Questions to Ask CIO in Roundtable
For attendees and moderators at a CIO roundtable, panel or peer dinner: twenty questions a technology leader can answer candidly in a room of peers, about deferred work, purchases they regret, spend ratios, incidents and where AI has actually shipped.
The questions
Open any question for the note
What is on your roadmap that you have quietly pushed to next year?
Why ask it
Deferred work describes real priorities better than the announced roadmap does, and it is safe to say out loud. It also gives the rest of the table permission to admit their own slippage.
What did you buy in the last two years that you would not buy again?
Why ask it
The most valuable question available in a peer room, and CIOs answer it far more freely than they discuss successes. Listen for whether the fault was the product or the rollout, because those are different warnings.
How much of your budget is run versus change, and is that ratio moving?
Why ask it
A number most CIOs know and rarely volunteer. It reframes every change-agenda claim made earlier in the session, since a function spending eighty percent on keeping the lights on has very little room to build anything new.
What is the last thing you decided to switch off?
Why ask it
Decommissioning is harder than launching and almost never discussed on panels. A CIO who cannot name a retirement is accumulating systems rather than managing an estate.
Where is your technical debt worst, and what does it stop you doing?
Why ask it
The second clause converts an engineering complaint into a business consequence, which is the form the rest of the room can actually use. Vague answers usually mean nobody has costed it.
Where has AI gone into production here, and what did it replace?
Why ask it
Cuts through pilot talk immediately. Ask for the workflow and the measured before and after; anything still described as an experiment after a year probably is not working.
What broke this year, and what changed afterwards?
Why ask it
Publicly discussable incidents are common ground at these events. Listen for whether the fix was technical, procedural or a person, since that reveals how the organization thinks about failure.
How are you deciding build versus buy now that the answer keeps moving?
Why ask it
Invites reasoning rather than a position, which is what makes it work in a group. The interesting part is usually the specific case where they went against their own default.
What have you accepted as residual risk rather than fixed?
Why ask it
Nobody should disclose vulnerabilities on a panel, and saying so when you ask keeps it comfortable. What you want is how they reason about risk they have consciously chosen to live with.
Which parts of the business genuinely trust IT, and which do not?
Why ask it
Gets past partnership language. Naming a function that does not trust IT, and why, is where the useful discussion in a roundtable usually starts.
What roles have you failed to hire, and what did you do instead?
Why ask it
Everyone in the room has this problem, so the answers are honest. The workaround matters more than the gap: contractors, managed service, retraining, or simply not doing the work.
How do you handle a team that wants to work on the new thing while the old thing still needs support?
Why ask it
A universal problem with no clean answer, which makes it good roundtable material. Watch for whether they have structural answers, such as rotation, or rely on goodwill.
What are you measured on by your CEO and board?
Why ask it
Reveals whether the role here is judged on uptime and cost or on revenue and product outcomes, which usually explains every other answer they have given.
What is everyone here excited about that you are deliberately avoiding?
Why ask it
Invites a contrarian position, which is what a peer discussion is for. The reasoning is the payload: lock-in, unit costs, missing skills, or a previous bad experience.
What advice did you take from a room like this one that turned out to be wrong?
Why ask it
Peer events manufacture consensus quickly and audit it almost never. Naming received wisdom that did not survive contact with their own estate is the most useful thing a panelist can offer.
How has your relationship with finance changed in the last two years?
Why ask it
Scrutiny of technology spend has tightened nearly everywhere, and this reaches it without asking anyone to criticize a named colleague. Listen for whether they now bring business cases or defend variances.
Which part of your job did you expect to have delegated by now and still have not?
Why ask it
Every technology leader has one thing they cannot hand over, usually vendor escalations or architecture arguments. The answer says as much about the bench beneath them as about their own habits.
How much of your estate could you describe accurately without asking anyone?
Why ask it
A question about the limits of their own visibility, which panels never put to them. Answers that begin with the word roughly are the honest ones, and they open a good discussion about asset data.
What have you learned to say no to that you used to accept?
Why ask it
Describes how a leader's authority actually changed over time rather than what their strategy says. It also hands the more junior attendees something concrete they can copy on Monday.
What do you wish people at these events would ask you?
Why ask it
A good closer that hands the floor back and tends to produce the thing they came to talk about but were never asked. It also gives the moderator a route into the next topic.
Asking Well in a Room of Peers
Practical guidance for the conversation itself
How to Ask in This Format
Ask one question, in under fifteen seconds
The most common failure at a roundtable is a two-minute preamble establishing the asker's credentials, followed by three questions stacked together. The panelist answers whichever is easiest. Keep it to a single sentence.
Ask what they can answer in public
Nothing unannounced, nothing about named vendors under contract, nothing about security specifics. Questions about ratios, retirements, hiring gaps and lessons from incidents are all safely answerable, which is why they get real answers.
Give the panelist somewhere to hide if they need it
Adding "without naming the vendor" or "at whatever level of detail you are comfortable with" makes a hard question askable. It also signals to the room that you are not hunting for a headline.
Follow up once, on a specific detail
The first answer is usually the prepared one. A single follow-up on a number or a date, rather than a new topic, is where roundtables produce their genuinely useful content.
Do not pitch
If you sell technology, any question that steers toward your category will be recognized instantly and will cost you the rest of the session. Ask something you actually want to know and introduce yourself afterwards.
If You Are Moderating
- Open with the deferred roadmap question rather than a vision question. It sets a candid register for the whole session, and the vision answers are the least useful part of any panel.
- Put the regretted purchase question to every panelist in turn. Nobody wants to be the one who claims a perfect record.
- When an answer arrives as a framework, ask for the case. "When did that last happen here" recovers more sessions than any other intervention.
- Redirect audience statements into questions within one sentence, or you will lose fifteen minutes.
- Keep the run-versus-change and measured-on questions in reserve. They are the two that reliably restart a flagging discussion.
- Close with what they wish people asked, and leave time for the answer rather than using it as a filler.
Common Pitfalls
Asking about strategy in the abstract
Questions about digital strategy and innovation produce fluent, interchangeable answers that could come from any panelist at any event. Anything with a number, a date or a specific incident in it will produce better material.
Asking a question that cannot be answered honestly in front of the CEO
Panels are often recorded or attended by colleagues and press. A question about internal conflict or a failed executive relationship puts the panelist in a bind and gets a careful non-answer.
Treating the roundtable as a one-to-one
Detailed questions about your own environment waste the room's time and rarely get useful answers without context. Save those for the coffee break, where they work much better.
Believing the AI answers
Reported adoption at these events runs well ahead of production reality, in both directions. Always ask what the process looked like before, who is still checking the output, and what happened to the headcount.