
Who specifically will do the work, as opposed to who is in this meeting?
The gap between the person who sells and the person who delivers is the single most common disappointment in advisory work, and it is entirely avoidable by asking.
A name, their background, and the fact that they are in the room now. In a small firm this is a short answer and it should be.
A team, a bench, a pod, or a promise to match you with the right person after signature.
Red flags
Things worth walking away from
Several of these are testable on any firm you are talking to, including this one. Ask, and see what happens.
- A roadmap arrives before anybody has looked at your systems. It was written for somebody else
- Findings are asserted rather than evidenced. Ask which export a claim came from and watch what happens
- Everything is urgent and everything is critical. A register where nothing is low priority has not been ranked
- The proposal cannot say what would make the engagement a success, in terms you could check
- They will not name a case where they told a client not to buy something
- The contract needs a long lock-in. Work that is worth renewing does not need to be compulsory
- Deliverables are described by their format rather than by the decision they let you make
- They agree with you quickly and often. You are buying an independent opinion, not a second voice for your own
The process
What a good selection looks like from your side
- Before you talk to anybody
Write down what has to be different
Two or three sentences about what is not working and what would count as fixed. It does not need to be precise, and it will change. Its job is to stop you being sold a scope somebody else defined.
- The first conversation
Ask the three questions early
Independence, ownership and who does the work. Ask them in the first call rather than the third, because they are the ones that eliminate firms and everything else is a matter of fit.
- The proposal
Look for what is excluded
Every proposal describes what is included. The useful reading is what is not: project work, vendor negotiation, incident support, travel, and what happens to the fee when any of those come up.
- Before you sign
Agree what you will hold in ninety days
Not a feeling of progress. Named artefacts, in your systems, that survive the relationship ending. If a firm cannot commit to that list, the engagement has no definition of done.
Questions
What people ask about choosing
How many firms should we talk to?
Two or three is usually enough, and more than that tends to produce a spreadsheet rather than a decision. The three separating questions eliminate most candidates in the first call, so the comparison you actually run is between the small number who answered all three well.
Should we ask for references?
Yes, and ask them a specific question rather than a general one. What did you hold at the end of the first ninety days, and did you still use it a year later. General reference calls produce general praise; that one produces an answer you can act on.
Does industry experience matter?
Less than people expect, with genuine exceptions where regulation shapes the estate, such as healthcare or financial services. Most of what this role does, budgets, vendors, risk, roadmaps, transfers well across industries, and a firm that only knows one sector often brings one answer to it.
What if we choose wrong?
This is what the notice period and the ownership question are for. A month-to-month arrangement where you already hold the documentation makes a wrong choice recoverable and cheap. A long lock-in where the work lives in their portal makes it expensive, which is the real reason both terms are worth settling before you sign.
Ask us the three questions
We would rather be evaluated on the arrangement than on the pitch. Thirty minutes, and bring the list.