What customers say. What customers do.
How attitudinal and behavioural evidence can be combined, and what to investigate when the two disagree.
For: customer insight, CX, product and analytics teams.
Speaking & media
A sharper discussion.
A practical next question.
Talks, panels and podcast conversations for research, analytics and leadership audiences.
I connect research practice with the organisational decisions around it. These proposed topics can be adapted to your audience, their questions and the format of your event.
Topics to discuss
Built around a decision
your audience recognises.
Why a good insight doesn't get used.
Decision ownership, incentives and the design of the insight function. A discussion about what happens after the read-out.
For: research leaders, senior managers and cross-functional teams.
Evidence for learning, or evidence for compliance?
What my doctoral literature review suggests about commissioning and analytical capacity, with the limits of that argument made explicit.
For: researchers, commissioners and third sector leaders.
AI in research: what would count as good enough?
Research quality, verification and human judgement. Questions to ask before accepting a faster output as a better answer.
For: insight and analytics teams assessing AI-supported work.
For event organisers
Bio. Photo.
Less chasing.
Use the professional headshot and short bio when introducing me. Please link to philpegram.com and confirm event details with me before announcing participation.
Download headshot
Short biography
Phil Pegram is Head of Insights, Analytics and Market Research at TalkTalk and a DBA researcher at Teesside University. A former Insights Director at dentsu and British Army veteran, he connects customer understanding, analytics and organisational decision-making. His research examines the conditions that allow organisations to use evidence, including commissioning, accountability and analytical capacity.
Start a conversation
Tell me about your audience.
Include the event or podcast, who it is for, the topic, date and format. I'll have enough context to respond.