Connecting Shoppers with Relevant Products with Crate & Barrel
Hear the hows and whys of Crate & Barrel’s SLI search solution replacement journey and how their move to Lucidworks solutions impacted customer retention, loyalty, and revenue growth.
Intended Audience
Tech and business leaders looking to leave customer pain points from your legacy platform behind.
Attendee Takeaways
Learn about the Crate & Barrel journey and get a deeper dive beyond the traditional use case of search.
Speakers
Peter Curran, General Manager, Digital Commerce, Lucidworks
Aaron Veit, Aaron Veit, Director, Digital Product Management, Crate & Barrel
[Peter Curran]
Hello, everybody. Thanks for joining us. My name is Peter Curran and I’m the General Manager for Commerce at Lucidworks. And today we’ll be talking with Aaron Veit, who’s a director at Crate and Barrel in the Digital Product Management group, about how Crate and Barrel is connecting shoppers with relevant products. Hey, Aaron.
[Aaron Veit]
Hey, Peter. Thanks for having me.
[Peter Curran]
Thanks for coming. This is what we’re gonna talk about today. We’re going to try to keep it an informal conversation, but talk a little bit about who Crate and Barrel is as a business. Why Crate and Barrel partnered with Lucidworks, go over the project that we did together, and the outcomes associated with that project. So first, Aaron, maybe you could just tell us a little bit about Crate and Barrel.
[Aaron Veit]
Yeah, so Crate and Barrel Holdings is the overseer of both Crate and Barrel and CB2. We have websites in US and Canada, and I have been with Crate and Barrel for a little over three years now.
[Peter Curran]
Cool. And what’s something about working at Crate and Barrel that people might like to know? What’s it like to work there?
[Aaron Veit]
We’ve got a pretty cool headquarters up in Northbrook, Illinois. It looks a lot like a lot of our retail shops. What is very cool about our headquarters is we have a test kitchen and what appears like a fake store inside our corporate headquarters; which allows all associates visibility into some of the new items before they hit the stores. So, really cool to be able to walk around a store inside headquarters and get to see all the new things before they’re are out in front of our customers.
[Peter Curran]
I’ve been once or twice. It is a really, really nice corporate headquarters. This vignette style that this particular slide is talking about is I think kind of represented on the slide itself here, where we see the sofa with the pillows, and the art, and the background, and the plant, and the bowl, the books, and all the rest of it, the rug, all of that stuff of course available from Crate and Barrel and all works together in that room. I remember shopping, you know, when I was much younger at other places where they used to just kind of group, all the sofas were in one place, and all the chairs are in another place. And there was a place with just pillows, and it’s kind of hard to see how the room comes together when you’re shopping like that.
[Aaron Veit]
Yeah. No, absolutely. And it’s definitely changed my personal shopping. You know, being more out in front of the product and getting to see it on a daily basis. It’s very cool stuff.
[Peter Curran]
All right, cool. So this conversation today is about the search experience on the Crate and Barrel websites. And this screenshot is taken from before you did your project with Lucidworks. So, maybe just tell me a little bit about, you know, what kind of experience were your customers and your associates having with your previous technology, and this wasn’t your first change, was it? You’ve been on other platforms previously as well?
[Aaron Veit]
Yeah. We have made our way through a couple of search providers over the past few years. We were using Endeca, then we migrated over to SLI prior to moving on to Lucidworks. And we knew that we needed to find a new partner that aligned with some of our objectives. You know, this, the search result page is not necessarily what we want to be serving up to our customers. So, you know, we referenced this often during the project for how we could get better.
[Peter Curran]
Yeah. You know, you can see in a page like this the logic of the search engine at work, and it’s a particularly deceptive example because here I have searched for magnet board and the search engine has chosen to drop the word magnet and only search for the word board, and then give me these different cheese boards and things, which are all obviously irrelevant. The problem is, the sad thing is, that you have magnet boards, right? You have this product then, you know, the user has kind of maybe led to believe that you don’t have that thing for them and potentially they’re going to leave. So, it’s a serious consequence. So maybe you could talk just a little bit about, you know, what that was like in your previous platform. It looks it like had limited machine learning features. Why does that matter? Why do you need machine learning features in a search platform today?
[Aaron Veit]
Yeah. I mean, we have a large and very talented merchandising team, but as we all know, there’s only so many hours in the day. You know, we have a busting SKU count that continues to rise as we continue to expand our assortment. But in addition to having a robust assortment that comes in many, many different customizable colors. So it’s, you know, really challenging for the merchant team to be able to stay on top of merchandising, and being so granular, and manually updating search terms along the way. We have product categories for both indoor-outdoor, all aspects of the home. So it really puts the burden on our merchandising team in order to be continuously making updates, and enhancements, and improvements to search results. So we knew that we needed to quickly be pivoting over to machine learning as our SKU count continues to grow.
[Peter Curran]
Yeah. So it kind of gives you the ability to scale and keep up with your assortment breadth. You know, you guys sell something for every room inside and outside of a home, or for business for that matter, and so, you know, I could see why it’d be difficult to keep up. The other side of that though is something I wonder about. So, it’s like, you can make the argument for machine learning, but then what is the point of the merchandising tool? Why, if you, if you want to kind of lean into automation, machine learning, why do you still need the merchandising tools?
[Aaron Veit]
So, with new products hitting the site all the time, we do want to make sure that some of our new product launches are bubbling to the top. We have new collaborations that are constantly hitting. We have seasonal items for, you know, everything from both holidays, and then for summer and winter, and so on. So we do want to make sure that we are relevant with the times, not only to show off the new products, but also the seasonal products. There’s a different product that a customer is gonna be searching for during the summer versus the winter. So, it is important to still have some level of merchandising capability.
[Peter Curran]
Yeah. I think the other thing too is that you guys are trendsetters, right? And so there are places in pages and parts of the user journey where you need to be very deliberate about inserting the merchandising team’s perspective on where the user is at in that journey. Kind of reminds me of the, you know, the self-driving car that you or at least I wouldn’t get into a self-driving car unless it also had a steering wheel and brakes. Right?
[Aaron Veit]
Yeah, absolutely.
[Peter Curran]
So what did the result of the project look like? You know, did you guys see any kind of commercial benefits from switching to Lucidworks?
[Aaron Veit]
So I’m not able to divulge too much of the specifics around the improvements, but I can tell you that we have noticed a higher conversion. There’s an increase in add to carts and order value, as we are continuing to serve up more relevant items to our customers.
[Peter Curran]
Yeah, it makes sense, right? If somebody, if you have a thousand people who come looking for a magnet board, some percentage of them are gonna leave. They’re not, even if they were there to buy, they’re gonna get flummoxed by the steps that you need to put them through in order to find the product. So it makes sense that you would see a better commercial outcome. So, maybe talk to us a little bit about what the process was like. So, it looks like from the timeline here that everything started for you in January of 2020. So you, I guess, went through your holiday peak period in 2019, and then started the process in 2020. How did you go about looking for a new vendor?
[Aaron Veit]
Well, we knew at that time that we were ready to make a change. The direction of our previous vendor wasn’t necessarily aligned with our goals. Their roadmap wasn’t shaping up how we were hoping, so we wanted to look for a new partner that did align with our objectives and was focused on enhancing the customer experience, and was focused on continuing to build out the roadmap to offer us opportunities to grow our search business. So it was time for us to start looking out into the open market. So, we had a formal RFP process, went out to more vendors than I’d like to admit, and, you know, we were able to have great meetings and conversations with the Lucidworks team, and that’s when we were able to start making some great decisions to partner with Lucidworks.
[Peter Curran]
Well, we appreciate the partnership. One of the things I see on this slide is talk about a proof of concept that you did about a year ago in the fall. Why did you want to do, vendors like to avoid proofs of concept. They like to go straight into the contract. Why did you want to do a proof of concept?
[Aaron Veit]
It’s nice to see a vendor’s product work with your own data You can see if the current issues you have with your old platform manifests itself in the new platform. So, you know, it’s nice to see features and functionality on a slide, but really taking it for a drive is where you get an opportunity to see how it really works out with your own data.
[Peter Curran]
We have had this kind of session at previous Activate events. Prospective customers like to come to this kind of session and kind of see what the experience of another customer of Lucidworks was like. Man, you guys got live pretty quickly, right? Looks like you were able to get everything up and going in something like three months. Maybe you could just tell us a little bit about what that project looked like. What were the things that you were thinking about and working on as you did the project?
[Aaron Veit]
Yeah. I mean, we, like you said, we had a really ambitious plan on getting all four of our sites live in a very short period of time. So, we lined up all the right internal and external resources in order to achieve our lofty goals. We were meticulously documenting along the way to make sure that we were prepared for when we were to start the project because had we not put in all the prep work and had all the dialogues leading up to the project kickoff we would have been behind the eight ball before we even started. So, we did our due diligence to make sure that we were as prepared as could possibly be, both from a resource perspective, but also documenting everything that needed to be documented for the kickoff, and ultimately start getting the work under way.
[Peter Curran]
Yeah. And the, you know, I think that some of the things that you were able to keep as part of the old system, like the way you were sending feeds in and things like that, those kinds of, if you can kind of count on some of those things as assumptions going into the project that they’re not going to change. That it kind of gives you a much better likelihood of a timely success than kind of tearing everything apart.
[Aaron Veit]
Yeah, absolutely. So we, you know, we took a real deep dive into how the data and signals needed to be delivered. There was a huge difference that Crate had data delivery information before our discovery kickoff. Then at our discovery workshops, both the Crate and Lucidworks teams were well-prepared. We came to the table with the right resources, as I mentioned, and, you know, both teams were asking the right questions. The integration communication and meeting structure worked really well. All team resources met twice a week to cover any outstanding questions and action items. We also, you know, I know both teams had internal, daily stand-ups and would add meetings as needed to make sure that we kept on pace to hit our dates. Crate had, you know, a lofty task plan designed and ready to go. And 10 days prior for both launches, we scheduled additional check-ins to make sure everything was set up. No stone was left unturned. We were ready to go.
[Peter Curran]
You were on it, right? You’re kind of managing it super closely and, you know, I know from the vendor perspective, it’s really helpful when the client kind of meets halfway or more than that, and kind of bringing the management and oversight and cadence to the project. You’re just about to go into it. I love it when our customers have steering committee meetings. I’m frequently able to participate in those from the vendor side, on the Lucidworks side. A lot of companies don’t do them. Looks like you guys were able to kinda, to keep those in your project.
[Aaron Veit]
Yeah, absolutely. You know, these were great meetings about milestone achievements, project phases, what to expect next. So we never had to guess what had been achieved and what was coming next. Then for the launch day, all hands on deck, make sure that everything went smoothly and fast files and issues were addressed immediately. So, the Lucidworks team was available to us to make sure that they were answering questions as needed and they continue to do so.
[Peter Curran]
And Lucidworks fully manages and runs all of your infrastructure and all of that kind of stuff for this as well.
Great. So, let’s look at after. So we talked about magnet board before here. You can see that there is in fact a magnet board. Kind of interesting that it suggests some other terms. Some of those are more relevant to one of the words; in one case, magnet, than they are to the word board. But the important thing is that the reversible whiteboards and the kid’s easel come back, showing the customer that you’ve got something for them.
When I look at another example of a before and after this is a kind of an interesting one, “boucle”. Not a word that I use every day. I bet you have a lot of words like this that are not necessarily part of everybody’s vocabulary, but like if you’re a designer and that is the fabric that you know is going to provide the right kind of texture and appeal to your client, it’s really important.
And so you can see at the top here before, it appears that it’s kind of a little bit hard to see, but it appears that it probably came back with some of the right products. Maybe the first one is a boucle product, but the others are velvet. Certainly the blue ones are velvet, not relevant.
And then in the after, it’s all boucle. So you see this kind of outcome. You’re giving people the breadth of your assortment with that boucle feature so that they can now narrow what they’re looking for and drill in.
Let’s look at this one. This is a search for “pampas”, and to my eye, I don’t see a huge difference at first glance. Pampas, sort of these big feathery plumes. I think it’s a grass. I don’t think it’s a feather, but kind of looks like feathers. And, you know, you’ve got them, it looks like they can be stuck into like a frog type thing, as well as arranged in a wreath. The only difference I can see between the first one and the second one is that, is that the first, the before, has the vase. Why do you think it’s important to be that specific about your sort of your critique of what search should come back with?
[Aaron Veit]
Yeah, I’d say one of the more important elements of the search experience is the filtering. Is the filters and sort features. These can easily get messed up if inaccurate search results are served up to the customer. So, being able to help customers narrow these results lead to improved conversion.
[Peter Curran]
Yeah, so you could imagine if there were not just one vase, but there were, you know, 400 vases and 11 pampas-relevant items, all of a sudden all of those filters in sorts would be all messed up. I wouldn’t be able to sort from lowest price to highest price, and actually hope to only look at the pampas products. So, cool! So, maybe we could talk for a minute around what’s new on the site from your previous platform, and then just a minute or so about what’s coming next. So, you’ve added search with fusion. It looks like you plan to add browse as well?
[Aaron Veit]
Yep.
[Peter Curran]
And then maybe you could talk just a little bit about type ahead. Product suggestions is something that you have seen in the type ahead experience for some sites for quite a while. It looks like that’s new to Crate and Barrel. Why did you decide to add that? Why do you think showing products in type ahead is an important thing to do?
[Aaron Veit]
We like to think of ourselves as an inspirational brand. So, it’s probably better to show that the customers what the options are. What else is out there. People do react strongly to our kinds of products when they see them. And also, it tests better. So, you know, it’s a nice way to drop the user to the bottom of the funnel, to a page where they can convert. So, this is more convenient for the customer and it cuts out the possibility that they’ll get distracted, in filtering, and sorting on the page. So, it’s really a nice feature.
[Peter Curran]
It’s interesting too that you have query for query recommendations on the search results page. I’ve seen that feature before in a few sites. I know eBay has used it for quite some time. Kind of really, really huge assortment sites, like Amazon and eBay tend to have that kind of thing. Cool to see it come to Crate and Barrel. Why did you guys decide to do that?
[Aaron Veit]
Yeah, I think it’s helpful for our customers. It gives them ideas for how to further explore. The site might be more relevant than whatever terms occurred to me the first time I search, and might also be, you know, give the customer better ideas, different ideas of other ways of searching for an item.
[Peter Curran]
I think it kind of gives the consumer a little bit of a feeling that they’re on an adventure, that they’re on kind of a treasure hunt, and they are being helped by the site in their journey to kind of discover different kinds of products. So, it’s helpful both when I’m on a mission to get something specific, as well as when I’m really just looking for inspiration and not necessarily looking to transact right now. So in terms of what’s next, I see some improvements to user experience, some backend things that you’re doing, like being able to experiment with one model against another model. Maybe you could talk a little bit about that last bullet there. What are you guys doing around omnichannel and why does it matter?
[Aaron Veit]
I mean, omnichannel is so important for us right now as the retail industry and, you know, all industries for that matter, are hit with the supply chain issues. We really want to give our customers the ability to shop where they want, how they want. So, being able to let customers know which items are available in stock, in stores, which items are available for buy online, pickup in store, which items can be shipped to the customer from a store. So being able to serve up these types of opportunities to our customers is really key for us going into peak. You know, as we look ahead to what’s to happen next with the supply chain? So, being able to offer different types of experiences for customers to shop where and how they want is super important to us.
[Peter Curran]
Yeah. I think I’m just speaking as a consumer, especially, I mean, in a pandemic, I want to kind of get in and get out. Or I want, if I’m going to get in the car and go down to the store, I want to know that I’m gonna walk away with the thing that I need, so I find it to be an incredibly useful feature and, you know, people who buy things online and pick them up in the store almost universally end up spending more when they go into the store. So it’s a sound commercial decision apart from just a really nice user convenience, so that’s great.
Well, that’s all the content that we have today. I know this has been useful for a lot of people. Aaron, thank you so much for your business and thanks for agreeing to do this with us. We really appreciate it.
[Aaron Veit]
Yeah. I really enjoyed the conversation. Thanks for having me, Peter.