B2B social selling customer interviews: seven questions that rebuild a buying episode

Good B2B social selling customer interviews reconstruct a recent buying episode. They do not ask whether someone "likes thought leadership" or would hypothetically reply to a seller. The seven questions below move from the account trigger to the commercial outcome while keeping observation separate from interpretation.

What to take away

Inclusion method and scope

This list was prepared on 5 September 2026 for research with people involved in purchases affecting English organisations or operating sites. A question was included only when it exposed a distinct decision point. It was justified by current government research guidance, a regulator's boundary or a first-party platform record.

Promotional prompts, personality questions and requests for confidential competitor information were excluded. The sequence is chronological, not a ranking.

Recruit actual or likely participants from defined account and role groups. The Government Digital Service recommends matching participants to the users and experiences being researched, while acknowledging recruitment bias. Record the invitation route, eligibility rule, refusals and incentive.

1. "What changed before you began looking for an answer?"

Ask for the event and date. A renewal, service failure or internal request may emerge, but do not supply examples until the participant has answered. This question distinguishes an active episode from a general opinion.

2. "What did you do first, and what happened next?"

The GDS guide to in-depth interviews recommends exploring behaviour, circumstances and problems. Let the participant describe colleagues, search, documents and social feeds in their own order. Do not introduce the seller's preferred platform.

3. "Which people became involved, and what did each need?"

Probe for actions rather than seniority labels. Ask who framed the problem, checked evidence, approved resources or raised an objection. If the account is registered as a company, the Companies House service can verify public entity and officer details afterwards, but it cannot confirm informal purchasing influence.

4. "Which claim or document changed your view?"

Request the source and what the participant concluded from it. A seller post, colleague's comment, tender specification and technical document are different evidence types. Preserve uncertainty if the item cannot be recovered. Do not convert remembered exposure into proof that a post caused the purchase.

5. "When did social activity help, distract or have no part?"

This wording allows a negative answer. LinkedIn's post analytics documentation defines impressions, reach, interactions and viewer demographics, and warns that several figures are estimates. An interview may explain an individual's use, but neither source alone establishes attribution.

6. "What made contact appropriate or unwelcome at that point?"

Ask what the participant expected and whether a research invitation felt separate from sales. The ICO explains that genuine market research is not direct marketing, but promotion or information gathered for later marketing can change that status. Keep the interview neutral and route legal application to a qualified reviewer.

7. "What happened after the decision, and what remains unresolved?"

Capture the selected route, no-decision outcome or later change. Ask what evidence was missing and which colleague now owns the issue. This checks whether the original problem was solved without asking the participant to praise the seller.

Consent, notes and analysis

Before the session, explain the purpose, recording, data use, sharing, retention and withdrawal. The GDS informed-consent guidance calls for a record of what each participant agreed to and a way to find their data if consent is withdrawn.

Use a participant code in notes, separate exact words from the researcher's summary, and never paste sensitive statements into a CRM.

After the round, group observations by buying stage, compare cases that reached different outcomes, and report the sample and English-location rule beside every finding. A handful of interviews can uncover a question worth testing, but cannot supply an England-wide percentage.

Before you act

Common questions

How should a researcher handle social activity during the interview?

Question five asks when social activity helped, distracted or had no part, which allows a negative answer. LinkedIn's post analytics documentation defines impressions, reach and interactions but warns several figures are estimates. An interview may explain an individual's use, but neither source alone establishes attribution.

What must be done before the interview session begins?

Explain the purpose, recording, data use, sharing, retention and withdrawal. The GDS informed-consent guidance calls for a record of what each participant agreed to and a way to find their data if consent is withdrawn. Use a participant code in notes afterwards.

Can the Companies House register confirm who influenced a purchase?

No. If the account is registered as a company, the Companies House service can verify public entity and officer details afterwards, but it cannot confirm informal purchasing influence. Question three probes for actions rather than seniority labels, asking who framed the problem, checked evidence, approved resources or raised an objection.