Three proposals arrive. All three name the same competencies, all three promise behavior change, and all three include a diagram with four quadrants. The differences that will actually matter are not on the page.
The disappointment rate suggests this is a widespread failure of selection rather than of supply. Reporting on a study of nearly 1,500 senior managers at 50 organizations, Harvard Business Review noted that three quarters were dissatisfied with their learning and development function. Only one in four said it was critical to achieving business outcomes.
Those buyers were not careless. They ran processes, compared credentials and checked references. The problem is that a standard selection process tests how well a firm presents, and presentation is the one capability you will never need from them again after the pitch.
Decide what you are buying first
There are three different purchases hiding under one phrase. Content, which is materials and facilitation against a model you have already chosen. Capability, which is a program designed around your business problem. And diagnosis, which is help working out what the problem is before anything is designed.
Most disappointment comes from buying the first and expecting the third. Write down which one you want before the first meeting. If you cannot, you want diagnosis, and you should pay for it separately rather than hoping it arrives free inside a program quote.
Test one: make them diagnose before they present
Give each firm the same short brief and ask them to come back with questions rather than a proposal. What you are watching for is the shape of the questions.
Weak providers ask about logistics: cohort size, dates, budget, platform. Strong ones ask what the senior team disagrees about, what happened to the last program, and who owns the business number. Then they ask what behavior the organization currently rewards that it says it does not want.
A firm that will not engage before a signed contract is telling you something accurate about how the work will run.
Test two: ask what they will refuse to do
This is the fastest question in the process. Ask each provider to describe a piece of work they turned down and why.
A real answer sounds specific and slightly awkward. A client who wanted a program to justify a decision already made. A cohort with no sponsor. A brief that was really a performance problem for one person. An answer that amounts to being flexible for every client means you are buying delivery capacity, which is fine, as long as you know that is what you bought.
The related question is what they would tell you not to do. A partner who cannot name anything in your own brief that they would drop is agreeing with you professionally rather than advising you.
Test three: find out who actually does the work
The people in the room at the pitch are frequently not the people who will run the program. Ask directly who will design it, who will facilitate each module, who will deliver any coaching hours, and whether any of them are subcontractors.
Subcontracting is normal and not a problem in itself. Undisclosed subcontracting is, because it means quality is being managed by someone you have not met.
Be similarly precise about what a coaching credential covers. In the 2023 ICF Global Coaching Study, 85 percent of coach practitioners held a credential from a coaching organization, up from 74 percent in 2019. And 93 percent offered services beyond coaching, most often consulting, training and facilitation. A credential confirms coaching training. It says nothing about whether the person has operated at the level of the leaders you are assigning them.
Test four: check what they own after the last session
Ask what happens in week four after the program ends, and who is responsible for it. Most programs fail in the months after delivery rather than during it, which is the whole argument in why leadership training does not stick.
The answer you want names a mechanism and an owner: a structured conversation between each participant and their manager, briefed by the provider, scheduled before the program starts. The answer you do not want is a platform, a nudge sequence, or an optional community.
If follow through is described as your responsibility, that may be reasonable. It should then be priced out of their fee and resourced by you, deliberately, rather than assumed by both sides and done by neither.
Test five: agree the measure before the contract
Ask each provider to propose two measures the business already tracks and would notice moving, and to name the risks to attribution. Then ask who owns each number today.
The quality signal here is honesty about causality. A provider who offers a clean return figure for leadership work is overselling, because isolating that effect in a live organization is genuinely difficult. A provider who proposes two imperfect measures and explains exactly why they are imperfect is someone you can work with when results are mixed.
Settle this before signature. After signature it becomes a negotiation about reporting, and nobody wins that one.
Test six: ask for a failure, then call that reference
Every reference list is curated. Ask instead for a client where the work did not go well, what the provider thinks caused it, and whether you may speak to them.
Some firms will decline on confidentiality grounds, which is legitimate. Fewer will be able to describe a failure in terms of their own contribution rather than the client’s. That distinction is the most useful signal in the whole process, because it tells you how they will behave in month four when something is not working.
Three ways this decision gets made badly
The pilot designed to succeed
A pilot run with twelve volunteers, a supportive sponsor and a strong facilitator will go well. It tells you nothing about the rollout, where participants are nominated, sponsors are busy and delivery is shared. Insist that a pilot includes at least one group that did not ask to be there.
The proposal written in your own words
Sophisticated firms mirror your brief back with better formatting. It reads as deep understanding and is actually an absence of a point of view. Check whether anything in the proposal contradicts what you said. If nothing does, nobody has thought about your situation yet.
Buying a person without saying so
Often the real decision is that you trust one individual. That is a reasonable basis for a purchase and a poor basis for a contract, unless their time is named and committed in writing. Otherwise the person who won the work appears at the kickoff and the close, and is absent in between.
The evidence to have in hand before you sign
Six things, and none of them are in a standard proposal. A written statement of the business problem in your own words, which they have challenged at least once. The named individuals who will design, facilitate and coach, with committed days. The four way split of the fee into design, delivery, assessment and individual support, so it can be compared to what a program of this kind normally costs.
Then: the two agreed measures with their named owners inside your organization. The follow through mechanism, with a date and an owner. And one reference conversation with a client where something went wrong.
If you have all six, the decision usually makes itself, and the remaining difference between finalists is genuinely a matter of fit. If you are missing three or more, you are choosing on presentation, which is how three quarters of senior managers ended up disappointed with a function they funded properly.
If you are building a shortlist and want the brief tested before it goes out, that is a short piece of work rather than a pitch. Our leadership development practice starts by reading the situation and framing the problem, before anything is designed.
Frequently Asked Questions (FAQs)
Should we run a formal RFP?
Run one if your procurement rules require it, and add a diagnosis stage in front of it. A standard RFP asks every firm to answer the same questions about your stated problem, which rewards whoever states it back most fluently. The diagnosis stage is where you find out whether anyone disagrees with the brief, and that is the information you actually need.
Do coaching credentials matter when choosing a firm?
They matter as a floor, not as a differentiator. Credentials are now held by the large majority of practitioners, so their presence confirms training and their absence is worth asking about. For senior work, the question that separates candidates is whether the individual has operated at or near the level of the people they will coach.
Is a big firm safer than a small one?
Larger firms offer continuity, bench depth and easier procurement. Smaller firms usually give you the senior people directly rather than at the pitch. The risk profiles differ rather than one being lower, so decide which risk you would rather manage: an unfamiliar facilitator in module three, or a single point of dependency.
How much should we pay for diagnosis?
Enough that it is real work with a deliverable, and separate enough that you can stop after it. Buying diagnosis as a standalone piece also lets you take the findings to a different provider for delivery. That is sometimes the right answer, and it is impossible once the two are bundled.
What is the single strongest signal in a pitch?
Whether they tell you something you did not want to hear before you have paid them anything. A firm willing to say that your sponsor is not engaged, or that the population is wrong, is showing you something. It is showing you exactly how they will behave when the program hits trouble.