Investment committees no longer accept a market overview, a competitive landscape slide, and five expert calls as a credible evidence base. Buyout deal value rose 44% to $904 billion in 2025, exit value climbed 47% to $717 billion, and with roughly 32,000 unsold companies worth $3.8 trillion sitting in PE portfolios, both bid-side and vendor due diligence are under more scrutiny than ever. At the same time, the expert network industry has grown at roughly 16% CAGR over the last decade, passed $2.1B in revenue in 2022, and is projected to keep compounding through 2030.
The bar has moved. The old CDD deliverable — a base-case growth estimate built on desk research and a handful of calls — has been replaced by what Bain calls full potential diligence: identifying specific, executable value creation levers, with every claim traceable to a primary source.
This guide is written for the four audiences Woozle works with every day: PE deal teams, hedge fund analysts, corporate strategy/M&A, and management consultants running CDD. It is opinionated, operator-grade, and designed to be applied to your next deal — not your next conference panel.
1. What primary research actually is
Primary research, in an investment context, is original, first-hand intelligence generated about a company, market, or thesis — as opposed to secondary research (filings, broker notes, databases, news) or management-supplied data. It has three core modalities:
- Expert interviews — 30–60 minute calls with former executives, customers, suppliers, competitors, channel partners, or regulators.
- B2B surveys — structured quantitative research deployed to defined respondent panels (e.g., 50 IT buyers, 200 SMB owners, 30 specialist physicians).
- Channel checks — targeted investigations of a specific link in the value chain (distributors, VARs, GPOs, end-buyers) to validate demand, pricing, inventory, or share dynamics.
An important distinction most buyers blur: an expert network is a sourcing layer, not a research output. Under the typical EN model, you are still responsible for scoping, writing the discussion guide, conducting interviews, and synthesising the result. A done-for-you (DFY) provider — the category Woozle operates in — owns the full process end-to-end and delivers a finished output.
2. When primary research matters most
Private equity deal teams
Primary research runs across the whole funnel — sourcing and screening, preliminary assessment, confirmatory diligence, and post-close monitoring — but it earns its keep during confirmatory diligence. Speaking with customers validates retention, switching behaviour, and willingness-to-pay. Competitor conversations reveal relative strengths, pricing dynamics, and emerging threats. Former insiders clarify cost structure and operational reality. Together, these inputs confirm or kill the assumptions inside the model.
Hedge fund analysts
Used to develop differentiated, non-consensus views ahead of earnings, product launches, or thesis catalysts. Common applications: channel checks on key revenue drivers (SaaS seat expansion, script trends, retail foot traffic), surveys to quantify churn or willingness-to-pay before public disclosure, and rapid expert calls to pressure-test sell-side narratives.
Corporate strategy and M&A
Acquisition target diligence, market-entry analysis, competitive intelligence, build-vs-buy. Corporate teams typically lack a dedicated diligence infrastructure and are the strongest candidates for DFY models.
Management consultants
A CDD report built on secondary research and desk analysis no longer wins mandates. PE clients expect a documented primary research programme underneath every conclusion: who you spoke to, how many, what they said, and how you triangulated.
3. How to design a primary research programme
Start from the thesis, not the expert
Decompose the investment thesis into 6–10 testable questions. Map each question to the expert archetype best placed to answer it. Sourcing before scoping is the single most common — and most expensive — mistake we see.
Stratify the panel
A defensible panel mixes:
- Former executives of the target and close comparables
- Current and lost customers
- Direct competitors and adjacent players
- Channel partners (distributors, VARs, GPOs)
- Regulators or industry body specialists where relevant
Most teams over-index on ex-executives because they are easy to source. Customers are where revenue truths live. If your panel is 80% former execs, your conclusions will be too.
Right-size the sample
Rough rules of thumb we use:
- Screening: 5–10 calls
- IC-grade CDD: 20–25+ targeted interviews
- Quantified claims (NPS, market share, switching intent): a B2B survey of 50–200 qualified respondents
An IC member can tell the difference between a CDD backed by five generic expert calls and one backed by 25 targeted interviews across customers, competitors, and specialists.
Discussion guide craft
- Open with low-stakes credentialling questions to confirm the expert is actually relevant.
- Front-load the one question you'd most regret not asking.
- Sequence: context → market → competitive dynamics → the target → forward-looking.
- Keep the spine consistent across calls so answers are comparable.
- Leave 15 minutes for follow-on threads — never script 60 minutes solid.
4. Running it well
Interviewing
- Let the expert anchor first — avoid leading questions.
- Ask for evidence, not opinion: "Why do you say that? What did you see?"
- Listen for what is not said; probe inconsistencies across calls.
- Always close with "Who else should we be speaking with?"
Triangulation and synthesis
- Build a Q×A grid: questions on rows, experts on columns, colour-coded by direction (supportive / contradictory / nuanced).
- Quantify qualitative signal — count concordance across N experts.
- Surface dissenting voices in the deliverable. Filtering them out is how IC members lose trust in the work.
- Link every claim in the final report to specific expert quotes. The evidence trail is the deliverable.
Compliance hygiene
The buy-side firm — not the network — carries ultimate MNPI and insider trading liability. Real compliance, not theatre, means:
- Pre-engagement screening against target and comparable employee lists
- Recorded compliance acknowledgement at the start of every call
- Cooling-off windows for former public-company employees (typically 6–12 months)
- No solicitation of MNPI, no trading-on-call
The biggest practical value driver in provider choice is how fast you can reach the right expert without taking compliance risk. A provider that is slow, or fast but off-target, is unusable.
5. Choosing your model: self-serve EN vs. full-service EN vs. done-for-you
| Dimension | Self-serve EN | Full-service EN | Done-for-you |
|---|---|---|---|
| Who runs the calls | You | You | Provider |
| Who writes the guide | You | You (light support) | Provider |
| Who synthesises | You | You | Provider |
| Pricing | Hourly per call | Hourly + retainer | Project-based |
| Best when | You have time and internal capacity | You need sourcing help on a live deal | Deliverable matters more than access; team is capacity-constrained; large samples needed |
The hidden cost of self-serve is analyst time — scheduling, prep, transcription, synthesis. On a fully loaded basis, DFY frequently wins for any programme above ~10 interviews or any project involving a B2B survey.
Most institutional investors (Blackstone, KKR, Apollo all follow this pattern) maintain relationships with 3–4 providers across categories: redundancy, competitive pricing, and backup access in niche sectors.
6. The vendor landscape (briefly)
Three generations of providers now coexist:
- First-generation networks (1998–2000s): GLG (1M+ experts, 19 offices, $550M+ 2019 revenue), Guidepoint, Coleman. Built the compliance framework that defines the industry.
- Second-generation (2006+): AlphaSights (founded 2008, London; fast, high-touch service to PE and consulting) and Third Bridge (founded 2007; differentiated through its Forum transcript library).
- Third-generation tech-led (2010s): Tegus/AlphaSense (merged; 260,000+ expert interviews in the library), NewtonX (survey-led), Dialectica, Techspert, Arbolus.
- Regional specialists: Capvision and SixDegrees in China; VisasQ in Japan (now with Coleman in the US); ProSapient and Atheneum in European consulting.
- Done-for-you: A structurally different model — Woozle's category — where the provider owns scoping, sourcing, interviewing, synthesis, and deliverable.
There are now 120+ expert networks globally (up from 50 in 2018), with a wave of boutiques breaking away from the majors since 2023.
7. What "good" looks like
An IC-grade primary research appendix contains:
- Panel composition (counts by archetype — customer / competitor / former exec / channel)
- The discussion guide(s) used
- Full transcripts
- A triangulation matrix mapping each thesis claim to supportive and contradictory signal
- An evidence trail linking every conclusion in the main deck to a specific source
- Both supportive AND threatening findings — not a curated narrative
If the IC member can audit any single claim back to a transcript line, the programme has done its job.
8. AI and the future of primary research
AI is reshaping the workflow in three places:
- Sourcing: AI-assisted matching is compressing time-to-first-expert to under 24 hours in most cases.
- Synthesis: Transcript-aware models can interrogate dozens of calls simultaneously to extract demand, inventory, pricing, and sentiment signals in minutes rather than weeks.
- Q&A grids: The Q×A triangulation matrix — the historical bottleneck — is increasingly auto-generated.
What AI will not replace: scoping the thesis, on-the-fly probing during a call, off-the-record context, and judgment about which expert to trust. The premium is moving toward bespoke sourcing and human synthesis. The first 30% of the work is commoditising fast.
9. Common mistakes (and how to avoid them)
- Starting calls before defining the thesis. Every call should test a specific assumption.
- Treating every call the same. Hypothesis-generation calls and validation calls require different guides.
- Skipping adversarial voices. Lost customers, churned employees, and direct competitors are where the deal-killing signal lives.
- Underinvesting in synthesis. Raw transcripts are not a deliverable.
- Trusting "professional experts." Operators doing 200+ calls a year have memorised the talking points. Fresh, off-circuit voices are worth more.
- Treating compliance as the network's problem. It is yours.
10. FAQ
How many expert calls do I actually need? For screening, 5–10. For IC-grade CDD, plan on 20–25+ targeted interviews, stratified across archetypes. For anything where you need to quantify a market or behaviour, add a B2B survey.
When should I use a survey instead of more calls? When the question is "how many" or "what proportion" rather than "why" or "how." NPS, churn drivers, willingness-to-pay, switching intent, and market share are survey questions.
How do I stay on the right side of MNPI? Pre-screen experts against target/comparable employee lists, enforce cooling-off windows, record compliance acknowledgements, and never solicit non-public specifics on revenue, pipeline, or M&A.
Self-serve EN or done-for-you? Self-serve if the deliverable is the call and you have analyst capacity. Done-for-you if the deliverable is a synthesised, IC-ready output, the sample is large, or the team is running the model in parallel.
Working with Woozle
Woozle Research is a done-for-you primary research provider for investment professionals. We don't sell access to experts by the hour — we run the entire research programme (scoping, sourcing, interviews, surveys, channel checks, synthesis) and deliver an IC-ready output. If you are scoping a CDD, building conviction on a public name, or evaluating an acquisition target and want a programme that holds up to IC scrutiny, get in touch with our team.
