Video Transcript

Hi, everyone. Thanks for having me here. I know this is the last session and there’s a quick wrap up, so I am officially between you guys and a beer on the rooftop. So, it’s scheduled for 30 minutes, but if you bear with me, I think I can do it in 20, and leave some time for questions at the end.

So, just before I start, big thanks to the Optimove team. I actually joined Groupe Dynamite June 17th, and I think we signed the contract by the end of August, right? Pretty close. So, we haven’t even gone live yet. We start the implementation in December, but hopefully, by the end of my presentation, you will understand why we acted so quickly and we signed up with Optimove.

After being here, super optimistic we made an amazing decision, and looking forward to work with the team and learn from them. So, without further ado, I want to ask a quick question, and I’m assuming not many of you will put your hand up, so you will not offend me. Who here has heard of Groupe Dynamite? Couple of Optimove people. If you were in Canada, at least 90%, if not 100% of you would have put your hands up.

So, we’re a pretty big retailer in Canada. We’ve been in business for over 40 years. We have over 400 stores, 200 of which are in the U.S., and we have over 5,000 employees. So, we’re a pretty big retailer. Our target customer, we have two brands. One is Garage, which, she targets 15 to sort of 22-year-old fast fashion, think you know, H&M, Zara, that kind of stuff.

And then we have Dynamite brand, which is when she gets a bit older, she starts to work. Dynamite targets the girl more 22 to sort of 35, with a bit more focus on workwear. But this video should sum up one of our brands pretty well.

♪ [music] ♪

– Cool. That’s Garage. Yeah, pretty cool brand. So, why was I asked to come speak here if we haven’t signed up with Optimove? Dynamite is currently going through a very large digital transformation. I personally haven’t seen anything like it and then I’ve worked for some pretty large retail companies. My background’s in telecommunications.

We called the transformation North Star. Basically, what happened about a year ago, kind of a marketer’s dream, the CEO went out to every single person in the organization and said, “Guys, give me a list of everything you wish you could have done but you don’t have the budget or resources to do it.” We’ve got a list of hundreds of initiatives.

Everyone wrote business cases and presented their business cases to the CEO. And from there, we prioritized and we created our workstream called the North Star. There are nine individual workstreams, ranging from pricing strategy, employee satisfaction, product marketing. And the one I’m going to focus on today is the digital workstream, which is actually part of the digital transformation.

And it’s funny what Boris was talking about, being best friends with your CTIO. Well, my CTIO is actually my boss. So, we worked together at a prior retailer. He joined Dynamite about a year ago, and he brought me in. And the reason he put me under himself is exactly what Boris was… We have such high velocity of programs that they wanted the marketing team under the technology team, under IT, so we can move faster, be 100% collaborative, and break down these barriers that Boris was talking about between marketing and IT.

And I can assure you it’s working really well. So, there was three main goals in the digital transformation. One is to take that brand that you saw and take it to the next level. We want to be the coolest, most relevant fashion brand in North America. We do not use the word retail. We completely dislike the word retail. We want to be a brand.

Retail has lots of negative connotations. People tend to focus on stores, things like that. So, we want to be fashion brands. We need to empower our employees because without your employees, you’re not going to deliver your results. And lastly, of course, you know, financially, we want 50% of all of our revenue coming through digital channels, which is a massive… There’s not many fashion retailers who are anywhere near 50%.

Most retailers, probably 10%, 15%, maybe 20% max of their revenue’s coming through digital. We want to be at 50% within the next 3 years. So, my workstream…as I said, there’s nine workstreams. Each workstream has a sponsor, who then rolls up to an executive committee, who effectively manage this transformation.

So, I head up digital. Under digital, I have four workstreams, which I’ll talk you through in a minute. I have over 38 separate initiatives. Just to give you an idea, Optimove, rolling out Optimove is one initiative. Rolling out our new website, the Salesforce Commerce Cloud, is another initiative. So, these aren’t small initiatives. They are big meaningful initiatives.

And as a result, I have about 20 people under my team, who each own these initiatives. And I’m going to get into it in a bit more detail about the process and how these teams works because they still have to do their day jobs, yet they’re also been asked to deliver these huge ass strategic initiatives. So, just quickly to talk through the four key pillars.

We have a workstream on customer acquisition, [inaudible], paid media, you know, Google, all the usual channels. But again, we’re bringing a data-driven approach to everything we’re doing. So, as an example right now, paid media is always a hard one because, you know, you want more money, but then you get challenged, “Was it really incremental?” So, the way we went about proving this was we took Canada and the USA, and we split them into three zones.

Each of the stores and cities were mapped against these three zones. One zone we called the blackout zone, and we turned off all of our paid media in that zone. Risky, right? But if people don’t believe in paid media, what’s the best way to prove the value than shutting everything off and then your store starts screaming for it? You kind of prove the value of it before you get too far.

The second stream was, like, business as usual, and then the third stream, that’s where we started laying our additional budget and money to prove the incremental ROI. So, again, everything we do, we take a data-driven approach. We don’t just loosely go and spend more of the money. Conversion was mostly around online optimization. We use a tool called Dynamic Yield, which I’m sure some of you have heard of.

There’s another tool called Optimizely. And it’s pretty straightforward. When people hit the website, it funnels half onto, like, one view and the other half see something else. And through aggressive testing, you can quickly work out, you know, showing, you know, one group of customers this versus that makes more sense. You test it for 7 days to 10 days, prove it statistically significant, and then you roll it up.

Loyalty, CRM, this is really what Optimove is here to help us with. So, I’ll talk about that in a minute. And then the foundation piece, all this technology and all the fun marketing stuff on top only works if you have a core foundation. So, as I said, as well as all the new marketing tools and technology, we are rebuilding our core platform. So, CRM, databases, Commerce Cloud with Salesforce, Service Cloud with Salesforce, rebuilding our mobile application.

So, there’s a huge program of work, and the breadth of this program is pretty significant. So, how does this all work? So, as I mentioned, we have workstream leads, and then each workstream lead has what we call a pizza team. Enough people to sit in a room and have pizza over lunch. So, we want to make sure everything’s collaborative and cross-functional. So, under the CRM portal, someone in my team would probably lead that initiative because it’s a CRM program.

And then under them, they’d have a data scientist to run the analytics. They would have someone from e-comm. They would have a designer, a copywriter, basically everyone they need to move fast and agile is in that team, and they all work cross-functionally. So, basically the way it works is you come up with the ideas, you prioritize the ideas, you plan your sprints because some of the tests are going to take longer to do than the others.

And then they go into a two-weeks period where they’re actually testing. And then they come out of those tests, and they analyze the results, and they start again. So, every two weeks, they’re testing aggressively. So, you know, all your batch-and-blast emails are going out. But as well as that, they’re testing against the different segments, testing the, you know, different creative, you know, any number of different tests.

So, it’s a lot of extra work for the team. And again, when they’re doing their day job, if you don’t do this, you will fail because they just don’t have time to… You literally have to lock people in a room. It’s like, “Guys, this is your number one priority. You have to focus on this,” and take the pressure of their day job away from them because it’s a hell of a lot of work. So, this will probably look familiar to some of the sessions we saw yesterday. So, again, we don’t do anything on gut-feel.

Everything’s based on the data. Everything’s based on testing and analytics. So, whether it’s an email test, or a website test, or, you know, any test for that matter, the way it basically works, you have your entire audience or database. Let’s pretend this is an email campaign. Okay. So, you have your entire database. Then in the pink, you have the group of people that you think is relevant for this particular message.

So, let’s pretend the email is going to sell…you know, promote our new dress category to our database. There may be people on the database that we don’t want to see in there too, you know, for people who are on journeys, new customers, whatever. Then once we have the total group, we take a smaller group of that and the two groups are statistically the same. And one is a control group, which gets the regular email, and one is the test group.

And then we test these over the next two weeks. And again, throughout the test, we look at the results of the lift. At the end of the two weeks, if we’ve seen success, we say it’s successful, we check, and then we scale it up. I’m sure those of you that use Optimove know well that’s what we’re doing. Well, we’re doing this today manually without any technology. So, you can imagine how hard this is.

I had to hire seven data scientists to enable us to do this. So, here’s an example, and this is actually real data, by the way, so feel free to take this test and try it for yourselves. So, this is the website. Basically, before, this is the new arrival section, the web merchandise team, who had been doing their jobs very well for the last few years, every time they put sale up the top, they would get more clicks and they thought it was a really good thing.

And the new products at the bottom, sorry, the bottom of the email. So, we said, “Okay. That’s interesting.Let’s switch it around.” So again, half of the customers would see what was there before, which would be the sale box at the top. And then the other half now saw the new arrival. So, let’s have a look at the results. So, we run the test for 14 days. The test group had a 3.62% conversion rate.

The customers that saw the new arrivals had a 4.3% conversion rate, and with the statistical analysis, we said that that ran at a 95% probability. So, that’s a check. That was a successful test. Sounds familiar, right? Someone, I think, talked about this yesterday. AOV, that’s the second key metric, because your revenue is obviously your traffic, times your conversion, times your AOV.

So, AOV, in this case, also went up 2.3% at a 80% confidence level. So, we say, “You know what, we’ll bank that money too?” Check, and we have a lift. And I obviously removed what the total revenue impact is, but… You know, you do that over a hundred million users and the money adds up pretty quick. I forgot to mention, everything we do through this initiative gets, you know, tested like this.

And then once you got a successful test, at the end of the two weeks, you sit in front of the CFO, you show the results. And if he believes you, check, and it goes into the incremental target that we’re all trying to hit to get to that 50%. Here’s another example. This is really straightforward, guys. There’s money all over the table when you start testing. So, before, the image on the website looked like that.

You know, we wanted to show the girl, she’s really pretty. Great. We said, “Hold on, what if we showed more of the product?” Because if you look at the first picture, it’s mostly the girl, right? You don’t really see the product. So, in the test, we actually just shrunk the image. So, to bring up more of the product and now you get the full product shot.

Okay. It makes sense. Let’s look at the results. So, in this case, no impact on the AOV. Actually, that’s not true. It was 2% higher, but it wasn’t statistically significant. So, we couldn’t bank it.

So, we said, “No impact.” Okay? But there was a significant lift on the conversion, 4.7%at a 92% confidence level. So, that’s another example of a successful test. Okay. So, we rolled that out. One click of a button, that rolls out.

So, that is now live every time you hit the website. So, the technology, as far as we’re saying, enables the marketing strategy. So, the ideas are there, you need the technology to deliver. And I think the last test is also a super simple one. So, everyone knows website speed is absolutely critical to conversion, to drop-off rates, etc.

So, you know, we’re the first to admit our website, you know, all will be replaced in the next year. So, we’ve teamed up with a company called Yotta. I’m sure some of the retailers in the room use these guys or know who they are. And their thing is that they can help speed up your website. So, again, we didn’t want to sign a contract with these guys and just, you know, jump head first into a relationship. We A/B tested the site, and we saw the site speed improve by 0.7 seconds, 0.68, if I added that up correct, which doesn’t sound like a lot, but in the world of e-comm, that’s a heck of a time.

So, then we have a look at the results and we see a 4% lift in our conversion rate due to the site speed. So, that’s just three examples of A/B testing. I use the website one because we have much more velocity on the website testing at the moment because we have the right tools to do it. And this is what we need Optimove for, is to help us get more velocity in our CRM testing.

So, this is what it looks like today. So, we have a loyalty program, very successful, and we have over 2 million members in the program. I think 70% of our entire revenue for the business is with the loyalty program. So, that’s great, check. But when they set the loyalty program up two years ago, they didn’t really put enough investment into the data. So, for my team today to pull a basic CRM campaign, take the red dress example.

Let’s send an email to people that shop red dresses, the time it takes for the team to do that. They have to manually extract the data. They have to probably reach out to the BI team. They might be busy. I have to go knock on someone’s door to make them do it. Three or four days later, they might get the list, they send it out. Great.

Okay. Ching. Job one. But then you have to track the results. So, then, you know, two weeks later, you look at the test in the control group, the team go back in to analyze it, and they realize that four days’ data’s missing because there’s no BI data governance. There’s no one looking at the data. Back and forth.

They have to work with the BI team. You know, three, four weeks later, we get the results, which isn’t very practical. Right? And if you want to test of the velocity we need to test, it’s not going to get you really far, which is why we signed up with Optimove. So, why did we sign up with Optimove? Firstly, as I just said, we didn’t really have a CRM system. So, rather than signing up with a, you know, archaic CRM system, there’s lots of them out there.

I’ve had them before. They’re not much fun. We thought, “Let’s go with the best.” We get all that data in one place. Plus, we get all the advanced models and analytics that come out of the box. I can only speak for retail but, you know, their retail models are, you know, they’ve got one-time buyer models, churn risk models. I’m sure they can help us migrate our customers from one brand to the next.

And, you know, I’m assuming some of you guys are using this stuff today. So, we’ve gotten pretty excited about the models. But then, for me, the big one is the next two, is the ability to do this cross-channel marketing very quickly. So, pull holistic customers. Send a control group, deploy the campaign, whether it’s on email, Facebook, direct mail, SMS, push, and then let Optimove do all the work.

The data comes in every day. And then, you know, marketing simplicity. You log into the platform, two weeks later, you can track your results. In fact, you can probably check the results after three days. If it’s not looking good, you could turn the campaign off. So, we’re really excited about this velocity that we’ve got to get with Optimove.

And then, of course, the last piece is through more testing. I think over 50% of our tests fail today, actually, which is really good because when you fail, you learn. And a lot of the tests that we fail on are things that we would have done anyway. So, someone from e-comm says, “Let’s do this.” And in the old world, they would just say, “Yup, sure.Let’s do it.”

Now, my team will say, “No, you can’t do that without testing it first.” Half the time they’re right. Half the time they’re wrong. But if you get it right 50% of the time and you get it wrong 50% of the time, chances are you won’t get anywhere. Right? You’re losing money here, you’re getting money there. So, by not scaling the failed tests and only scaling the positive tests, you’re more likely to have a positive lift throughout your program.

And then the last slide… And I’ve still got 11 minutes left, so that’s good news. Is just to show you guys the velocity of what we’re doing in such a short space of time. And the only way, again, to do this is with executive-level alignment. Take…

Everything I said in the beginning. Taking people out of their day jobs, telling them that this is a priority, and then putting the right resources and budget behind it. So, you know, we knocked on our CTIO’s office and said, “This is what we need in the next 12 months,” and this is what we’re getting. So, we’ve already rolled out Dynamic Yield, which is all the A/B content testing. Service Cloud launched two weeks ago. So, now our call center can access our 360 data.

We’re going to enable chat and other features within the call center so we can deliver that right level of customer experience. Yotta went live, I forget now, two months ago. That was the speed testing that improved the speed of the website. I think Optimove is up next. We’re going to try and be live by February, so we kick off the project once we get back from those lovely conference.

CrowdTwist, I don’t know if anyone’s heard of them, but they’re the number one loyalty platform in Forrester. So, yeah, we’re relaunching our entire loyalty program. That’s another initiative. That’s a huge initiative. Some companies would take two to three years to do that. We’re also signing up to the buy now, pay later option, where you can check out and not pay anything upfront, and then you set it up on Storm, as I heard someone talking about that this morning.

Commerce Cloud I’ve already talked about. And just to sort of throw it in for good measure, we’re also ramping out our entire HR processes because you can’t get very far when you’re working with archaic HR processes and your people are getting caught up, you know, manually doing expenses, and things like that. So, that’s probably just a flavor. There’s probably another 20 initiatives that we’re working on. But all that’s going to get done in the next 18 months.

And that’s what a digital transformation looks like. Excellent. Now, it’s time for questions. I don’t know if anyone’s brave enough to ask any questions, but we can talk about the rugby. I know there’s someone from Ireland here. Congrats to anyone from England.

They did not deserve the World Cup. Cool. Questions? Yes, sir.

– So, how do you account for the differences in the product? So, you said, one of the examples was you moved the new arrivals lower down. What if those new arrivals were better or worse than previous new arrivals?

– Well, in this case, it wasn’t whether they were better or worse, it’s just that the visibility of the new arrival product was so far down the website that people weren’t scrolling down. So, by increasing the new arrivals above the fold, we got more clicks on the new arrival section. Obviously, your new arrival is more likely to be your best product at a higher price point.

So, we found that by doing that, we got the right lifts and the right clicks. So, yeah, it did compensate our sale items. But if you’re in retail, you know, you want to be selling as much of your full-price product as you can, as quickly as you can. Any other questions? Great. Oh, there’s one here. –

– Hey. So, you’re obviously just targeting Generation Z. So, in regards to that, how many, like, different segments do you believe exist within that?

– That’s a good question. I can’t really answer that right now because I don’t have Optimove. Look, when I joined, they were really not doing any segmentation, to be completely honest with you. [inaudible], you know, the typical retail, one email a day, same email to everyone, so that’s kind of where we are. We’re obviously testing against that. So, the first thing we did was our clearance segment. So, we have identified customers that shop clearance.

We did a DSL analysis, which basically means you put your entire database into DSLs, starting with the best customers at the top, based on, you know, percent of their purchase, you know, on new arrivals. And we quickly found that the customers in the top DSL not only bought reg, but they spend way more money. And then right at the bottom, and we love all of our customers, don’t get me wrong, but right at the bottom of the DSL analysis, the 10th DSL, were these customers that only shop clearance.

So, we’ve started that. And then we’ve tested against that and we’re trying to figure out what’s the right mix of sale messaging against each of the segments. Because, you know, even your best customers, you know, if you can get an extra spend out of them by adding a clearance product to their basket, it’s not a bad thing. So, that’s one thing we’re doing. And then the other thing that we’re doing right now, which, you know, I don’t know who lives in North America, but I’m from Montreal, Canada.

And if you go there in the winter, it’s, like, minus 20, minus 30. Don’t ask me why I moved there. I’m from New Zealand. It’s fricking cold. And right now, our customers… And when it’s cold, the head office is in Montreal… You know, the designers are cold, so they add the snowflakes to the emails and, you know, the girls are wearing all the outerwear, and we send that same stuff to the girls in California, which makes absolutely no sense, right?

So, that’s kind of the next level of segmentation is, “Okay, we have 40 stores in California.We know they buy a different product. Why don’t we take the time to design the emails?” So, that’s our next thing. But, you know, we’ve talked about it. You start small, one segment at a time. You can’t change the world in one day. So, we’re testing against that. But, you know, ask me that question next year.

Hopefully, we’re doing 30, 40 segments. I don’t know. Yeah. Cool. One more question, yeah?

– Yeah. So did anyone…? So, she asked, “Do you feel it’s aggressive?” And the answer is absolutely. Good news is my Boris, he’s on the hook to deliver. So, no, like, jokes aside, we bought him 40 developers, 25 developers, 7 data scientists.

You know, we have a check, like, literally an open checkbook. If we’re working on initiatives and we have a strong business case, we will move fast. So we have all the support all the way up. There’s no politics. Everyone’s behind this. It’s absolutely the key initiative for the company. So, when you have full alignment from the top down, it is easier.

Of course, it’s still stressful. And, you know, you wake up and, you know, you’re a week late here and a week late there when you have so many programs running at the same time. The goal is to try… Like, next June, we have to launch a loyalty program, the CDP, e-comm, all at the same time, because one of the things we don’t want to do is implement some of these products on the old technology and then have to do it all again.

So, that’s actually the tricky part, is the timing of all these initiatives to go live at the same time. And if one of them slips, you potentially slip all the projects. So, that’s probably the hardest part of it is just getting the timing right, you know, juggling all those balls at the same time. Yeah. Cool. One more question.

– If I remember right, you said that you have 400 stores, right?

– Yeah. Four hundred stores. Yeah.

– And you said in 3 years, you’re going to have 50% is going to be online to sales. So, how many stores do you think you’re going to close or we are talking about new customers?

– Yep. We’re going to open more stores, to answer the question. And let me… We’re growing aggressively. So, let me clarify. I did say 50% digital, but I did not say 50% on e-comm. So, let’s think, what does that mean?

So, iPads in the store. So, we’ve just rolled out iPads to all of our associates in the store. We’re launching online exclusive. So if a customer walks into the store, maybe she’s plus size, you know, we don’t have a lot of that stuff in the store. The associates will now be incentivized to go online and, you know, through client telling them how to sell them a product online. So, I expected a huge portion of that digital to come through the stores.

You know, we’re also looking at how the stores can help us with our return processes. Our return process today is not customer obsessed, so we want to make sure that there’s an easy return process. And, you know, we’ve only just rolled out Ship from Store. So when you buy on e-comm, now the stores are also helping with fulfillment, which most retailers are going to go, “We didn’t stop there.We took it to the next level.”

So, rather than having an e-comm order… So, I mean, hopefully, you know, most retailers, when you buy online, believe it or not, a lot of the products get shipped from the stores. Okay? Our warehouse is in Montreal, Canada. We have customers in California. It takes about 10 days to send the product over. Again, that’s not very customer obsessed.

So, now, we send the items directly from the nearest store in California. But with the data science team, we took it to the next level. And again, given we’re big on technology and gamification, we basically created an app. We call it the Uber app. It’s not branded Uber because we’d get in trouble for that, but it’s basically the Uber app. And just like when you want to get a taxi and you look up Uber, you see the five drivers, right, and it sort of crowdsources and says, “Who wants to pick up this order?”

The AI, our data science team have built a model that sends out this request to five-plus stores that have the inventory on hand. There are about three or four rules. It’s not just imagery. It’s, you know, what’s their turnover of the inventory? How far are they from the customer? And then it sends out a request to the stores and the stores get incentivized on how quickly they can pick up the order so we can get out to the customer quickly. So, we’re just tearing everything up and, like, building everything from scratch.

So, digital is really omnichannel. Ninety percent of customers will start their journey on the website or maybe on a Facebook ad and come to the website. So, digital, yeah, let me clarify. It’s the overall commerce or digital experience, but again, that includes our associates in the store. I think I’m out of time.

Cool. Thanks, everyone.