Online grocery took off in 2020, and the influx of new visitors, coupled with supply chain uncertainties, made customer experience more crucial than ever for these retailers.
Thrive Market, a membership-based digital grocer, embraced the challenge. For our recent “Ecommerce and Retail Customer Experience 2021” report, we spoke with Sasha Siddhartha, co-founder and CTO of Thrive Market, about how the company uses first-party data to improve personalization and, in turn, drive customer loyalty.
What does customer experience mean to you personally?
I run the product technology for Thrive, and my role is connecting the dots—business strategy, consumer experience, and the technology strategy to deliver them.
If I were to zoom out all the way, customer experience would be every interaction that a consumer has with a brand or a business. Through my lens specifically, I look at it as all the interactions a consumer is having with us primarily on the online platform.
[These interactions take place] in our web and mobile applications, but also through other communication channels, whether that’s email, SMS, our member services contact center, or the self-assistance programs we have in place. Every one of those touchpoints is part of the customer experience and journey.
From a business perspective, our role is to make sure that experience—of interacting with us as a business and a platform—is as useful, delightful, and high-quality as possible, with the maximum number of aha moments and the minimum number of problems.
How has the pandemic affected Thrive Market’s growth trajectory?
It became the moment for our mission. Prior to the pandemic, we provided a great deal of convenience to our member community. But after stay-at-home orders [were announced], we were in a position to keep people safer and make sure that, wherever they lived, they had access to their healthy essentials, delivered to their door.
It was a galvanizing moment that allowed us to be the most focused and the most customer-centric we’d ever been, during a period when we were all feeling a higher level of stress as employees and in our own lives.
What’s really exciting to see, even as things start to go back to normal, is that a permanent habit shift has happened for a lot of folks. Everyone’s had a chance to consider the online grocery category, so the level of user education we needed to do with new customers, that problem’s now been solved.
How have you created better customer experiences throughout the pandemic?
We put our member hat on, and put our human head on a lot.
We were facing dramatically accelerated demand, combined with unprecedented instability in the supply chain—order delays, shipping delays. We had to decide whether to continue to let customers go and take as much revenue as possible. Or, we could strike a balance and make sure our members had a positive experience and reasonable SLAs [service-level agreements] for when they would get their orders.
It was no-brainer for us to take the latter approach.
At the peak of the pandemic, we had what we called store hours. We didn’t open 24/7, even though we were an ecommerce store, just so we could make sure the customers who were purchasing could get their orders in a reasonable time frame.
We also focused on strategic projects that we would not have had time to invest in otherwise, such as dialing up personalization or working on our recurrent shipments platform, through which we help our members form healthy habits.
We completely revamped the recurring shipments program over the course of 2020 and started to scale it up in Q4. So far, customer feedback has been really positive, and we are seeing improved engagement, order rates, and lifetime value as a result of investing in the program.
Speaking of personalization, what aspects of those efforts are you particularly proud of, in terms of driving revenues, customer satisfaction, or some combination thereof?
Our guided shopping quiz is where all the personalization starts. If you go to a traditional grocery store with a goal in mind—starting a diet, gaining weight, or losing weight, or whatever it is—you’re on your own. You’d have to browse all the aisles and read all the nutrition labels to figure out the perfect products for achieving your goal.
We position our program as a helpful personal shopper that finds out what really matters to you and your household. That includes the categories of products you’re interested in—the kinds of dietary and ingredient restrictions you want to subscribe to, what your goals are, and so on. Then, the program recommends the perfect set of products for you.
Here’s the way that manifests in the user experience: Users take the questionnaire, and then we walk them through the store one aisle at a time, focusing on the areas they’ve expressed interest in, and providing personalized recommendations for products in each of those categories.
We take everyone through that experience, and afterward, they can check out right then and there, or they can continue to browse and shop in a more traditional ecommerce manner.
One really interesting thing about the model is that it provides a better experience, but it also takes a certain gumption to make a user take a quiz just to shop.
As soon as we launched this program, there was a pretty dramatic improvement in first-order conversion rate—about 16%. Probably even more exciting was our ability to cut down the time to checkout.
Our average order has about 15 items in it. Through traditional ecommerce search and browse, customers would take quite a while to build that basket of products. The guided shopping experience allowed us to cut down that time by more than 25%. Not only are we converting more members and helping a lot of them get value out of the product, but we’re also saving them time by bringing the products to them.
The other major benefit of this program is it solves our cold-start problem. When you want to start personalizing, you typically look at someone’s browsing history, purchase history, and a bunch of other data to figure out what the best products are for them.
For new visitors, you usually have a very thin data set to personalize with. What the program has been able to achieve is allowing customers to give us a tremendous amount of relevant data, which in turn allows us to create a bespoke experience for them from day one.
So, there’s the answers to the quiz, but there’s also every single page on the site being sorted, and rank-ordered, based on the affinity between the user and the product according to their answers to the quiz. All those affinity models are synced to our marketing and CRM [customer relationship management] systems. That way, every email and every marketing communication the user gets from day one can leverage that customer data.
How do you determine the right amount of information to collect? Or is the calculation that Thrive Market attracts the types of customers who want to interact with guided selling?
It’s all of the above. We definitely need to be able to strike a balance, because it’s a privilege to get customer information. There needs to be a very clear quid pro quo that we can message to members, that says we will provide you immediate value if you provide us data.
That’s true at a high level. It’s also true for each specific data point. The questions we ask have to be relevant, and they have to make sense in terms of how we would use the answers to provide better recommendations. There is a limit to how many of these questions you can ask before you start to see fatigue and drop-off.
We did extensive A/B testing and multivariate testing before we launched the version in production today. The interesting thing we saw is that [this type of data collection] works in your favor if you approach it authentically, provide a real value on the back end, and communicate some of that value in your customer messaging. We certainly see drop-off in the questionnaire, but it’s low. I think more than 85% of people complete it.
It turned out to be a win-win for us. We drove up membership and sign-up rates. We drove up first-order rates for each person who signed up. But it definitely required massaging to get the balance right.
How do you personalize promotions?
We want the promotions to be relevant. My biggest pet peeve from the early days was that consumers who clicked on the Vegan filter on our site would still get the beef stick promotion going out the door. It’s something to laugh at, but you can imagine it erodes their trust that the brand understands where they’re coming from.
Now that we have this 360-degree view of our customers, and synced that view to all our marketing systems, every single promotion is personalized. We send a promotion to only the subset of the audience that it would be relevant to. So, that’s step one: Promote in a way that drives high relevancy, and send people products they’re likely to be interested in.
It’s good for the business and the consumer. We see higher engagement rates around those promotions, even though they go out to a much smaller subset. Sometimes we get potentially 80% of the revenues on the promo off 20% of the audience, because we’re sending it to the right group.
We score every customer in our file every single day based on different scoring methodologies. Ultimately, we want to focus on people who have the potential to have a successful, long-term relationship with the brand. And we’ve got the data and structured insights to identify them pretty accurately.
When we see high-value customers start to lapse out—maybe they had a negative experience on an order, or something happened in their lives, and they forgot about Thrive Market for a bit—we know based on lifetime value which specific customers to focus on reactivating.
All of that leads to less promotional dollars spent on our end, which is great for the business. Not to mention, we have better brand equity overall, since we don’t want to be a promotional player.