Market Sizing with Expert Interviews: A Consulting Methodology

Every market sizing deck ends with a number. The quality of that number is not determined by the databases your team pulled from, or the top-down versus bottom-up reconciliation you ran, or the growth assumptions you laddered onto trailing periods. It's determined by how many expert interviews you ran to triangulate the number, and how defensible those conversations are when the client pressure-tests the work.

Most consulting market sizing projects know this and under-run the interview layer anyway. Why? Because primary research is the part of the methodology that doesn't scale with desk hours. Your Insights Associate can pull every syndicated report in a day. They cannot run 15 expert calls in a day. And yet the defensibility of the final number lives in those 15 calls.

This post is a methodology for running the expert interview layer of a market sizing study. It covers the four expert types you need, what each can actually tell you, how to triangulate their inputs into a number that holds up to client pushback, and how to make the execution math work when your Insights function is already at capacity.

Why expert interviews matter more than desk research for sizing

A typical market sizing study has three layers:

1. Desk research. Syndicated reports, analyst data, company filings, trade publications. This produces the starting-point numbers — but every consultant has the same numbers, and they're almost always wrong about current conditions because the underlying surveys are 12–24 months stale.
2. Client-provided data. Sales data, customer lists, internal segmentation. This sharpens the picture for your client's slice of the market but tells you nothing about the rest of it.
3. Primary research. Expert interviews that validate the desk research, correct stale assumptions, and surface segments the syndicated reports missed. This is where your deck becomes defensible.

The final number in the deck is almost never from layer 1 or layer 2. It's from triangulating across all three, and the triangulation is only credible if layer 3 is real. Clients don't ask "where did you get this number" to be polite — they ask it because they're about to make a capital allocation decision against it.

(For broader context on how strategy firms are integrating AI into primary research workflows, see our playbook for consulting Insights Centers.)

The four expert types that matter for market sizing

Market sizing interviews are structurally different from commercial due diligence interviews. You're not asking about a target company; you're asking about a category. The four expert types below each answer a different slice of the sizing question.

1. Customer-side experts (across segments)

Who they are: Decision-makers at companies that buy the category. Procurement leads, department heads, or senior operators who make the buying decision or sign off on budgets.

What they tell you about the market:

• What percentage of their category budget goes to the type of product you're sizing
• How they segment the market mentally (not how the syndicated reports segment it)
• Which vendors they actually consider, and which ones don't make their shortlist
• How the budget has moved over the past 2 years — growing, flat, or shifting to adjacent categories

How many to run: 8–12 interviews across 3–5 customer segments. The goal is representation, not volume.

The question that makes the interview useful:

"When you think about your annual spend on [category], what percentage goes to [type of vendor], and what's that absolute number roughly? How has that moved over the past 2 years?"

That question, asked to 10 customers across segments, gives you the raw inputs for a bottom-up number that's anchored in real spending behavior instead of analyst extrapolation.

2. Distributors and channel partners

Who they are: The intermediaries who see the flow of product between vendors and end customers. Distributors, resellers, channel managers at platform companies, or partners who aggregate vendor catalogs for buyers.

What they tell you about the market:

• Actual unit volumes (what's moving through the channel this quarter vs. last quarter)
• Which vendors are gaining or losing shelf/listing share
• Emerging categories that don't yet show up in syndicated reports
• Price realization across the channel (list price vs. actual transaction price)

How many to run: 3–6 interviews, depending on channel complexity.

Why they're under-used: Distributors are often excluded from market sizing research because the team doesn't think of them as "experts." They are — they're just on the supply side of the transaction rather than the demand side. A good distributor interview can anchor your TAM calibration faster than three analyst reports.

3. Former vendor executives and product leaders

Who they are: People who used to run product, sales, or strategy at companies competing in the category. Former VPs of product, former heads of sales, former category GMs.

What they tell you about the market:

• What their former employer's internal forecasts were (often sharply different from public guidance)
• What the real competitive dynamics are, beyond what annual reports admit
• Where the category is heading — which segments are growing and which are dying
• Which adjacent categories will reshape demand over the next 2–3 years

How many to run: 3–5 interviews.

The unique value: Former executives can tell you things current ones can't — strategic bets that were killed, customer cohorts that underperformed, acquisitions that didn't work. This is where you find the hidden dynamics that change your sizing by 20%.

4. Industry analysts and trade association staff

Who they are: People whose full-time job is tracking the category from the outside. Independent analysts (not the brand-name research firms), trade association research directors, niche publication editors.

What they tell you about the market:

• Historical data you can't find in syndicated reports
• Regulatory changes that will reshape demand curves
• M&A activity across the category and what it implies for consolidation
• Adjacent categories that are converging into yours

How many to run: 2–4 interviews.

The trap to avoid: Big-name research firms (Gartner, IDC, Forrester) are sources of the numbers you're trying to triangulate, not independent validators of them. Their interviews will just confirm their published reports. Find smaller, specialized analysts whose incentive isn't to defend their own numbers.

How to triangulate into a defensible sizing number

Once you've run 15–25 expert interviews across the four types, you have a stack of inputs. The question is how to turn them into the single number that goes on page 8 of the deck.

The triangulation pattern that works:

1. Start with the bottom-up customer-side number. From your 8–12 customer interviews, compute a weighted average spend per segment, multiply by segment size, and sum across segments. This is your primary TAM estimate.

2. Cross-check with distributor volume data. Take the unit volumes distributors reported, multiply by realized price from your customer interviews, and compare to the bottom-up number. If they're within 15%, your bottom-up is defensible. If they're >25% apart, something is wrong — usually your segment sizing or your price realization assumption.

3. Use former executive interviews to pressure-test growth rates. Your desk research will give you one growth number. Customer interviews will give you a different one. Former executives will give you a third — usually the lowest, because they know what internal forecasts looked like vs. public guidance. The growth rate in the deck should be the one you can explain using all three inputs, not a best-case average.

4. Use analyst interviews to frame category boundaries. Analysts are best at telling you what to exclude from the market you're sizing — what's adjacent but not actually in the category, what's a different buyer, what's a different use case. This tightens your number by removing over-counting.

5. Run your final number back past 2–3 customers and 1 analyst. The last step is a sanity check: does the number feel right to someone inside the category? If five independent experts all say your number feels 20% too high, it probably is.

This 5-step triangulation is the difference between a sizing deck that the client pushes back on and one they build their strategy around. It's all human analysis work. But it's only possible if the underlying 15–25 expert interviews actually got run — and that's the execution math problem.

The execution math problem (again)

Market sizing studies have the same capacity problem as commercial DD, with a slightly different shape:

• Interview target: 15–25 calls across four expert types
• Study window: 3–6 weeks, depending on scope
• Insights Associate capacity: 3–5 interviews per week including scheduling, running, and synthesis
• Associates on the study: 1, maybe 2

The math says an Insights Associate can realistically run 12–18 interviews over 4 weeks of elapsed time — just enough to hit the lower bound of the target. For studies with ambitious scope, headcount blocks the interview plan before the study even starts.

The traditional outcomes are familiar:

• Cut the interview count to what the team can run. The deck ships with 8 interviews behind it instead of 20, and the triangulation layer is thinner than the partner wanted.
• Pay an expert network. For sizing studies, per-hour expert-network credits add up even faster than for CDD because the interview counts are higher and the experts are less directly "on-point" so you need more of them to cover the same ground.
• Extend the study timeline. Not usually an option — market sizing engagements have fixed client deadlines.

A concrete sizing study interview timeline

Here's what a 4-week market sizing study looks like with AI-moderated interviews running in parallel:

Week 1, Days 1–3: Scoping and sourcing.

• Lock the four expert-type target lists with the case team
• Draft one script template per expert type (customer, distributor, former exec, analyst)
• Define the triangulation inputs you need from each type
• Begin expert sourcing across all four types

Week 1, Day 4 – Week 2, Day 3: First interview wave.

• Run 8–10 customer-side interviews (parallel)
• Run 3–4 distributor interviews
• Transcripts land within hours of each call
• Insights Associate reads transcripts and tags inputs against the triangulation framework

Week 2, Day 4 – Week 3, Day 2: Second wave + early triangulation.

• Run 3–5 former-executive interviews
• Run 2–3 analyst interviews
• Begin bottom-up vs. channel triangulation on the numbers that have landed so far
• Identify the 2–3 questions that weren't answered and need gap-fill interviews

Week 3, Days 3–5: Gap-fill and synthesis.

• Run 2–4 targeted gap-fill interviews on unresolved questions
• Complete the 5-step triangulation
• Pressure-test the final number with 1–2 customers and 1 analyst

Week 4: Deck write-up.

• Associate writes the sizing section of the deck, referencing specific interview inputs at each triangulation step
• Partner review and refinement
• Ship

The difference from the traditional workflow is that the interview-running step happens in parallel across the four expert types instead of serially across one associate's calendar. Your Insights Associate's time moves from running calls to reading transcripts, triangulating inputs, and writing findings — which is the work they're actually supposed to be doing.

What NOT to automate in a sizing study

Three parts of the market sizing workflow should not go through AI:

Client-facing hypothesis setting interviews. Every sizing study has 1–2 conversations with the client's own leadership to understand what sizing question they're actually trying to answer. Those need a human with the relationship.

Sensitive segmentation interviews with competitors. If part of your methodology involves talking to competitors (directly or through former executives who recently moved), the framing and rapport matter more than the raw data extraction. Humans run those.

The triangulation and synthesis. AI produces transcripts. Humans triangulate 25 interviews into a defensible number. That's the work that justifies the consulting fee.

Getting started

If you have a market sizing engagement starting soon and want to try this:

1. Pick an upcoming study. Not one in mid-flight — give yourself a week of setup and script drafting.
2. Commit to the full interview target. Don't pre-cut for capacity; target 15–25 and plan backward.
3. Write the four script templates with the triangulation framework in mind. Each question should map to a specific triangulation input.
4. Run the first wave in parallel. Customer and distributor interviews can go simultaneously; you don't need them sequenced.
5. Triangulate as you go. Don't wait until all 25 interviews are done before you start comparing inputs. Early triangulation tells you where the gap-fill interviews need to focus.

Market sizing is one of the clearest examples of where primary research changes the quality of a deliverable that's otherwise indistinguishable from every competitor's version of the same study. The triangulation layer is the difference. The interview count is what makes the triangulation layer real.

---

InsightAgent runs AI-moderated expert interviews for strategy consultancies, research firms, and direct investors. See how the consultancies workflow fits together.