Custom Research: one specific question, properly answered
Bring us the question nobody has published an answer to. We sharpen it, choose the method it calls for, investigate, and report what we found and what we could not establish.
In brief
Custom research investigates one specific question when no published report answers it. We sharpen the question with the person who will decide, choose the method the question needs, report findings as they arrive, and finish with the answer first, then the evidence, the confidence and the gaps.
- Who brings it
- A strategy, innovation or product lead holding a question that a decision depends on.
- What you provide
- A named decision maker, access to the right people, and patience with a question that changes shape.
- What you receive
- A question note, interim findings, and a written report that opens with the answer.
- Where it ends
- At what the evidence supports. Advice on what to do, and delivery, are scoped separately.
On this page
Why is a topic never an answerable question?
Research is sold in categories: a technology study, a market report, a regulatory review. A question that crosses these lines belongs to none, so it is split between providers who each answer the part they know, or it is answered with the nearest generic report. Nobody then owns the join, and the join is usually where the decision sits.
The question is also often the weakest part of the commission. It arrives as a topic, such as how AI affects the supply chain, which no research can finish. A topic produces a long document that informs everybody and decides nothing. The result is a decision taken on instinct, with a report on the shelf that was never quite about it.
Which method does the question call for?
The method follows from the kind of question, and we write it down before fieldwork with what it cannot show. A question that crosses lines is cut into parts, each part gets its own method, and the answer joins them.
| If the question is | It usually needs | It cannot show |
|---|---|---|
| Can this be done | Hands-on tests, documentation, accounts from people who have tried | Whether it holds up over a long period or at larger scale |
| What do buyers or users actually do | Conversations with them, public filings, behavior left in public records | How common the behavior is across a whole population |
| What is allowed, and what changes | The regulator's own texts, guidance and enforcement notices, read for the provision | How your counsel will read it; the general counsel's office decides that |
| What will it cost to run | Operators' accounts, vendor documentation, a cost model with stated assumptions | Your actual bills, which depend on choices not yet made |
| Something that crosses these lines | A cut into sub-questions, each with its own method, then joined | Anything the weakest sub-question could not support |
From the first conversation to the written answer
- 01
The question session
We sit with the person who will decide, and with whoever must act on the answer. We keep asking what they would do differently on each possible answer, until a question remains that evidence could settle. The artifact is a question note: the question, the decision behind it, and what would change the decision.
- 02
Method and its blind spots
For each part of the question we name the sources, the techniques and what they cannot reveal. The sponsor agrees this before fieldwork. If the question turns out to be more than one question, we say so here and agree the order.
- 03
Fieldwork with interim findings
We report as findings arrive, on a short call with the sponsor, so surprises reach the decision maker early. When a finding shows the question was framed wrongly, we change the scope in writing and carry on, instead of finishing the wrong job.
- 04
The challenge session
Before the report is final, the draft answer goes in front of the sponsor and the skeptics they name, and they try to break it. Whatever survives goes in the report, and the objections that did not get answered go in the gaps.
- 05
The report
A written report that opens with the answer, then the evidence, the method and the confidence on each finding, and closes with what we could not establish and what it would take to find out.
A pattern we are called into
A legal technology product lead and a question that cuts across departments
The product lead at a legal technology company is asked whether to build drafting assistance for law firms. Engineering says it can be done. Sales says firms are asking for it. The general counsel's office has not decided how professional duties on confidentiality and supervision apply. Each department has an opinion, and none has an answer.
- The question note splits the topic. It separates whether drafting can be made reliable on the firms' documents, who in a firm would buy and what they would stop paying for, and what supervision duties require of a tool.
- Each part gets its own method. Hands-on tests on masked sample documents, conversations with buyers at firms, and a reading of the professional rules and regulators' guidance.
- An interim finding reframes the commercial part. Buyers describe reviewing as the painful job, not drafting. The scope is changed in writing so that the buyer conversations test review assistance as well.
- The challenge session tests the answer. Engineering and sales both try to break it. Sales objects that some large accounts want drafting, and the report records that as a gap and a segment to examine.
Where it endsThe answer is to build review assistance first and drafting later, under stated conditions. The report is plain about the one reading of supervision duties that counsel still needs to settle, so the product lead does not carry it as an assumption.
A report that opens with the answer and closes with the gaps can be used in a meeting. One that opens with background cannot.
What the report looks like
The first page is the answer, in a paragraph that stands alone, followed by the confidence we place in it and the single most important condition on it. A leader who reads nothing else has what they need to act or to ask the right next question.
The body goes finding by finding. Each finding states its evidence, the method that produced it and how well it is supported. The source log behind it records every source and how it was checked, so a skeptical colleague can follow any claim back. Where sources disagree, the report shows the disagreement.
The last section is the gaps. It says what we could not establish, why, and what it would take: access to a particular person, a test that needs production data, a ruling not yet made. Often that section is the most useful next step.
Ownership and use of the report are set in the statement of work. Commissioned reports are confidential to you by default.
Commission a question, or look for an existing answer first
We say so when an existing report already answers the question, and point you to it. That costs us a commission and saves you the engagement.
Custom research fits when
- A decision depends on the answer and someone is named to make it
- The question crosses technical, commercial and regulatory lines, so no single provider covers it
- Published work answers a nearby question, not this one
- You can say what you need to know, even if you cannot yet say how to find out
Look elsewhere when
- A published report or a short call with an expert would settle it
- The question is a matter of values or preference, which evidence does not decide
- You need a quantified measure across a population, which belongs with a survey specialist
- You want recommendations on what to do, which is advisory work and scoped differently
You give up a quick, generic answer and spend your sponsor's attention on the question. You get an answer shaped to the decision, with the gaps shown.
What a commissioned answer cannot promise
Not every question has a finding. When the evidence does not support an answer, the report says that first and says what is missing, which lets a decision proceed on known ground. We do not guarantee a particular finding.
- The question must be one that evidence can answer. Where it is not, the first step is to reshape it with you.
- The work needs a named decision maker who takes part in sharpening the question and reviewing the findings.
- Some questions need people or data that the evidence trail cannot reach. The question note names them at the start.
- We read regulation and report what it says. Your counsel decides how it applies to you.
Published work that shows how we investigate
Commissioned reports are confidential, so the evidence of method is our own published research. The prioritization tool and the decision brief help a leader frame a question before commissioning anything.
What buyers ask before commissioning custom research
What if we only have a topic and not a question?
That is the usual starting point. We spend the first conversations turning the topic into a question that evidence can answer, and into an idea of what answer would change the decision. If it cannot be done, we tell you before research begins.
How is this different from Technology Research and Market Intelligence?
Those services cover technologies and markets. Custom Research covers whatever the question requires, including questions that cross both and add regulation, operations or people. We often borrow their methods, and we write the method fresh for each question.
Do we own the report?
Ownership and use are set in the statement of work. Commissioned reports are confidential to you by default, and nothing from them appears in our published work without your written agreement.
Can the question change during the work?
Yes, and it often should. Interim findings exist to expose a question that was framed wrongly, and we adjust the method and the scope in writing when that happens, so the final report answers the question that matters.
Related
Is there a question your decision depends on that nobody has answered?
Describe what you are weighing in a few lines. A senior practitioner replies with a suggested first conversation.