B2B research by company size: SMB vs mid-market vs enterprise
A finding that holds for a 20-person startup can be wrong for a 20,000-person enterprise. This guide explains how to segment B2B research by company size and reach each tier with the right respondents.
Segmenting B2B research by company size means studying SMB, mid-market, and enterprise buyers as distinct markets rather than averaging them together, because a finding that holds for a 20-person startup is often wrong for a 20,000-person enterprise. The practical challenge is defining the tiers clearly, designing one study that stays comparable across them, and recruiting respondents you can actually confirm belong to each tier. This guide covers where the lines sit, why each segment behaves differently, and how to reach the right buyers for each.
Company size is the firmographic variable that quietly breaks more B2B research than any other. Two buyers can share a job title, an industry, and a problem, yet make the purchase in completely different ways because one works at a scrappy startup and the other inside a global enterprise. If your research blends them, your averages describe a buyer who does not exist.
Where the lines sit, and why you must draw them
There is no universal legal definition of SMB, mid-market, and enterprise, which is exactly why you have to set your own thresholds before fielding. A common working definition uses headcount: SMB at roughly 1 to 100 employees, mid-market at roughly 100 to 1,000, and enterprise at 1,000 and up. Government frameworks vary widely, with the U.S. Small Business Administration using industry-specific size standards and the European Commission defining SMEs by headcount and turnover together.
Pick a definition, write it down, and apply it consistently. Many teams layer revenue and operational complexity on top of headcount, because a 200-person professional services firm and a 200-person deep-tech company have very different procurement realities. The point is not to find the one correct boundary. It is to classify every respondent the same way so your segments remain comparable this quarter and next year.
This is a specific application of the broader discipline of defining segments before you study them, which we cover in market segmentation in market research and in the practical walkthrough of how to run a customer segmentation study.
Why each tier is a different market
Company size changes who decides, how they decide, how long it takes, and what they care about. Here is how the three tiers typically diverge.
SMB. One or two people usually make the call, often the founder, owner, or a single functional lead. They buy fast, care about price and ease of setup, and have little patience for long sales or research processes. They are allergic to anything that wastes their time. The upside for researchers is speed. The risk is that their context, few seats and no dedicated procurement, looks nothing like enterprise reality.
Mid-market. A small buying committee forms. Procurement starts to appear, integration and scalability become real concerns, and the buyer weighs growth against the cost of switching later. Mid-market is the tier most often misclassified, because it borrows behaviors from both neighbors depending on the company’s trajectory.
Enterprise. Large buying committees, formal procurement, security reviews, legal involvement, and multi-month cycles. Gartner research has long found that a typical enterprise buying group involves six to ten decision-makers, each arriving with their own information. Price is rarely the first question. Risk, compliance, integration, and vendor stability dominate.
The table below summarizes the differences that should shape any study spanning more than one tier.
| Dimension | SMB (1 to 100) | Mid-market (100 to 1,000) | Enterprise (1,000+) |
|---|---|---|---|
| Who decides | 1 to 2 people, often the owner | Small committee, early procurement | 6 to 10+ stakeholders, formal procurement |
| Top priorities | Price, speed, ease of setup | Scalability, integration, growth | Risk, security, compliance, stability |
| Buying cycle | Days to weeks | Weeks to a few months | Months to over a year |
| Budget authority | Discretionary, informal | Departmental, approvals needed | Multi-level, finance and legal gated |
| Hardest part of research | Low patience for long studies | Ambiguous classification | Reaching and verifying real buyers |
Designing one study that holds across tiers
Most teams want to compare tiers, which means fielding a single study carefully rather than three disconnected ones. The structure that works has three parts.
A shared core of questions every respondent answers, so you can compare tiers on the same dimensions. Tier-specific branches where the buying process genuinely differs, for example a procurement and security section that enterprise respondents answer and SMB respondents skip. And quotas that guarantee each tier has enough respondents to analyze on its own.
The failure mode is fielding to whoever is easiest to reach, which skews heavily toward SMB respondents because they are cheaper and faster, then reporting a blended average. That average represents no real buyer and leads product and go-to-market teams astray. Set quotas up front and hold the line on them.
Pricing and packaging questions deserve special care across tiers, because willingness to pay and the meaning of a given price are wildly different for a 20-person team and a 20,000-person division. We go deep on that tension in pricing tier packaging for SMB vs enterprise.
Recruiting the right respondents for each tier
Segmentation only works if you can actually place each respondent in the right tier, and this is where most B2B research quietly fails. On open opt-in panels, respondents routinely misreport company size to qualify for a study and its incentive. A sample that looks like a clean enterprise read can be half mid-market in disguise.
The fix is to recruit from verified professional audiences where you can confirm employer, company size, role, and decision-making authority before anyone enters the study, then screen on firmographics as a gate rather than a trust-me question. Verification is what lets you field a true enterprise sample or a genuine SMB sample instead of a blurred mix that undermines every subgroup comparison.
Each tier carries its own recruiting reality:
- SMB respondents are relatively plentiful but impatient. Keep studies short and incentives fair, and expect higher drop-off on long instruments.
- Mid-market respondents are the hardest to classify. Verify company size directly rather than trusting self-report, since this is the tier where misclassification does the most damage.
- Enterprise respondents are the scarcest, most senior, and most expensive to reach. Plan extra recruiting time and budget, and lean on verification to confirm that a claimed VP at a large company really holds that role.
We cover the enterprise end of this in detail in how to recruit enterprise buyers for research and in a worked example, recruiting mid-market IT managers for enterprise security software studies. For the SMB and startup end, our roundup of the best user research platforms for startups and SMBs compares tools suited to lighter, faster work.
Sample sizes and how to read the tiers
If you want to compare tiers quantitatively, treat each as its own segment that must read reliably on its own. In practice that often means 150 to 300 verified respondents per tier for stable subgroup analysis, more if you plan to cut each tier further by industry or region. Undersized tiers produce swings that look like insight but are really sampling noise.
For qualitative work, 8 to 15 interviews per tier usually surfaces the dominant patterns. The purpose is depth, not statistics, so continue until new interviews stop producing new themes within each tier.
Because enterprise respondents are the limiting factor on both cost and timeline, design around them. It is often smarter to lock the enterprise sample first and scale SMB and mid-market to match, rather than overfilling the easy tiers and scrambling for enterprise at the end.
For the mechanics of running the quantitative half at volume without sacrificing data quality, see our companion guide to B2B quantitative survey research at scale with verified professionals. And when a size tier intersects with a specialized audience, such as engineers inside large organizations, combine this approach with the methods in B2B developer community and DevRel research.
Common mistakes to avoid
Leaving the thresholds implicit. If three analysts classify “mid-market” three different ways, your segments are fiction. Write the definition down and enforce it in the screener.
Blending tiers into one average. The most common and most damaging error. Always report tiers separately before any combined figure.
Over-recruiting the easy tiers. SMB respondents flood in because they are cheap and fast. Quotas prevent your enterprise read from being a rounding error.
Trusting self-reported company size. Without verification, firmographic screening is an honor system that respondents have every incentive to game. The Pew Research Center’s methods work is a good reminder of how much sampling and classification decisions shape results.
Using one incentive and one length for everyone. A fair incentive for a time-poor founder and for a senior enterprise executive are different numbers, and a study length that works for SMB will drive enterprise drop-off.
Putting it together
B2B research by company size is less about finding perfect boundaries and more about discipline: define the tiers explicitly, design a single study that stays comparable across them, set quotas so each tier can be read on its own, and recruit from verified audiences so every respondent is really in the tier you placed them in. Do that, and you replace one misleading average with three sharp, actionable reads, one for each market you actually serve.
Frequently asked questions
What counts as SMB, mid-market, and enterprise in B2B research?
There is no universal standard, but a common working definition uses headcount: SMB is roughly 1 to 100 employees, mid-market is roughly 100 to 1,000, and enterprise is 1,000 or more. Many teams also layer in annual revenue and IT or procurement complexity. The important thing is to define your own thresholds explicitly before fielding so every respondent is classified consistently and segments stay comparable over time.
Why segment B2B research by company size at all?
Because company size changes almost everything that matters: who makes the decision, how many people are involved, how long the cycle takes, what the budget looks like, and what risks the buyer weighs. A single blended average across tiers hides those differences and produces findings that are true on paper and wrong for every actual segment. Segmenting by size keeps insights actionable.
How does the buying process differ across company sizes?
In SMBs, one or two people often decide quickly and care most about price, speed, and ease of setup. In mid-market, a small committee forms, procurement starts to appear, and integration and scalability matter more. In enterprise, large buying committees, formal procurement, security reviews, and legal involvement stretch cycles to months or more. Research questions, screeners, and incentives should reflect these realities.
How do you recruit respondents from a specific company-size tier?
Recruit from verified professional audiences where you can confirm employer, company size, role, and decision-making authority, then screen on firmographics before anyone enters the study. Open opt-in panels make this hard because respondents misreport company size to qualify. Verification lets you field a clean enterprise sample or a true SMB sample instead of a blurred mix that quietly undermines your conclusions.
What sample sizes do you need per tier?
If you want to compare tiers quantitatively, treat each as its own segment with enough responses to read on its own, often 150 to 300 verified respondents per tier for reliable subgroup analysis. For qualitative work, 8 to 15 interviews per tier usually surfaces the dominant patterns. Enterprise respondents are the hardest and most expensive to reach, so plan recruiting time and budget accordingly.
Can one B2B study cover all three tiers?
Yes, and often it should, but only if you design it to. Use a shared core of questions so results are comparable, add tier-specific branches where the buying process genuinely differs, and set quotas so each tier has enough respondents to analyze separately. The failure mode is fielding to whoever is easiest to reach and then reporting a blended average that represents no real buyer.
Reach every company-size tier with verified buyers
CleverX connects B2B and research teams to more than 8 million verified professionals across 150+ countries, screenable by company size, role, industry, and decision-making authority. That means clean SMB, mid-market, and enterprise samples you can actually trust, with results in days rather than weeks. Book a demo with CleverX.