May 29, 2019 · 55m · y-combinator
Analyzing Billions of Transactions to Understand Consumer Behavior - Michael Babineau and Kevin Hale · Y Combinator
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this Y Combinator podcast episode, host Craig Cannon and partner Kevin Hale interview Second Measure co-founder and CEO Michael Babineau about building a self-service analytics platform that transforms messy credit card transaction data into actionable financial and competitive intelligence. Babineau discusses the company's engineering origins, data pipeline architecture, hiring methodology, and real-world consumer behavior insights.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The partners as informed peer 3.5 Guest teaching 4.4 Guest disagreement 0.5 The partners pushing back 1.3
speaking balance: gold is the partners, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 51:08 Defending competitors against host dismissal
Babineau pushes back directly against Hale's assertion that Citi and Mint are bad at processing data, explaining the fundamental divergence in technical constraints and product goals.
Hardest push from the partners ▶ 15:48 Questioning unit economics data visibilityCraig Cannon halts the conversation to question how Second Measure could possibly claim visibility into unit economics from credit card data alone.
Biggest teaching moment ▶ 38:12 Baffling scale of transaction string entropyBabineau stuns the host by revealing that within 50 billion transactions, there are a billion descriptions and over three million distinct representations for Macy's alone.
The partners hold their own ▶ 40:02 Deconstructing user fatigue in crowdsourced taggingKevin Hale demonstrates sharp product insight by diagnosing why Mint's crowdsourced correction model fails as repeated friction burns out user willingness to categorize data.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The partners as informed peer | Guest teaching | Guest disagreement | The partners pushing back | Why |
|---|---|---|---|---|---|---|
| Origins and Core Vision of Second Measure | 3 | 5 | 0 | 1 | Kevin Hale and Craig Cannon explore how Babineau transitioned from gaming data infrastructure at EA to financial transaction intelligence. Babineau educates the hosts on how unsophisticated multi-billion-dollar hedge funds were with raw big data. | |
| Transaction Data Versus Traditional Market Research | 4 | 4 | 0 | 0 | The conversation contrasts slow, low-sample consumer survey methodology against direct transaction observation using the Chipotle food poisoning incident as a case study. | |
| Early Customer Acquisition and Data Boundaries | 3 | 3 | 0 | 1 | Babineau describes leveraging YC meetings with VCs into software sales, delineating what direct credit card data captures versus wholesale/B2B blindspots. | |
| Due Diligence and Competitive Intelligence for Investors | 4 | 4 | 1 | 3 | Cannon presses Babineau on how Second Measure calculates unit economics for due diligence, prompting Babineau to clarify that they only track top-line consumer spend rather than internal merchant costs. | |
| Data Storytelling and Editorial Press Strategy | 2 | 3 | 0 | 1 | Babineau explains their PR strategy of partnering with journalists covering tech IPOs like Uber and Lyft using a dedicated internal editorial team. | |
| Analyzing Consumer Behavior: Stitch Fix Case Study | 4 | 4 | 0 | 0 | Babineau shares counterintuitive findings from Stitch Fix data showing it expands overall apparel spend rather than cannibalizing department stores, with Hale providing behavioral insights. | |
| Market Shifts: Peloton Versus SoulCycle | 3 | 5 | 1 | 2 | Babineau outlines Peloton overtaking SoulCycle ridership and reframes Hale's assertion regarding Amazon's Prime strategy by clarifying revenue composition metrics. | |
| Product Roadmap and Hiring Quantitative Talent | 4 | 4 | 0 | 1 | Babineau describes prioritizing core statistical and mathematical foundations over specific software tooling when hiring data science talent. | |
| Data Science Interview Philosophy and Common Pitfalls | 3 | 4 | 0 | 0 | Babineau details their take-home interview assignment, explaining that the biggest pitfall for candidates is naively assuming data is clean and overlooking underlying distortions. | |
| Resolving Complexities in Raw Transaction Data | 3 | 7 | 1 | 2 | Cannon expresses shock that financial transaction strings are not standard or normalized. Babineau educates the room on the reality of three million unique text variations for a single company like Macy's. | |
| Enterprise Expansion, Investment, and Future Edge | 5 | 5 | 2 | 3 | Hale challenges why established banks or tools like Mint fail to clean transaction data effectively. Babineau reframes the problem of high-precision tracking for specific enterprise tickers versus general consumer budgeting. |
Statements from this episode (13)
Most hedge funds operate without any in-house coders, Michael Babineau says
“In reality most most hedge funds have a handful of analysts and just some back office support, right? They don't have any coders in house.”
Investors generally lack the technical skills to process messy credit card data
“Credit card transaction data is, it's a messy and, like, there, it's a messy data set with unstructured data problems baked into it. And the skill sets of investors, even the more technical ones those tend to lean more towards, like, time series analysis as op…”
Babineau: Second Measure's First Customer Was a VC Who Later Invested
“So our very first customer was a VC who also ended up investing in us.”
Most Bay Area venture capital firms are Second Measure customers, Babineau claims
“And now most most of the VCs here in the Bay Area, they are our customers, you know”
Stitch Fix did not cannibalize department store spending, Second Measure found
“And what we found is that Stitch Fix had
no impact on department store spend.
Right?
People just started spending more on clothes.
Period.”
Peloton monthly active members have surpassed SoulCycle riders, Second Measure data shows
“I mean, the short version is that Peloton has now surpassed SoulCycle in terms of, like, the number of active Peloton members, right, and this is based on a spending based on spending behavior. Active Peloton members on a monthly basis have surpassed the numbe…”
Babineau: Amazon revenue is increasingly reliant on Prime subscribers
“Increasingly, Amazon is looking more and more like a subscription business. Like they're increasingly reliant on Amazon Prime customers for their revenue.”
Lapsed Amazon Prime subscribers continue spending more than they did before subscribing
“People who became an Amazon Prime subscriber, even if they lapse, right, even if they are no longer a subscriber they're still spending more on Amazon after than they did before.”
Data science tools can be taught on the job, but math cannot
“It's our view that if you come in with that, that strong that strong, like, you know, mathy foundation, that learning the tools, like, the tools can be taught, right? We can, like, we're happy to help to help people get onboarded with, like, using Python. Like…”
Babineau: Data scientists' biggest mistake is assuming data is clean and perfect
“I'd say the, like, the number one mistake that people make is that they you know, they assume too much of the data. They assume the data's perfect, right?”
Second Measure sees 1B unique descriptions across 50B+ transactions
“So within our data set we have So we're looking at like, 50 plus billion transactions. We have one billion unique transaction descriptions.”