The Buyer's Guide to Sample Sources · MR Framework

Most buyers walk onto a farm looking for a pet.

There are four ways to source B2B participants, and each one treats people differently. This guide explains the four models in plain language, shows why so many studies end up with the wrong one, and gives you the questions that separate a well-run operation from a bad one.

Why buyers get it wrong

You were taught how to order food. Nobody taught you how to buy sample.

By the time you could order for yourself, you knew a drive-thru from a tasting menu. You learned it by eating, and nobody had to explain it twice. You would never walk into a drive-thru and ask for a sommelier.

B2B sample buying never got that education. Most buyers learn the market from sales decks and word of mouth. So they treat every supplier as interchangeable, assume a vendor that ran a fast consumer study well can deliver verified IT Directors the same way, and end up at the farm asking for a pet.

Drive-thruZooProgrammatic and river. Fast, cheap, high volume.
Chain restaurantFarmExchanges and aggregators. Consistent, system-run, at scale.
Neighborhood regularFamily petsProprietary panels. Known history, managed relationships.
Chef's tableSafariCustom recruitment. Bespoke, slower, built for your study.
The four models

How each kind of supplier treats its participants.

Every model is good at what it is built to do. The mistake is asking one to do another's job.

Where it breaks

Buying from the wrong part of the kingdom.

The mismatch runs in two directions, and both cost you.

The most common mistake

Going to the farm to buy a pet

Asking an aggregator or programmatic marketplace to reliably deliver verified decision-makers with real purchase authority, at farm prices and farm speed. The supplier says yes, the screener gets quietly widened to hit quota, and a large share of completes get cleaned out later.

The expensive mistake

Booking a safari to visit a petting zoo

Paying custom-recruitment rates and timelines for a directional read that a well-run panel could have filled. The study gets slower and costlier without a better answer.

Think it through

The fix is not always to trade up. Ask what decision the data will drive. If it is pricing, market sizing, or a go or no-go, buy accordingly. If it is a directional read, put the budget toward the study that needs it.

Quick comparison

What buyers should know about each model.

ModelWho sells accessRelationship depthQuality predictabilityTrade-offs
Family PetsPanel owner, direct relationshipHighKnown historyHighIf well runLimited scale; may need supplementing for rare audiences.
Farm AnimalsExchanges, brokers, aggregatorsMediumFunctionalMediumDepends on standardsGreater scale but mixed provenance and variable quality.
Zoo AnimalsProgrammatic platforms, routersLowTransactionalLowHighly variableFast fill but fraud risk, unknown history, professional participants.
SafariSpecialist recruiters, field agenciesHighAuthenticHighIf well runSlow, expensive, low scale, but reaches truly hard-to-find audiences.
Do your homework

Every part of the kingdom has bad operators.

Not every pet comes from a good home. Puppy mills sell pets too, and farms get shut down for neglect. The sample industry works the same way. The model a supplier claims tells you what they should be. Only your questions tell you what they are.

Pets

The puppy mill

Calls itself a proprietary panel but recruits cheaply, profiles thinly, and surveys the same members until they are professional participants.

Farm

The neglected farm

An aggregator that cannot tell you which sources fed your study, does not de-duplicate across them, and passes problems downstream.

Zoo

The roadside zoo

Programmatic traffic with no meaningful fraud, device, or attention checks. Fills fast because nothing is stopping it.

Safari

The unlicensed guide

A recruiter who promises rare executives but cannot show how identity and role are verified before the survey starts.

Alarm

Every supplier will tell you their quality is excellent. That is the sales pitch, not evidence. Do not buy B2B sample on a capabilities deck alone. Ask for specifics, and treat vague answers as answers.

Five questions to ask before you sign any bid

  1. Where exactly will my participants come from?Ask for the source mix by percentage: owned, partner, routed. "Our network" is not an answer.
  2. How were they recruited, and how are they rewarded?Recruitment channel and incentive design drive who shows up and why.
  3. How do you verify that a person holds the role they claim?Title self-report alone is a screener that leaks.
  4. What did you remove on your last comparable B2B study, and why?A supplier that tracks removals can answer in numbers. Compare against the GDQ benchmark.
  5. Who else are you routing my study to?Ask them to bid on their own resources unless they state otherwise.
Think it through

Vetting takes time and track record. Knowing which operators are well run is exactly what a named partner shortlist in Build It gives you, so you are not learning it one bad study at a time.

Know where your participants come from before you pay for them.

Frame It gives you the full partner evaluation framework. Build It gives you a vetted, named shortlist matched to your studies.