Watch
How DoorDash is Powering Omnichannel Excellence with AI & Data | Live at Shoptalk 2025
Episode 242 •
27th March 2025 • Omni Talk Retail • Omni Talk Retail
00:00:00
00:16:31
Shownotes
In this Shoptalk 2025 episode, recorded live from the Meta Podcast Studio,, Chris Walton and Anne Mezzenga chat with Jessica Lachs, Chief Analytics Officer at DoorDash, about her remarkable rise from startup GM to data leader. Jessica unveils DoorDash’s newest partnerships with DSW and Klarna, discusses the role of AI in grocery substitutions and targeted retail media, and explains why data is the foundation of customer-centric innovation.
Key Moments:
0:20 – Jessica Lachs joins Omni Talk from Shoptalk 2025
1:20 – Her startup journey: from GM to CAO at DoorDash
3:00 – Learning SQL and Python from scratch
4:10 – Launching new partnerships: Klarna and DSW
5:45 – DoorDash’s evolving verticals beyond restaurants
7:00 – Retail media and new DoubleDash ad platform
8:20 – Role of AI in customer personalization and targeting
9:45 – AI-powered grocery substitutions and smarter alternatives
11:40 – Creating seamless omnichannel experiences
12:55 – Solving the first-time search and personalization challenge
14:30 – Advice for brands: invest in analytics before tech
15:45 – Building an intentional data strategy and empowering analysts
16:30 – Wrap-up with “Life Hacks with Lachs” podcast teaser
#doordash #shoptalk #retailinnovation
This podcast uses the following third-party services for analysis:
Podcorn - https://podcorn.com/privacy
Transcripts
Speaker A:
Welcome back, everybody.
Speaker A:We are here with our final interview of Shop Talk day two.
Speaker A:This is Omnitok Retail.
Speaker A:I'm Anne Mazinga.
Speaker B:And I'm Chris Walton.
Speaker A:And standing between us, we have one of the OGs of DoorDash.
Speaker A:This is like a super exciting interview, Chris, because Jessica Lax, the Chief analytics officer, standing between us.
Speaker A:You were what.
Speaker A:What employee were you at Doordash?
Speaker C:You know, I don't know the actual number, but it was probably in the 20s.
Speaker A:In the 20s?
Speaker A:Yeah.
Speaker A:And how many people are at Doordash now?
Speaker C:More than 20,000.
Speaker A:I mean, you've seen some things.
Speaker C:I have.
Speaker C:My lips are sealed.
Speaker A:We don't have to talk about all of them right now, but I'm really excited to dive in with you today.
Speaker A:Before we do that, I just want to give a quick thank you to Meta, who has made all of our coverage here at Shop Talk today day two, possible.
Speaker A:Thank you to Meta, where you can unlock seamless customer centric experiences with omnichannel ads.
Speaker A:Thank you again to them for making all of our coverage possible.
Speaker A:All right, let's dive in.
Speaker B:Yes, Jessica.
Speaker B:So like Ann said, you were one of the first team members at Doordash, but let's talk about your background a little bit too.
Speaker B:Like chief analytics officer.
Speaker B:I think you may be the first one of those we've ever had on our show.
Speaker B:I'm honored how you became that and what is it that you do on a daily basis?
Speaker C:That's a great question.
Speaker C:So I am a bit of an anomaly, which is I have a job that I shouldn't have.
Speaker C:I actually started off as a GM over 10 years ago.
Speaker C: So I started in: Speaker C:And since then, I've sort of navigated into the analytics space by just being endlessly curious, really.
Speaker C:So.
Speaker C:So I have a natural curiosity and I ask a lot of questions, and there was no one to answer those questions except for myself.
Speaker C:And so I had to figure out how to acquire the skills to be able to answer those questions.
Speaker C:Started asking more questions, answering more questions, built a team to help me answer those questions, and that's how I moved into the analytics space.
Speaker B:Wow, interesting.
Speaker B:But are you, like, a quant in any way, shape or form, or, like, you know, like, what's your.
Speaker B:You gotta have some, like, math in your background, right?
Speaker B:Or, like, what's going on?
Speaker C:So I do have some math in my background.
Speaker C:I actually started off as an investment bank banker.
Speaker C:So I spent several years after graduating in the investment Banking space at Lehman Brothers.
Speaker C:So that'll date me.
Speaker B:Ah, yes.
Speaker C:And then I went to business school where I started a company called giftsimple, which ultimately failed.
Speaker C:But it was that failure that created the introduction to Tony Hsu, our co founder and CEO through Sequoia.
Speaker C:And he liked my background, that I wasn't just a banker with an mba, but I had tried to start something, had that entrepreneurial spirit.
Speaker C:And so when he was looking to hire his first GM and help launch a bunch of cities, my resume was sent to him and he thought it looked interesting.
Speaker B:Wow.
Speaker B:And so then you.
Speaker B:And then it transpired into you taking on more of an elevated role too, and learning how to do that as you figure out what that role is about.
Speaker C:Exactly.
Speaker C:So I knew some basic SQL when I had started.
Speaker C:Started, but our data, there was no one to access the data for me.
Speaker C:I had to actually figure out how to do that myself.
Speaker C:We didn't have structured data models that would make queries.
Speaker C:Exactly.
Speaker B:Wow, that's amazing.
Speaker C:Yeah.
Speaker C:So self taught.
Speaker C:And SQL and Python, out of the goodness of his heart.
Speaker C:One of our early engineers actually tutored me on the weekends to teach me Python and that's just the beginning of where I am today.
Speaker B:That's a true startup story.
Speaker B:That is awesome.
Speaker B:God, I love that story.
Speaker A:And so cool.
Speaker A:I mean, you're definitely in line with the kind of people that we align with.
Speaker A:Where, you know, doing.
Speaker A:Doing what you cook yourself.
Speaker C:Yeah, exactly.
Speaker A:Doing it and actually getting that experience to help you figure out.
Speaker A:Yeah, that one didn't work.
Speaker A:But how are you going to apply what you learned, the next role?
Speaker A:Well, I want to shift gears just a little bit because you've made some pretty big announcements, I would say in the last couple days, you being doordash, I'd love for you to share a few of those with our audience.
Speaker A:We did cover one of them on our Fast 5 this morning about partnership with Klarna.
Speaker A:But what are you launching here at Shop Talk?
Speaker C:Jessica?
Speaker C:So we've got a number of announcements.
Speaker C:So you mentioned the Klarna partnership.
Speaker C:We also announced a partnership with DSW because as you know, you can get a lot of things delivered through doordash, not just restaurants.
Speaker C:We have a huge new verticals business across grocery, retail, convenience, pets, flowers.
Speaker C:And so DSW is one of the latest partnerships within our retail vertical.
Speaker A:That's awesome.
Speaker A:There are such things as shoe emergencies.
Speaker A:I feel like that's always, everybody's always like, do you need shoes in 30 minutes or less?
Speaker A:Absolutely you need shoes.
Speaker A:Especially when you're at a conference in landlocked Las Vegas.
Speaker A:I am super excited about the door or the DSW DoorDash partnership and may.
Speaker B:Or may not need those today.
Speaker B:Actually, I do.
Speaker A:Yes.
Speaker A:I will be doing that immediately after this interview.
Speaker B:Explain the Klarna Partnership, too, if you don't mind, for those maybe that aren't as familiar with Klarna.
Speaker C:Yeah, I mean, I've seen the memes like you probably all have, but, you know, as I mentioned, we don't just deliver burritos.
Speaker C:There are a lot of things that you can order on DoorDash that are quite pricey, that are larger ticket items.
Speaker C:We were talking earlier about ordering ipods on the AirPods.
Speaker C:Sorry, yes.
Speaker C:Dating myself again.
Speaker C:I loved my ipod.
Speaker A:Yeah, obviously we all did.
Speaker C:But you can order AirPods through DoorDash or even a leaf blower from the Home Depot.
Speaker C:And so for shoes.
Speaker C:Exactly.
Speaker C:Shoes from dsw.
Speaker C:And so if you.
Speaker C:For folks, the Klarna Partnership is really about flexibility.
Speaker C:So we want consumers to have the payment options that work best for them, whether that's PayPal or Apple Pay or traditional credit card or Buy Now, Pay later through Klarna.
Speaker C:It's really just about flexibility.
Speaker C:And so you also can use Klarna if you want to get an annual DashPass subscription for a discounted price and then break up that rather than paying it all at once.
Speaker A:Oh, my gosh, I didn't know about that little angle.
Speaker B:Now we're in the life hacks, eh?
Speaker B:Life hacks with lax, I think.
Speaker B:Yeah, that's what we got going on here today.
Speaker C:All right, it's gonna be my new podcast.
Speaker A:Thanks for that.
Speaker B:That's a good one.
Speaker B:All right, so retail media, it's been the talk of this conference.
Speaker B:It's been the talk of many conferences over the past years.
Speaker B:It's been the talk of the industry for the past little while, too.
Speaker B:What have you, you know, it's a very important part of, you know, what you guys do on a daily basis as well.
Speaker B:What have you seen that, you know, separates the retailers that get the retail media side from those that don't?
Speaker C:When you mean retail media, you're specifically talking about, like, advertisers within the platform.
Speaker C:Yeah, I think, you know, we have 42 million monthly active users that are using the DoorDash or Volt platform globally.
Speaker C:And so it's a massive audience to get your products in front of.
Speaker C:And while we do a lot of things using our own data to improve personalization, you have.
Speaker C:There's a great opportunity to acquire new Consumers spread your reach, showcase new offerings through.
Speaker C:Through advertising.
Speaker C:One of the other large announcements we just had was advertising through Double Dash, which is at checkout, after you placed your order, you can add on.
Speaker C:That's great.
Speaker C:You can add on more items to your order from other stores.
Speaker C:And that's a great way to get visibility in front of consumers who may not even know that you're on doordash, who may not be aware that you sell certain items and to be able to get additional incremental sales on the platform.
Speaker A:Right.
Speaker A:That's so cool.
Speaker A:I have to ask, in order to enable that, especially to the level that you're doing that, how does AI come into play there?
Speaker A:Are you tapping that?
Speaker A:That is a technology to help kind of place the right ads in front of the right customers at the right time.
Speaker C:Absolutely.
Speaker C:So we have an enormous amount of data, a lot of.
Speaker C:If you think about all of the information we have that we collect from the actions that consumers take on the platform, their engagement, their search behavior, what they've ordered before, when they've order, what promotions they've taken advantage of on past orders, we have all of that information and we can use it in order to surface more relevant selection for them to reduce friction and really make browsing and choosing what you want delightful.
Speaker C:One of the most fun problems that we are using AI for is actually grocery substitutions.
Speaker A:Oh really?
Speaker C:If you think about it, we have hundreds of thousands of SKUs.
Speaker C:And if somebody orders chocolate chip cookies, something that I often do, and the store is out of chocolate chip cookies, we have to figure out what are the right substitutions to make for that order.
Speaker C:If you order Chips Ahoy chocolate chip cookies, in the old days when we would do user testing, we would see that things like chocolate chip cookie dough ice cream would come up as a substitution.
Speaker C:Exactly.
Speaker C:Fuzzy matching of the name, while delicious, I don't think that that was a great substitution.
Speaker C:I think we can do better.
Speaker C:And so what we would hope to see is maybe another chocolate chip cookie brand like Entenmanns or Tate's as an alternative.
Speaker C:And so through our data labeling, through getting actual data from consumers about the substitutions that they want, we're able to learn and run more sophisticated algorithms that surface the right substitutions at the right time for the right consumer based on their pre conferences.
Speaker A:Yeah, that's so true, actually, because when I was placing, as I mentioned, one of my hacks for shop talk, or any conference for that matter, is having a dash pass subscription.
Speaker A:I did notice yesterday and I typically just avoid the Substitutions and say no substitutions at all cost.
Speaker A:Because I just.
Speaker A:If you don't have it, I don't want it.
Speaker A:It's not usually right.
Speaker A:But I have to say, this time, doing this, I did go through a few items that were served up to me, like, this might be low.
Speaker A:Do you want this item?
Speaker A:And for the first time, it was more accurate.
Speaker A:Like, it was actually raspberries that were maybe organic raspberries versus, like, raspberry jam or something, which.
Speaker C:Exactly.
Speaker A:That's just.
Speaker A:That's.
Speaker A:It's a testament to how you're really trying to make that shopping experience better and more reliable for me as a doordash customer.
Speaker C:Exactly.
Speaker C:It's all like a crawl, walk, run kind of process, similar to the origins of doordash in general, which is you start doing things, then you do them better and smarter, and you learn from your own mistakes and you learn from the data that you collect.
Speaker C:So that's obviously my passion.
Speaker C:And then you can get even better and really make it magical for consumers.
Speaker A:And the technology is adapted to allow you that at a speed that makes sense, too.
Speaker A:Right?
Speaker B:Yeah.
Speaker B:I would call out two things you said there, one of which is how you talked about if you've been on the platform before, it starts to know you and it knows your habits.
Speaker B:And I think doordash does make it easier than most other platforms of, you know, me reordering the things that I've ordered before, which it makes it just so easy to use.
Speaker B:And the second point I call out is this is the first time in history again that we've had entamins dropped on a podcast, so nicely done again.
Speaker B:All right, so I want to ask you this question.
Speaker B:So, Chief analytics officer, you get probably a bird's eye view into a lot more data in terms of where things are going than, say, other people.
Speaker B:Where does that data tell you that the Omnichannel retail experience is heading next?
Speaker C:So I think you bring up a really great sort of keyword, omnichannel.
Speaker C:I think it's important to meet consumers where they are, whether that's online, through the doordash platform, or in store.
Speaker C:We want to make sure that they can find what they need, when they need and how they need it.
Speaker C:And your in store shopping behavior, being able to tie that to your digital shopping behavior is really important because it helps us to learn what your preferences are.
Speaker C:Do you like organic raspberries or not?
Speaker C:And so the more information we have, the better we can make the recommendations and improve the shopping experience for consumers.
Speaker C:So I Think there's a lot more that we can do, particularly on the edges.
Speaker C:We focus a lot on what are the most frequent queries, and you talk a lot about repeat behavior.
Speaker C:But the hardest problem is the first time you encounter something, the first time you search for a new category or a new item.
Speaker C:How do we get that right?
Speaker C:How do we figure out based on consumers that look like you?
Speaker C:Metaphorically speaking, yeah, of course.
Speaker C:But how do we learn from consumers who have similar behaviors that you have and see what they've selected so that we can do the same and make your experience better that first time instead of having to always learn from your own behavior on the platform?
Speaker B:Got it.
Speaker C:We're just at the beginning.
Speaker C:As you mentioned, the technology is developing incredibly fast, and I'm so excited about what the future will bring.
Speaker C:There are so many challenging problems we have to solve across our three audiences, consumers, dashers and merchants across all of the verticals and items and skus that we have.
Speaker C:It's just a really ripe environment for hard problems.
Speaker C:And you know, as at doing what I do, I love the hard problems.
Speaker B:Right.
Speaker B:But so if I recap.
Speaker B:Recap what you're saying.
Speaker B:Basically saying on the margins, there's still just so much, so much, so many rows to be hoed, so to speak, on just improving the personalization for the returning shopper as well as the new shopper.
Speaker C:Exactly.
Speaker A:Okay, well, Jessica, because we have you here, I want to close with any advice that you have for any of our retailers and brands listening at home or from the show, whether that's advice on how to attack unified commerce going into the year ahead to survive a startup as one of the first 20 people before 20,000 employees start.
Speaker A:What.
Speaker A:What knowledge can you share, impart on our audience?
Speaker C:Oh, wow.
Speaker C:That's a tough question.
Speaker C:I think.
Speaker C:So I'm going to, because of my role and because I am the first chief analytics officer that you have interviewed on your podcast.
Speaker C:I'm going to go.
Speaker C:I'm going to go with the analytics base, which is one you should have an analytics team and that you should measure as much as possible, because having that information is really the foundation on which so much can be built.
Speaker C:Whether that's tracking the right metrics for your business so that you can hold yourself accountable to hitting your goals and pivot when necessary when things aren't working out so that you can understand if the things that you're shipping to consumers or the business decisions, decisions you're making have the desired effect to solve the root problem for your customer, which is Ultimately, what you're trying to do or to build these new AI models on top of, you need that data foundation, or the models aren't going to be that great.
Speaker C:So I think it all comes down to thinking about analytics, thinking about data and being really intentional with your strategy and making sure that you open a seat at the table for whoever you have on the analytics team to help make those better business decisions and help you make them more quickly.
Speaker C:So I just sort of advocate for all the analytics folks out there who have great ideas that want to be heard.
Speaker A:Well, help uncover the questions or the.
Speaker C:Problems that you're trying to solve before.
Speaker A:You throw teams or technology at them.
Speaker C:Absolutely.
Speaker A:Okay.
Speaker A:Well, I don't know.
Speaker B:That's awesome.
Speaker A:I don't know if there's anything more to say after that.
Speaker A:Thank you so much, Jessica, for your time today, for sharing your wisdom with all of our Omnitalk audience.
Speaker A:Thank you.
Speaker A:Finally again, to Meta once more.
Speaker A:We really appreciate your support of our Omnich coverage.
Speaker A:Chris, anything you want to say before we wrap up day two?
Speaker B:No, but I'm going to be the first subscriber to Life Hacks with Lax, I think, Ann.
Speaker A:So maybe we can convince her to come under the Omnitok umbrella.
Speaker A:Coming to a podcast network.
Speaker B:Yeah, right.
Speaker B:The Omnitalk podcast network.
Speaker B:All right.
Speaker B:But until next time, be careful out there.
Follow
Links
Chapters
Video