What structured customer interviews mean for professional services marketing

What to take away

The Data Protection Act 2018 sets the rules for storing interview recordings and notes, so decide consent and retention before the first call.
The ICO's direct marketing advice generator helps you check whether follow-up contact is lawful, because a research call is not a marketing permission.
Six to twelve clients across two or three buying roles usually beat a long list of loosely relevant names.
Capture the language buyers already use, then feed that wording into buyer personas.
Ad targeting only improves if interview findings change which job titles and sectors you exclude.

Start with a written inclusion rule

Before booking anyone, write one sentence that says who qualifies: usually a client or lost prospect who made or influenced a buying decision in the last 18 months.

Testable inclusion rule

Name the buying decision
Set an 18-month date window
Name the role in buying group
Colleague can apply without asking
Exclude complaint or debrief contacts

Keep the rule narrow enough that a colleague could apply it without asking you. A testable rule names the decision, the date window and the role, for example a buyer who approved a retainer in the past year. A vague rule produces interviews that cannot be compared.

If you serve England and Wales, interview buyers from England and Wales rather than a general mix of international contacts. Exclude anyone whose feedback already sits in a complaint or a tender debrief.

Pick roles, not job titles

The unit of analysis is the role in the buying group, not seniority. A managing partner, an in-house counsel and a finance lead describe the same matter differently. Cover two or three roles at least twice each so one strong personality does not set the tone.

Role beats job title

Recruit by role

Unit of analysis: Role in buying group
Coverage: Two or three roles twice
Veto holders: Procurement, finance, risk
Result: Explains why deals slowed

Recruit by contact list

Unit of analysis: Main contact only
Coverage: One strong personality
Veto holders: Missed entirely
Result: Misses why deals slowed

Recruit from the buying group, not the contact list. Procurement, finance and risk often hold a veto that never appears in the pitch, so a sample drawn from your main contact will miss why a deal slowed.

The wider sampling method is set out in the professional services marketing audience research guide, which explains how interviews sit alongside enquiry data and win-loss records.

Write questions that survive a transcript

Six to eight open questions is enough for a 45 minute call. Ask about the last decision rather than opinions of your firm. "Walk me through how the work was approved" produces usable material; "how do you rate our service" produces politeness.

Pilot the guide once before the first live call. Cut any question that invites a yes or no answer, then keep the same order for every participant so themes line up across transcripts.

Do not ask buyers to design your service. You want their words, constraints and approval steps.

Sort the data protection position first

Recordings and notes are personal data. Under the Data Protection Act 2018 you need a lawful basis, a retention period and a route for withdrawal. Say how long you will keep the recording and who will hear it.

If you plan to contact participants again with marketing, check that separately. The ICO's direct marketing advice generator walks through the conditions, including consent and soft opt-in.

Use the findings to rewrite, not to decorate

Three outputs justify the effort: a language list of the nouns buyers use, a friction list of the steps where decisions stall, and a proof list of the evidence buyers want before signing.

Those outputs are the raw material for professional services marketing buyer personas, because personas built from interview language survive a real sales conversation.

Close the loop with paid targeting

Interview findings should change targeting settings, not just a slide deck. If buyers consistently sit in one sector and one role, that is an exclusion decision as much as an inclusion one. The tool for testing it is LinkedIn Campaign Manager, where job function and seniority filters can match what the interviews showed.

Decide how to handle each situation

Choose

Positioning language: Six to eight recorded interviews with written consent
A client asks to be quoted: Written approval of exact wording and context
A participant withdraws: Delete the recording and log the date
Re-contacting buyers later: Check marketing conditions first
Findings conflict with belief: Report the conflict and sample size

Avoid

Positioning language: A firm-wide survey with no follow-up
A client asks to be quoted: Assuming a quote is publishable
A participant withdraws: Keeping notes because they are coded
Re-contacting buyers later: Treating research consent as marketing consent
Findings conflict with belief: Averaging answers into a safe summary

Common questions

How many interviews are enough?

For a single service line, six to twelve interviews across two or three buying roles usually reveals repetition. Stop when two consecutive calls add no new themes.

Do I need consent to record a call?

Yes. Tell the participant what you will record, how long you will keep it and who will hear it, then note the consent in writing.

Can I use interview quotes in proposals?

Only with permission for that specific use. Ask separately about attribution: many buyers allow an anonymised quote but not a named one.

Should interviews replace win-loss analysis?

No. Interviews explain the reasoning behind a decision; win-loss records show the pattern across many. Run both and compare where they disagree.