Jun 19, 2015 · 20m · mad

David Glueck, Bonobos // Data Science & Engineering at Bonobos (Hosted by FirstMark Capital)

David Glueck · 16m spoken Matt Turck · 23s spoken

0:00 / 0:00

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

In this presentation hosted by FirstMark DataDrivenNYC, David Glueck details how Bonobos built an integrated data science, engineering, and analytics organization from the ground up to drive e-commerce growth. He outlines the company's organizational philosophy, flexible cloud infrastructure, practical experimentation with email and recommendations, and strategies for cross-platform customer tracking.

Chapters · click to jump

1 statement ▾

Bonobos was the largest U.S. retail company founded online in 2015 open →

1 statement ▾

Direct-to-consumer models allow superior inventory optimization by avoiding physical store distribution open →

2 statements ▾

Data teams make a critical mistake by de-emphasizing operational reporting open →

Separating data engineering from data science ties expensive employees' hands open →

3 statements ▾

Bonobos boosted email campaign revenue 4x purely through smarter targeting open →

Simple recommendation algorithms deliver 80 to 90 percent of the total benefit open →

E-commerce purchase decisions are highly elastic, but basket sizes are not open →

1 statement ▾

Native apps distract engineering teams by forcing them to build features thrice open →

2 statements ▾

Email unsubscribe rates are a poor optimization signal because they are rare open →

Ignoring marketing emails is just as bad as actively unsubscribing open →

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 2% of the talking time here. How this is scored →

Matt as informed peer 0.7 Guest teaching 2.0 Guest disagreement 1.3 Matt pushing back 0.5

speaking balance: gold is Matt, purple is the guest (3 minute bins)

Sharpest disagreement ▶ 4:30 Rejecting Siloed Data Engineering

Glueck forcefully criticizes conventional corporate structures that separate data engineering from data science, calling it hiring expensive people and tying their hands behind their back.

Hardest push from Matt ▶ 14:27 Host Presses for Live vs Wishlist Clarification

Host Matt Turck interrupts Glueck's broad answer on real-time tracking to explicitly demand whether the capability actually exists live at Bonobos or is just on their wishlist.

Biggest teaching moment ▶ 8:50 Counterintuitive Email Frequency Dynamics

Glueck educates the audience on why e-commerce brands send aggressive email volumes, citing a 6-month study showing 5 incremental purchases for every 1 unsubscribe.

Matt holds his own ▶ 13:42 Host Probes Real-Time Behavioral Trigger Capabilities

Host Matt Turck demonstrates industry understanding by immediately pivoting the presentation to real-time behavioral event triggers on the web platform.

the scores for every segment, with the reasoning behind each

Topic	Matt as informed peer	Guest teaching	Guest disagreement	Matt pushing back	Why
Defining Data Science and Delivering Business Value	0	1	1	0	Glueck gives a monologue introduction detailing Bonobos' e-commerce background and framing data science around operational business results rather than abstract definitions. The host is silent throughout this presentation segment.
The 'One Team, Three Goals' Data Team Structure	0	2	2	0	Glueck critiques industry trends that de-emphasize business intelligence or separate data engineering from data science, arguing that separate teams tie data scientists' hands behind their back. The host is inactive during this presentation segment.
Organic Technology Architecture and Cloud Integration	0	2	2	0	Glueck challenges the traditional pyramid stack tech analogy, advocating instead for an organic tree-like cloud architecture. The host does not speak in this presentation segment.
Key Practical Learnings in E-Commerce Optimization	0	2	1	0	Glueck outlines practical e-commerce takeaways regarding email frequency economics, recommendation algorithms, and promotional elasticity. The host does not intervene during this monologue presentation.
Q&A: Real-Time Tracking and Product Design Feedback	3	3	1	3	Host Matt Turck opens the Q&A by asking whether Bonobos uses real-time behavioral triggers, pushing Glueck to clarify whether that capability is live or just planned. Glueck explains their reliance on product similarity while working toward real-time personalization.
Q&A: Cross-Platform Tracking and Mobile App Strategy	1	2	1	0	Audience members ask about cross-platform tracking, native mobile apps, and email segmentation. Glueck defends avoiding a native mobile app to prevent engineering clutter, while host Matt Turck manages the Q&A timing.

Statements from this episode (10)

Assertion Contradicted

Bonobos was the largest U.S. retail company founded online in 2015

“And now we're the largest U.S. Retail company that was founded online.”

David Glueck Jun 19, 2015 ▶ 0:15

Insight

Direct-to-consumer models allow superior inventory optimization by avoiding physical store distribution

“And being that we don't have to worry about distributing our inventory across hundreds or thousands of stores we have a real opportunity to do that better than the traditional retailer, if we can get smarter about it.”

David Glueck Jun 19, 2015 ▶ 2:15

Insight

Data teams make a critical mistake by de-emphasizing operational reporting

“I think what I've seen as a trend in the industry, a lot of data teams are starting to try and de-emphasize the BI in reporting. They kind of see it as, ah, well, you know, it's a lot of work, which it really is, and in the end it doesn't add as much value to …”

David Glueck Jun 19, 2015 ▶ 3:45

Insight

Separating data engineering from data science ties expensive employees' hands

“If you're hiring a bunch of data scientists and then making them dependent on a different organization for engineering is basically Hiring some really expensive people, and then tying their hands behind their back, and you know, that's not a smart thing to do,…”

David Glueck Jun 19, 2015 ▶ 5:03

Assertion Not checkable as stated

Bonobos boosted email campaign revenue 4x purely through smarter targeting

“We were able to find, even without changing, ah, the emails that we were sending out, the different product emails that we were sending out, just by being smarter about who we sent each email to, we were able to get sometimes as much as Forex, the revenue for …”

David Glueck Jun 19, 2015 ▶ 8:05

Insight

Simple recommendation algorithms deliver 80 to 90 percent of the total benefit

“You get 80, 90% of the benefit from a very simple algorithm, and most of the really hard math and work goes into closing the gap on the rest of it.”

David Glueck Jun 19, 2015 ▶ 10:01

Insight

E-commerce purchase decisions are highly elastic, but basket sizes are not

“What we found is that the decision to buy is very elastic. So you can make small changes there, and you can really influence whether people are going to make that purchase or not make that purchase. But what's a lot harder to do is to influence what they're go…”

David Glueck Jun 19, 2015 ▶ 12:34

Insight

Native apps distract engineering teams by forcing them to build features thrice

“Because it's a, you know, huge distraction. You have to develop every feature, you know, three times for iPhone, Android, and the web.”

David Glueck Jun 19, 2015 ▶ 16:45

Insight

Email unsubscribe rates are a poor optimization signal because they are rare

“What we found basically is that the unsubscribe rate's not a great signal just because so few people do it.”

David Glueck Jun 19, 2015 ▶ 19:34

Insight

Ignoring marketing emails is just as bad as actively unsubscribing

“It's almost, it's just as bad if someone just ignores us as if they unsubscribe from us, and so we have to look at that as being the same thing, basically.”

David Glueck Jun 19, 2015 ▶ 19:48

David Glueck, Bonobos // Data Science & Engineering at Bonobos (Hosted by FirstMark Capital)

How this conversation actually went

Statements from this episode (10)

Turn any episode into a week of clips.