Podcast: App engagement to millions of users (Walmart Savings Catcher)
Podcast: App engagement to millions of users (Walmart Savings Catcher)
In this podcast series we ask experienced Appreneurs for one success story and one fail.
Stuart Argue is one of the core technologists that drove Walmart’s eReceipts Platform, in this talk he shares insights from rolling out the Savings Catcher App that had engagement to millions of users within a few months of launch.
A key acquisition element was the linking of real-world to the App download via the promotion on every printed purchase receipt. Not many pureplay Apps get a chance to do this but for brands and retailers its a great, free, compelling method to get in front of users.
For all our interviews we ask the interviewee to score their own anecdote based on 3 axes of: acquisition, UX and retention. In this hypergrowth retail example Stuart self-scored his learnings and experiences from Savings Catcher as 1,9,7 because of the uptick in utility that both Walmart and the consumer got from Native Apps with bar-code scanning and push notifications as part of the user lifecycle.
Let us know on our Twitter account how you score it.
Stuart talks about how with a massive potential user-base and support from top-level management, the Savings Catcher rollout had plenty of real-world acquisition support. But with a lot of UX planning and native App capabilities like scanning and Push Notifications, they delivered a compelling value prop for Walmart shoppers.
DAVID: Gidday. This is David Jones from Street Hawk, and I’m here today with Stuart Argue. Stuart was one half of the team of the company called Grabble. The other half was Anthony Marcar. Those two guys were in the first cohort of Startmate teams that went through. They came to us with a really cool piece of hardware that allows you to actually intercept receipts at the cash register.
And I, for one, said to them: “It will never work. How are you gonna actually deploy this to thousands and thousands of cash registers?” They said, “Don’t worry about that.” Anyway, they got acquired pretty quickly by Walmart, and for the last few years, Stuart’s been on a wild right inside Walmart. So I hope you enjoy this talk. Stuart, what have you been up to lately?
STUART: Well, lately, Walmart bought our technology. Anthony and I, we really knuckled it down to try and get it out to scale, and for us that means getting it to 4,200 stores and 50 cash registers per store. So it took a little time but…
DAVID: No suprise!
STUART: Yeah. We knocked it down. We didn’t just target by too much. But the first step was really to get that hybrid device and make it work on a Walmart point of sales system. As soon as we did that, it got in front of the board, and the board said, “This is great. Do it in software.” So I guess, David, you’re pretty right in that getting hardware out to all these stores wasn’t going to work, but the concept and the way that we tackled the problem was equally applied in the software world. So, we built this e-Receipts Platform. It’s really a platform to build a lot of other products on top of it.
DAVID: Yes. Just going back to the e-Receipts aspect of it, what would make a consumer be interested in what you actually trying to achieve?
STUART: Yeah, good question. When we’re a small startup, we went out and talked to people. We thought only about one in ten people even really cared about their receipts, so we’re already operating on the assumption that most people didn’t care. But we wanted to build a platform on top of it, and I think when we were pitching early on, it was around loyalty. That kind of stuck with people, but even still it’s what feature do you want to build on top of it?
The nice thing for Walmart is they have a business model around “Every Day Low Prices”, and that means effectively no loyalty, as in everyone should be treated the same. So we were really going in with e-Receipts to create a loyalty product but with no loyalty aspect to it effectively. [Laughter] It was a really subtle thing. I mean, a lot (60%) of transactions at Walmart are cash and untraceable and things like that, and we wanted to have a much closer engagement with the consumers while they’re in the store to be with other great products and have a really great mobile experience being in-store shoppers. That’s really the team that Anthony and I built out at Walmart.
DAVID: Given that there was a kind of a disconnect in the business model and what you were trying to achieve, what actually made Walmart interested in you guys then wasn’t just the technology, it wasn’t because you’re two very good-looking gentlemen, or some other reasons.
STUART: Definitely not the latter. [Laughter]
DAVID: Trick question.
STUART: I think there’s a really strong digital and multichannel strategy put together by some really, really smart guys, great executive leadership around really knowing that mobile was the future and seeing eReceipts as a platform at help drive engagement and help understand our customers better individually, tying their purchase history to a profile and then being able to build products that leverage that and are really customer-focused. The vision was two years out, so there’s some really, really solid ex-entrepreneurs are in leadership roles at Walmart.
DAVID: So they saw that and they definitely wanted to go in that direction and they’re willing to actually spend two years getting that right.
STUART: Yeah. I think 18 months was probably the initial estimate, but we were talking to lots of people around the same time, and these were the only guys that flew one their CEOs in of the eCommerce business. He looked us in the eyes and said, “We’re going to ship this if you can build that.” There’s a lot of optimism, and they saw the value and were willing to take a risk, so our hats off to them for sure.
DAVID: How has that been realised today? If you’re a Walmart shopper, how do you know you’re using the technology formerly known as Grabble?
STUART: Well, the main product we built on top of eReceipts was called Savings Catcher. Effectively once you purchase your products at Walmart, you interact with the eReceipts product formerly known as Grabble, if you want to call it that. We will go out and look at competitive pricing in the local area, competitive advertised pricing. If there’s anything cheaper, we’ll give that money back onto a virtual gift card to be spent back at Walmart. Or you can even sign up for an American Express-Walmart combined product called “Blue Bird” which is a debit card, and under that model, you get double the money back that you can re-spend at Walmart.
So, it’s a really strong value proposition to the customers, saying, “Look, Walmart’s the cheapest. Don’t bother shopping around, you just come to us every time.” That rolled out starting in August this year, and I think for a couple of weeks, a lot of America knew about it and were using it. Of course, there were some interesting times with so much engagement.
DAVID: Can you share any numbers in terms of what you’re doing, in terms of daily and monthly actives and stuff like that?
STUART: I probably can’t share direct data but I guess there’s some public data that I can share. The app was sitting in the top five on the App Store for about six weeks.
DAVID: Six weeks? That’s huge!
STUART: I think the Android app publicly, according to Google Play, has between 10 and 20 million installs [EDIT: Stuart updated me later its actually between 10 and 50 million installs]. You can guess the numbers just by how many people shop at Walmart. It’s a huge brand, and every one of those people gets exposed to the product. Not all come obviously, but we’re talking pretty significant percentages.
DAVID: Right. So how does a consumer actually kind of get to know about Savings Catcher? As you just sort of implied, there may be some signage in store or…?
STUART: Yes, there’s definitely a lot of signage but also huge media spend on TV advertising. I think we sponsored the World Series, things like that, radio advertising. I was even driving to work one day and heard an ad for Savings Catcher in Spanish. The only word I understood was “Savings Catcher” and I thought, “Cool.”
DAVID: So you were listening to a Spanish radio station?
STUART: Well, I was just looking through the channels as I do driving up the (U.S Route) 280 there. But, yeah, typically the call to action, the big thing is on the bottom of the receipt and we inject an image and a QR code onto the receipt, and the image is the call to action saying, “Hey, get the iPhone or Android app. Scan this QR code to get started.” I think we tried lots of different copy, so that’s definitely not the exact copy. I’ll get an image for your blog post, if you like.
STUART: Yeah. They engage with the app, they scan the QR code with the app, they wait a couple of days and we send the push notification back to them saying, “Hey, there’s some savings. We can put it on the gift card,” or “Sorry, no savings this time but try again next time,” effectively.
DAVID: That’s interesting. So that instead of just relying, you know, in terms of your kind of acquisition point to go from, you know, real world or meatspace into the app space, you’re using the receipt, as opposed to like an in-store signage. That was something that was right there right at transaction time that was meaningful to the user. Every time they went in and bought something, that thing was sitting there, so that somebody might see that three, four or five times before they ultimately actually responded to the call to action, I guess.
STUART: Exactly. Now don’t get me wrong. There’s a lot of signage in the stores basically about every register. Every second register, there’s a sign, and some stores have this televisions on closed loops advertising the product. So there’s a lot of advertising around. They’re trying to drive the right message, the message that Walmart’s the cheapest and here’s a way to prove it. It’s just the image and the QR code is the actionable item to tell them exactly what to do, so we actually used a short URL to drive the download; wmt.co does detection of what type of phone you have and takes you to the right App Store or go to the mobile web. We built the product to cross iPhone, Android, mobile web and desktop.
DAVID: So I’m guessing that probably Walmart didn’t think they’re made to run a lot of promos in store or anything like that. Was there a particular point in time where acquisition was struggling because you thought it would be easier than what it was?
STUART: No. Actually this product is one of those one in a hundred that just blew up. The major stakeholder was the marketing group, so they were trials run and we’re in one market for a long time before going to other markets. So it’s a pretty well-known proposition, but the early markets we went into didn’t have native apps and there was no QR code scan and things like that. That was just mobile-Web and desktop. So the percentage adoption went up significantly once we had that image on the receipt, the QR code and native apps. The experience was so much better on native apps; it can send push notifications and we can do all sorts of things. That just doesn’t quite work when you’re just sending emails and things like that for the mobile-Web and desktop users.
DAVID: Sure. The role of the push a couple of days later, was that something you did straight away or is that something that you evolved over time?
STUART: Again, that was off the bat. We have a very big and active UX team that I think we’re revising the wire frames and the flows for a good six months, getting it out in front of various test users and things like that and walking through it. We definitely changed the language and the value props a few times beforehand but the core technology on the app is pretty straightforward. We kept it pretty simple for the user. But what goes on behind the scenes to get all the data out of the stores and everything else is where a lot of the complexity lies.
DAVID: So effectively, your UX team was basically looking at the user’s life cycle that you actually launched with what – we tend to “rabbit on” about this term “launch with an engagement strategy”, you know, so you don’t lose 80 percent of your customers before you figure out how to actually ignite their interest. You actually think that through before you launched. That sounds like you did that.
STUART: Yeah. And I think a lot of the challenge there in the UX time was working with marketing or not getting the exact value proposition right, because Walmart is the cheapest most of the time, so all the time these people aren’t getting money back. And so really coming up with the language that when they don’t get money back because Walmart was the cheapest to say that they’re a smart shopper and all these other things, but at one point we were tallying-up how much they saved, but it’s only against advertised prices, so it’s selling Walmart a bit short. So we ditched the whole “how much you saved” aspect and really focused on the core value prop that Walmart’s the cheapest, and in the event that they’re not the cheapest, we’re going to give you the money back to make it equal cheapest, and then “double your money back” launched about two months after, afterwards with American Express.
DAVID: But there was that thing in there that had something to do with “there are savings available for you” or something like that mentioned that.
STUART: Yeah. Sorry I probably mis-phrased. When the push notification comes through, it says, “Hey, we did find some cheaper prices and here’s the money back on a gift card.”
DAVID: Ah, okay, right, very cool.
STUART: So the biggest thing was some percentage of people, first-time users, they can now get money back and some arn’t. Then looking at retention, the ones that don’t get money back, say, two or three times in a row, do they keep doing it and things like that. There are all sorts of permutations there. On our side of the world, we’re really trying to drive engagement; on the marketing side of the world, they’re really trying to get brand loyalty. So we come from slightly different angles but it was one of those products where it was really well aligned.
DAVID: Right. So it sounds like there was an inflection point right at the start, but was there a particular sort of aspect to the application design or that acquisition stuff that we’ve been talking about that really was a game-changer or an inflection point for you guys?
STUART: Yeah. I think because our initial pilots didn’t have the native apps, that when we launched with the native apps, that was definitely a huge inflection point, I think, within a month. The non-native apps had been running, I think, for two or three weeks prior to the native apps going on. I think within a month they were doing – and in a lot of markets even well before that, (for another) five months. So I think at launch time within a month, I think we’re doing 5X the number of receipts through native than through the existing platforms that people are already aware of.
DAVID: Oh, wow. How were you actually even doing scanning through HTML?
STUART: It wasn’t. So that’s exactly the issue. The eReceipt also has a 19-25 digit number on it that they could type in.
DAVID: Good luck with that!
STUART: Yeah, exactly. We did trial taking photos through mobile-Web very early on and uploading the photo and doing OCR on the photo, but it was a pretty ordinary experience, so we ditched that.
DAVID: So you got something that was virtually impossible down to kind of like a five-second experience. A lot of people complain about QRs and stuff like that, but that’s a huge uptake in functionality just by giving somebody the ability to scan it really quickly.
STUART: Yeah, definitely. I hate QR codes. It pained us a lot in Grabble times to finally end up putting a QR code on the receipt, but it works in the end. When you’re giving somebody clear value, they’re happy to learn how to scan a QR code. But if there’s no clear value, if it’s a poster on a wall with a QR code, I don’t think many people scan those.
DAVID: Yeah. It’s generally kind of thought of to be a big fail but it’s really interesting that, okay, there’s a QR code that was tied directly to the potential to actually get a really different user experience, I guess. You know, that you’re really getting it kind of like a 10x improvement or the promise was there that if you scan that, you’re going to get a lot of benefits out of it.
STUART: Exactly. We also captured it by number and validate that. I’m not sure if it will launch soon but the ability to type your phone number into the debit reader as well instead of scanning that QR code. Well, that functionality is built and ready and waiting.
DAVID: Okay, that’s good. So just in terms of those three axes of acquisition, user experience, engagement and retention, is there a way you can kind of like score that on a 1 to 10 for each of those?
STUART: I think for us, in our group, user experience was the most important thing. There is always a clear value prop for acquisition, so user experience and retention, I’d put it quite a bit higher. For us, experience (UX) was like a 9. Retention was probably closer to a 7, and acquisition, for our team, was like a 1. That’s our team; if you asked marketing, it would be exactly the opposite. [Laughter]
DAVID: Yes, the bigger picture.
DAVID: That’s very cool. That’s great. That’s kind of like an experience of where you really got some upticks. Is there anything that you kind of feel as though you had a massive fail? It sounds like, you know, you went down a lot of blind path and things like that. But, yeah, just tell us of one fail possibly to do with the development side of things or user experience or something?
STUART: Yeah. From the user experience side, we did underestimate the number of users we would get by quite a significant margin, and that ramping up support on that is pretty hard. So when you go from basically zero to millions of users a day, getting your support right – and we launched with a few defects, nothing particularly serious, but really triaging support. Some of those poor support people begging us to close out some defects as soon as possible.
DAVID: Yeah, it’s funny. It’s kind of like the complete counter to what most sort of mid-tier retailers or mid-tier brands and startups have, which is they don’t actually have enough of a user base to get any statistically meaningful value out of it. You got the opposite problem, which is, you know, I’ve heard this with Google. Whenever they roll out a product, it just hit massively with a huge amount of people and you don’t get the chance to do many split tests, you know, if it’s around a defect that you aren’t aware of or things like that. That’s a completely different sort of problem to try and solve.
STUART: Yeah. We used a lot of tools to help us there, trying to categorize these things and look into them.
DAVID: You mean crash reporting or…?
STUART: No. It’s mainly through support you get, so taking support tickets and then we investigating the sessions that they had and diving in and trying to work out this problem that they’re having. They might be a support ticket saying, “I didn’t get the money on my gift card” or something like that. We’d have to go in and we might work out, well, actually, no. That’s because they clicked the button to, say, redeem it onto their “Blue Bird” card.
Initially in the first few weeks, looking and investigating something like that might have taken us 20 minutes per support ticket, and then we really worked on our tooling around being up to triage those much faster in finding exactly what happened and then making sure if there was some copy or some other tweaks that needed to be made to make it more clear and get them into the next release.
DAVID: Did you go down to button click instrumentation than inside the native apps?
STUART: No, not button click. We had most of the stuff, a lot of actions, resulted in service calls, so I’m really looking through Splunk and I’ll make sure and things like that, ad hoc to work out what’s going on.
DAVID: Okay, cool. All right, that’s very good. Let me just see what other questions I’ve got to ask you to close out because we’ve hit our 20 minutes. Anything you want to share about what you guys are going next, or you yourself?
STUART: Yes. I’ve done three years at Walmart. We built a lot more than just Savings Catcher. We built three other products on the eReceipts Platform. So it’s time that I’m moving on to the next startup and work in the health IT space. That’s exciting for me.
[At Walmart] We’ve built up a great team. I mean, it may have come off a bit like a couple of days to build all this stuff. It’s definitely not the case. The number of engineers just touching Savings Catcher would have been about 40 or 50, the number of user experience people ten, marketing ten. It’s just a huge effort and lots and lots of quality people doing cool stuff.
DAVID: So, yes. To me, from the outside, Walmart looked like they’re really taking the mobile “bull by the horns”. They brought everything in-house; they did Walmart Labs. Then later on, 12 months later, you see Westfield introduces Westfield Labs, and I think Walmart might have psychologically thought, “Well, if we leave this digital space too long, then Amazon will “eat our lunch”. For other retailers and big brands out there, do you think that ultimately the only way you can really actually get that kind of huge impact is like don’t outsource your mobile to an agency, or is that… should you in-source everything? Do you really need to throw everything, a lot of resources at it, as if your life depended on it? What’s the take on that?
STUART: Yeah, a really good question. I think if you’re talking specifically about in-store experiences, there are really great reasons to do them in-house. If you’re connecting to or need to modify things like point of sales or in-store systems, I think having consultants coming is not the right approach. In-sourcing is definitely the right approach there. I think Walmart took it with Walmart Labs to quite an extreme, and I think it’s working really, really well. Not only does Walmart Labs have a pretty positive brand in Silicon Valley and pretty easy to recruit for, I think they’ve been very effective especially on the mobile phone. Having said that, I mean, Tesco’s doing some cool things. We look at what they’re doing occasionally. There’s some good stuff there. But I think for Walmart, anyway, it has the money to spend. I think they acquired five companies, including us, in a very short order to create Walmart Labs, and it paid off.
DAVID: How high was the actual support for that?
STUART: There’s lots of CEOs with Walmart. So the CEO of the e-commerce business, Global E-commerce, along with an SVP, senior vice president, really formulated the strategy. I guess before we were acquired they went to the board, and the board liked it, and from then on, especially the SVP was a driving advocate of all of this stuff. Yeah, I think the Walmart Labs model, if you can afford it, is a great path.
DAVID: Yeah, I’m doing the numbers here. It sounds like about a hundred million. Am I cost-cutting?
STUART: I think publicly it’s probably a lot more than that to set up something like Walmart Labs.
DAVID: All right, very good. Thank you very much. I hope everything works out well for you in the next phase, this new startup in health. It sounds pretty interesting, so good luck with that. Thanks so much for sharing your experiences on that. That’s great. I might see if I can get you back some stage to hear a little bit more, if you don’t mind.
STUART: Sounds great. Well, thank you very much, David. I appreciate the chat, and hope that it’s useful to some people, and I hope I didn’t say anything that I shouldn’t have. [Laughter]
DAVID: We’ll find out. Cheers, mate.