Run B2B demand generation audience research: a practical guide

What to take away

Define the decision before the research

Audience research fails when it starts with a method. Someone suggests a survey, someone else wants interviews, and eight weeks later you have a deck nobody acts on. Start with the decision instead. Write one sentence that names the choice you face and the deadline attached to it.

A workable version looks like this: by the end of Q1 we must decide whether to keep selling to mid-market manufacturers or move up to enterprise accounts in the same sector. That sentence tells you what evidence matters and what does not.

Write a research brief that a supplier can price

A brief is not a wish list. It states the decision, the audience, the questions, the output and the deadline. Keep it to two pages. Anything longer tends to hide the absence of a decision.

Include the commercial constraint. If the sales team can only handle 20 new accounts a quarter, research that identifies 4,000 prospects is not useful. State the capacity limit so the research is scoped to what the business can actually serve.

Name the person who will act on the findings. If no single person owns the decision, the project will drift. Give that person the right to sign off the brief before fieldwork starts.

Agree what counts as a good answer

Before fieldwork, write down what would change your mind. If the answer comes back as X, we will shift budget to channel A. If it comes back as Y, we will not renew the agency.

This sounds mechanical, but it prevents the common outcome where research produces interesting context and no decision. It also protects the budget conversation later, because you can show what the spend was meant to settle.

A short exercise helps: ask each stakeholder to write their current assumption on a sticky note. Collect them. You now have a list of hypotheses to test rather than a vague brief.

Map the buying group and the data you already hold

The unit of B2B demand generation is the buying group, not the individual. A typical mid-market software purchase involves a budget holder, a technical evaluator, an end user and often a procurement or finance contact. Each has a different reason to say yes and a different reason to say no.

Map those roles for your three largest recent wins and your three most painful losses. The pattern usually shows which role you have been ignoring. For many suppliers it is finance, which explains why late-stage deals stall on payback period rather than features.

This mapping work sits alongside the sector-level view in B2B demand generation target audience in England, which covers how regional industry mix shapes who is worth targeting.

Segment on behaviour, not job titles alone

Job titles are a starting filter, not a segmentation. Two operations directors at similar-sized firms can behave completely differently depending on whether they are replacing a system or buying one for the first time.

Segment on what you can observe: pages viewed, content downloaded, events attended, previous purchases, support tickets raised. HubSpot's guidance on understanding and organising customer data is a reasonable starting point for structuring those fields so segmentation stays maintainable as the database grows.

Keep the number of segments small. Four to six is usually enough for a mid-sized B2B business. Beyond that, message variants multiply faster than the team can produce or measure them.

Audit first-party sources before buying anything

Most teams already hold more audience insight than they use. Pull together CRM records, closed-lost reasons, website analytics, webinar attendance, sales call notes and inbound enquiry logs. Even a modest business will have several thousand data points.

The gap is usually structure, not volume. Closed-lost reasons are free text, analytics is not joined to CRM, and call notes live in someone's inbox. Fixing that is dull work with a high return, and it is often the point at which a demand gen agency earns its fee by turning scattered records into a usable view of the market.

If you would rather buy the work in, the same brief applies. UK B2B agencies such as The Marketing Practice, Earnest and Stein IAS work in this market, and the scope they quote follows the decision you describe rather than the method you name.

Set a rule for data quality. Any record without a sector, company size band and last activity date is not usable for segmentation. Clean or discard it before you build anything on top.

Choose methods that fit the question

Different questions need different methods. A pricing question rarely suits a survey. A market-sizing question rarely suits six interviews. Match the method to the decision, then accept the trade-offs.

Use this sequence as a default for a B2B audience research programme in England. Start with an audit of the first-party data you already hold. Add desk research from public sources such as Companies House filings and ONS business statistics to size the sectors and regions involved. Run six to ten qualitative interviews across the buying group. Test the themes that emerge in a quantitative survey. Synthesise the evidence into a one-page decision memo with named owners and dates. Then convert that memo into CRM targeting rules and message changes before the next planning cycle.

The interview stage deserves care. Poorly run interviews produce polite agreement and nothing else, which is why the practical guidance in B2B demand generation customer interviews is worth reading before you book any calls.

Run interviews and surveys without leading the witness

In interviews, ask about past behaviour rather than future intent. What did you do the last time you evaluated a supplier like US is far more useful than would you buy this. People are unreliable narrators of their own future decisions.

Keep the sample honest. If you interview only your happiest customers, you will learn why people like you, not why the market buys. Include at least two people who chose a competitor.

Surveys work best for confirming patterns, not discovering them. Keep them under eight minutes, avoid double-barrelled questions, and never ask a question whose answer you cannot act on. The tooling comparison in B2B demand generation survey tools in England covers the practical differences between the main platforms. For most B2B programmes that shortlist comes down to SurveyMonkey or Typeform for speed and cost, SmartSurvey as the UK-based option, and Qualtrics where advanced analysis and panel management matter.

Handle tracking, consent and claims properly

B2B audience research still involves personal data. Website analytics, email engagement and CRM enrichment all fall under UK GDPR and the Data Protection Act 2018, and cookies, pixels and similar tracking technologies sit under the Privacy and Electronic Communications Regulations, known as PECR. The ICO's guidance on online tracking explains the consent and transparency expectations that apply when you use those technologies to follow individuals. PECR also governs email consent and the soft opt-in rules that most B2B teams run into, so check it before you add contacts to a nurture sequence.

Ethical standards matter beyond the legal minimum. The DMA code sets out the conduct expected of organisations running data-driven marketing, and it is a useful internal reference when a tactic is legal but feels intrusive.

If your research feeds advertising claims, the regulator's enforcement powers are real. The ASA's sanctions regime covers what happens when ads breach the CAP Code, including the online enforcement notices that can affect paid search visibility.

For teams building internal capability, formal training helps. The CIM's data and insights courses cover research design, analysis and interpretation for marketing teams that need to run this work in-house.

Turn findings into targeting, message and spend

Research that ends in a report has failed. The output should be a small set of changes to who you target, what you say and where the budget goes. Write them as decisions with owners and dates.

Translate each finding into a targeting rule. Not manufacturers are a good fit becomes manufacturers with 200 to 2,000 staff, an existing ERP system and a named operations lead. Specific rules can be loaded into your CRM and used by sales without interpretation.

Then translate findings into message. If the buying group's main objection is implementation risk, your content should address migration and downtime, not feature comparisons. This is where most demand generation underperforms: the targeting is fine and the message answers a question nobody asked.

Set measurement that matches the buying cycle

B2B cycles are long, so weekly lead counts tell you very little. Measure at the buying-group level and use time windows that match your actual sales cycle. If deals take five months, a 30-day attribution window will mislead you.

The wider market picture affects how you read results. Economic conditions, sector hiring and budget cycles all move pipeline independent of your activity, and the annual view in B2B demand generation trends and outlook is a useful reference when a quarter looks unusual.

Agree two or three measures and stick to them. Common choices are qualified buying groups created, progression rate between stages, and cost per qualified opportunity. Add a qualitative check: ask sales whether the inbound conversations match the audience you intended to reach.

Review quarterly and retire what stops working

Set a quarterly review with a fixed agenda: what changed in the audience, what we stopped, what we started, what we will test next. Keep the audience model as a living document rather than a one-off project deliverable.

Retire segments that no longer produce pipeline. Audiences move, particularly in sectors exposed to regulation or technology change. A segment that worked in 2024 may be saturated or consolidated by 2026.

Finally, keep a short record of what you tested and what you learned. Six months later, nobody remembers why a channel was dropped. A one-page log prevents the team from repeating the same experiment and drawing the same conclusion twice.

Common questions

How much should a small B2B team spend on audience research?

For a team on a modest budget, phase one should cost little beyond internal time. Most value sits in your existing CRM and call notes.

Commission external interviews and a survey, and expect a four-figure sum. A team paying £6,000 for ten interviews and a 200-response survey is within normal market range. This is illustrative, not a benchmark.

Can we run audience research without a dedicated researcher?

Yes, provided someone owns the decision and the analysis. Marketing managers with strong analytical habits run credible interview and survey programmes regularly. Where it breaks down is when nobody has protected time, so assign the work explicitly rather than adding it to an already full role.

How many interviews are enough?

Six to ten interviews with people who were involved in real purchase decisions usually reveal the main themes. Beyond that, you tend to hear repetition rather than new information. If you need to know how common a theme is, move to a survey rather than adding more interviews.

What is the most common mistake in B2B audience research?

Treating it as a one-off project. Audiences shift as sectors consolidate, budgets move and buying groups change composition. A model built once and filed away will be stale within a year, which is why the quarterly review matters more than the initial report.