Toward Autonomous Research with Edward Olayemi

Inside Toward Autonomous Research

1. —

   Meet Edward Olayemi

   The episode opens with Edward's background: head of customer excellence and user research at FairMoney, with a foundation in statistics that pulled him into research early. He sets up a wide-ranging conversation on how research has moved from clipboards in the field to digital surveys, why quantitative data alone gets teams into trouble, and where he sees research heading, autonomous and conversational.

2. 01

   From Statistics to Research

   • How SPSS and Excel first opened the door, back when the tools were new to his region.
   • Running his first big analysis for a director, a two-year project that taught him methodology the hard way.
   • Realizing a survey instrument wasn't fit for purpose, and why that lesson stuck.
3. 02

   Research in University vs Industry

   • Why student research often aims at graduating, not depth.
   • How industry research ties directly to acquisition, retention, revenue, and growth.
   • The industry demand for root-cause analysis down to the most granular detail.
4. 03

   How Research Changed: From Field to Digital

   • The old way: walking neighborhoods with a device to log GPS coordinates by hand.
   • 300 surveys that once took three weeks, now pushed out over email, WhatsApp, or a WhatsApp status.
   • New reach through web scraping, sentiment analysis, and quick anonymized polls on X or LinkedIn.
5. 04

   Edward's Research Stack

   • Customer.io for feedback collection and configuring NPS and in-app surveys.
   • BigQuery as the data warehouse, with Python for deep-dive and root-cause analysis.
   • Claude for sentiment work, classification, and finding patterns in verbatim responses.
6. 05

   Why AI Isn't Hype

   • The tool he's building: product managers prompt a question, and the system fetches the right customers, composes the survey, fields it, and returns recommendations.
   • Human-in-the-loop survey design, with reports turned around in under 30 minutes.
   • His boss's line that stuck with him: the people who refuse to use AI are the ones who get replaced.
7. 06

   The Hardest Part: Getting Reliable Insights

   • Correlation vs causation, and the trap of reading one product's score in isolation.
   • "Product contamination," when anger at one product bleeds into ratings of another.
   • Why segmentation and persona matter: you engage high-net-worth and mass-market customers differently.
8. 07

   Where Teams Get Research Wrong

   • Over-relying on quantitative metrics without the qualitative "why" behind them.
   • His deep-dive method: call the customer who gave a low score to understand what they really meant.
   • The finding that 40% of the time it's a communication gap, not a broken product.
9. 08

   Don't Take the Customer's Word at Face Value

   • Validating claims against actual transaction data, which sometimes tells a different story.
   • Running two-level validation, with a second team independently checking the assumption.
   • The rule underneath it all: go deeper before you conclude.
10. 09

    Is AI Bringing Qualitative Research Back?

    • Why AI makes qualitative work more accessible, even conducting interviews and calls.
    • A homegrown transcription tool that handles Yoruba and Pidgin before analysis.
    • The case for building AI into your qualitative methodology, not just your surveys.
11. 10

    Researching the Nigerian Market

    • A digitally savvy, mobile-first base, reflected in FairMoney's ~30 million downloads and high response rates.
    • A market that's willing to speak up: as Edward puts it, Nigerians don't hide their pain.
    • Why segmentation still matters, from tech-savvy users to SMEs reached through agency banking.
12. 11

    The Future: Autonomous and Conversational Research

    • His magic-wand fix: a self-driving research team that frees people for strategic work.
    • Conversational surveys over WhatsApp, including voice notes customers can leave in their own words.
    • Advice for newcomers: learn research methodology, never stop at quantitative, and always stand in the customer's shoes.