Questions to Ask Consultants
Questions for the buyer's side of a consulting conversation, whether you are choosing between firms or scoping work with an independent adviser. They cover who will actually do the work, how fees and scope changes are handled, what you own at the end, and how the engagement can be stopped.
The questions
Open any question for the note
Who will actually do the work, and how much of it will you do yourself?
Why ask it
Pitch teams and delivery teams are often different people. Ask for names, seniority, and the share of hours each will bill, then put that staffing in the contract, because the gap between the partner who sold the work and the analyst who does it is the most common complaint about consulting.
What is the closest engagement to this one that you have run?
Why ask it
You want a specific client situation, not a category. Listen for whether the similarity is in the problem or only in the industry, and be wary if the closest example turns out to be a colleague's project they were not part of.
What happened at that client after you left?
Why ask it
This is the question that separates advice from results. Consultants who follow their clients can tell you what was implemented, what was abandoned, and why. Not knowing the outcome is a fair answer and also tells you their engagement ended at the presentation.
What have you already assumed about our situation that we should correct?
Why ask it
Every proposal rests on assumptions formed in two meetings. Surfacing them early costs an hour now instead of three weeks later, and a consultant who can state their assumptions crisply is likely to manage the project the same way.
How would you frame this problem differently from the way we have?
Why ask it
If they simply accept your framing, you are buying labor rather than judgment. A useful answer challenges the boundary of the question, and it is also your best early sign of whether they will tell you inconvenient things later.
What do you need from us, and how many hours of our people's time will that take?
Why ask it
The hidden cost of consulting is internal: interviews, data pulls, workshops, review cycles. Get an estimate in hours per week and by role, because a project that quietly needs half of your analyst's time is a staffing decision, not just a purchase.
How is the fee structured, and what triggers a change order?
Why ask it
Fixed price, time and materials, and retainer each fail differently: fixed price invites narrow scope, time and materials invites duration, retainers invite drift. Ask for the specific conditions under which the price changes, in writing, before you sign anything.
What is the rate for each person on the team, and how are junior hours billed?
Why ask it
A blended rate hides the mix, which is how a team can shift toward juniors without the invoice changing. Rates by name, plus a cap on the hours the most junior members can bill, keeps the staffing you were sold from eroding quietly.
Which expenses are billed on top of the fee?
Why ask it
Travel, data subscriptions, survey panels, software licenses, and administrative percentages are the usual additions. Ask for a cap and for receipts, since expense terms are where an otherwise clear proposal often turns out to be open-ended.
What is the first real deliverable, and when do we see it?
Why ask it
An early, concrete output within the first few weeks tells you whether the project is moving and gives you an exit point if it is not. Proposals where nothing lands until the final presentation give you no way to intervene while intervening is still cheap.
When this is finished, what will we be able to do ourselves, and what will still need you?
Why ask it
You are asking whether the engagement is designed to end. Look for training, documentation, and models your team can run. A consultant whose answer implies permanent dependence has told you their business model, and it may still be acceptable if you go in knowing.
Who owns the documents, models, and code when we are done?
Why ask it
Many firms license their frameworks and tools rather than transferring them, which can leave you unable to rerun an analysis your own data produced. Settle ownership and any license terms in the contract, particularly for anything you will depend on operationally.
Will any part of this be subcontracted or delivered from another office?
Why ask it
Subcontracting is common and not automatically a problem, but it changes who has access to your data, which time zones the work happens in, and who is accountable. You want it disclosed now rather than discovered from an unfamiliar name on a file.
Do you work with our competitors, and how do you handle that?
Why ask it
Sector experience usually implies competitor work, so a blanket denial is less credible than a described process. Ask about team separation, exclusivity for the engagement period, and what happens to what they learn from you.
What data and which people do you need access to in the first two weeks?
Why ask it
Access is the most frequent cause of delay, and the delay is usually charged to you. Getting the list up front lets you resolve permissions, legal review, and executive calendars before the clock starts rather than during it.
What do you do when the answer is something our leadership does not want to hear?
Why ask it
Ask for a specific occasion. Consultants who have delivered unwelcome findings can describe how they did it and what happened. Anyone who says it has never come up either has not looked hard or has learned to write around it.
Can you tell me about a project that went badly, and what you took from it?
Why ask it
A candid answer with their own role in it is a reasonable predictor of how they will behave when your project runs into trouble. An answer that blames the client's lack of commitment is a preview of the conversation you will have later.
At the halfway point, how will we know this is working, and what happens if it is not?
Why ask it
You are asking for a checkpoint with teeth: a named measure, a date, and a defined response. Without one, a struggling project usually continues to the end of the contract because stopping requires someone to admit it in public.
How do we end this early, and what does that cost?
Why ask it
Read the termination terms closely: notice period, work in progress, and any minimum commitment. A consultant comfortable with a clean exit is confident in the work, and knowing the number keeps a poor fit from becoming a full-length engagement.
Which clients can I speak to, including one who chose not to continue with you?
Why ask it
Reference lists are curated by definition. The second half of the request is the informative part: either you get a useful conversation about fit, or you learn how they handle a request they would rather refuse.
If we do nothing at all, what do you think happens?
Why ask it
The answer tells you how they size the problem, and whether they are willing to say the situation is tolerable. A consultant who describes doing nothing as catastrophic in every case is selling; one who can describe the real cost of delay is thinking.
Would you do a smaller first piece, so we can both test the fit?
Why ask it
A short, priced diagnostic gives you a working sample and gives them a chance to scope accurately. Firms geared to large engagements often decline, which is useful information, and the ones who agree usually produce a better proposal for the main work.
Buying consulting well
Practical guidance for the conversation itself
Before you brief anyone
Write the decision, not the topic
Consultants respond to what you hand them. A brief that says we need a growth strategy invites a broad, expensive project. A brief that says we must decide by September whether to keep the second facility open produces a tighter scope and lets you compare proposals against each other.
Name the internal owner first
Someone on your side needs the time and the authority to answer questions, unblock access, and accept deliverables. Engagements without a real owner drift, and the consultant's time gets spent chasing meetings you are paying for.
Decide what you would do with each possible answer
If a finding would not change anything, do not buy it. Working this through in advance also stops you from commissioning research to settle an internal argument that is really about authority rather than evidence.
Talk to more than one firm, even if you know who you want
Competing proposals surface different framings of your problem, which is free thinking, and they give you a reference point on price and staffing. Sole-sourcing tends to produce the scope the firm prefers to sell.
Terms worth pinning down in writing
- Named staffing with seniority and expected hours, plus a requirement that substitutions be approved.
- Rates by person rather than a blended rate, and a cap on total fees and on expenses.
- The specific conditions that trigger a change order, and who has to sign one.
- Deliverable dates, including at least one substantive output in the first few weeks.
- Ownership of documents, data, models, and code, and license terms for anything not transferred.
- Confidentiality and data handling, including whether your data leaves your systems and where it is stored.
- Termination for convenience: notice period, what is owed, and what you receive on the way out.
- Conflict of interest terms, including any exclusivity during the engagement.
What goes wrong
Hiring a consultant to avoid making a decision
When the real purpose of the engagement is to give a decision external cover, the work is judged by whether it supports the conclusion someone already reached. It is expensive, everyone involved can tell, and it wastes the one thing an outsider can genuinely offer.
Buying analysis when you needed implementation
Diagnostic work and delivery work require different skills and different people. Be explicit about which one you are purchasing, and if you want both, ask how the team changes at the handover and who stays.
Letting scope grow without a change order
Additional requests during a project feel small individually and reset the timeline collectively. Route every addition through the change process even when the consultant offers to absorb it, so the record of what was promised stays intact.
Accepting a blended rate and a nameless team
Without named staffing you have bought capacity rather than expertise, and the mix can shift toward juniors while the invoice looks the same. Ask who is on the project this month, every month.
No plan for the work after the final presentation
The most common fate of a consulting deliverable is a well-received presentation followed by nothing. Agree before the project ends who owns each recommendation, what the first three actions are, and when someone will check whether they happened.