How early-stage startups can get value from data, fast

On March 29th, Whaly hosted a webinar with one of our customers, CrushON, a circular fashion marketplace.

It was an inspiring and insightful discussion with Leila Veerasamy, CrushON's Chief of Staff, on how she managed to set up a data foundation with limited resources and no data background. Read her interview below to learn how she achieved this, and how you can do the same!

The key? Get yourself a self-service BI platform like Whaly that doesn't require technical expertise to set up.

Q: Can you give us a little introduction to yourself, your role, and to CrushON?

LV: I’m Leila, and I’m the Chief of Staff at CrushON. For those who are familiar with the Chief of Staff role, you’ll know that this is a very vague and wide-ranging role. For me, it consists of assisting the CEO with his operational and strategic projects. The specific tasks for which I was onboarded onto CrushON was really to take on various projects such as investor relations, developing organizational processes for the company, hiring, and one of my latest and important projects has been organizing the data strategy for our brand.

I have a background in development, having studied economics and international development. I began my career in the social sector, and then pivoted into sustainable fashion, first in India, and now with CrushON in Paris.

Q: Can you elaborate a bit more on how your business works as a B2B Marketplace for circular fashion?

LV: CrushON has indeed pivoted recently. We were mainly a B2B2C marketplace up until late 2022, and now we’re mostly a B2B marketplace for circular fashion.

Just to give you some background on the history of the company to contextualize our model — CrushON was really created with the mission to eliminate textile waste. We noticed that the traditional, very linear “take-make-waste” production model of the fashion industry was not a sustainable one. When we looked into developing solutions that would make fashion more circular, we found that the highest-impact solution really lied in the resale sector — essentially, vintage and second-hand fashion. So CrushON set out on this mission to help professionalize the second-hand vintage sector by connecting professional and semi-professional vendors from this industry to different distribution channels. Distribution can range from retailers, online distribution, marketplaces, and more.

What we do now is help connect these vendors (SMBs) to customers via our platform, by plugging in their shops into all different types of online and offline marketplaces. So for the vendor, we’re more of a B2B marketplace where vendors can come onto our platform, shop around and identify where they want to plug in their catalog, and do so in our subscription model. From a consumer standpoint, we’re more like a traditional marketplace or multi-designer store.

Q: Could you go over your data journey prior implementing Whaly as your BI platform, and how you were able to extract insights to influence your business decisions?

LV: We’re a fast-growing, early-stage company and our model has evolved and grown quite a lot. This is a blessing in terms of business development, but it’s a bit of a curse in terms of data! Early-stage companies tend to operate with really lean teams, and data isn’t usually your highest immediate priority. You’re really focusing on developing your product and learning as you go.

In our previous B2B2C model, vintage and second-hand sellers could work with us and distribute their items through our 3 different channels: 1) online website (marketplace), 2) retail shops, and 3) events wing (vintage markets, where our end users could have the same experience but across different cities across Europe, usually in a limited-time opportunity like over a specific weekend). Our focus during this time was really on growth and sales. We have a community of over 2,000 vendors directly selling to the consumer in these 3 channels. It was more logical at that time to focus not only on GMV or Sales Rate as our North Star KPI, but also to do so in terms of these 3 business verticals.

All of our data was managed and visualized on spreadsheets (Google Sheet and Excel), and it was very Ops-centric. We had separate teams: events team, retail team, and online team, and each team would look into the performance and evolution of their own business verticals. They would track progress and performance day by week or week by week on their own in their own spreadsheets. Then, periodically, their team’s data would be congregated into one dashboard on Google Sheet that the whole organization could check out once in a while, usually on a quarterly basis.

It worked out for that time when we were in MVP stage and figuring things out, but it meant that these teams worked in siloes and there were no synergies or collaboration with each other. This wasn’t a sustainable way to work with data since it didn’t allow us to have a bird’s eye view on data — and we weren’t able to access and learn from data in real-time.

We also used individual tools like Google Analytics for website traffic data, Mailchimp for user acquisition, and CRM tools like HubSpot to help track evolution and churn of the vendors, but this data was scattered across different source systems.

Q: Where and when did the need to put a BI platform in place come from? What was the moment that was like, OK gotta do something about this?

LV: I don’t think there was a single defining moment — it was more about the growing pains and slowly realizing that it was becoming more of a headache, and more urgent. For us, data wasn’t really centralized. We had a way of doing it, but it was very time-consuming and manual, and to be honest, it wasn’t very attractive.

It was also problematic that our team leads weren’t interacting with other team’s data very much. It was really just the C-suite who would have a bird’s-eye view on data and would regularly consult each team’s data. In terms of individual teams, they would just look into comparing certain events from the past, within their own team’s remit. If the events team ran an event, for example, they would look back on the previous one and see which event yielded better results.

For us, we’re still a 4-year old company and are definitely considered early-stage, especially with our management style which means we didn’t have to be VC-led. Over the past 4 years, however, we’ve definitely accumulated a lot of historical data. By the time we switched to Whaly, we already had around 500,000 visitors a year and a vendor community of over 2,000 vendors.

We’re growing very fast, and as we scale up, that whole idea of having a bird’s-eye view on data and performance collectively in real-time was increasingly important. In order to benefit from that, each team’s ‘builder’ had to figure out how to spend less time building out the data, but more time consuming it and analyzing it, to extract valuable insights from it.

It had become so time-consuming for us to update all of our source documents, dashboards, etc., that we didn’t end up using the dashboards as much as we wanted to. So that’s when it became essential to move to Whaly’s BI platform.

Q: Once you decided on using Whaly as CrushON’s BI platform, how long did it take you to set up your first dashboards and start getting value from data? What did you do to get there?

LV: It happened really quickly, to be honest. We had a great onboarding process with Whaly where we were able to test out the product for one month. Within that POC period, we were able to get our first real dashboard up and running, with our very own data. We were able to connect all our different channels and sources (retail, events, online) all in one place.

While this was possible with our other tools (with a massive amount of effort and time), it was too time-consuming and wasn’t automated. And because we were using Google Sheets, there were issues with properly connecting sources and adjusting data.

While the total POC lasted about a month, we already had our first dashboard with GMV & Sales Metrics up and running within 2 weeks, connecting sources from our three different channels. It did take the full month to have a polished version with more granular data beyond the GMV - like number of unique users across all different channels, or the customer acquisition cost.

What was interesting about the process is that it was still really easy considering I don’t have any data science background - I was hired more as a “generalist” in my Chief of Staff role, not as a data specialist - so I enjoyed using a no-code platform that empowered me to simply plug-and-play and get what I need. The value that we got out of Whaly was pretty immediate. Building a dashboard and seeing how different channels interact with one another made us understand a lot of different things about our business (seasonality, trends, etc.) and helped us confirm our intuition that we should shift to a B2B model.

Q: What are the main metrics that you’re tracking in your dashboards? How did your data needs evolve with your recent shift to a B2B model?

LV: We used to look more into Sales (B2C metrics). While these Sales and GMV metrics are still important - we’re also now focusing more on traditional SaaS metrics. Our vendor activation rate, churn rate, the onboarding evolution, our monthly recurring revenue - things like that.

It was easy for us to shift and look into different sets of metrics, because how Whaly works is that as long as you have your sources connected, and you have a way of cleaning your data, you can then really choose which visualization you want to have in your dashboards, and focus on certain metrics over others. You can always run different queries and get different questions answered about any of your data. In this sense, it was easy for us to go from looking at B2C metrics to more B2B metrics with Whaly.

Q: What has data and having Whaly as your BI platform allowed you to do that you weren’t able to do before? How has it been game-changing?

LV: Whaly made data much more accessible to a wider range of team members in the company. It also really helped us solidify and clean up our processes. Like I said, a big part of my role is around organizational management, so this helped us figure out how to prioritize and focus on the most important parts of our business.

We care and value our customer retention (end customers), but we’re a lot more focused now on seller retention. It’s helped us understand what to do on a day-to-day to improve seller retention - for example, whether developing new offers for our vintage seller community is attractive to them, and how these new offers are helping them perform better. Or, do we need to get a new retail partner or distributor for them?

This has allowed us to focus on seller pain points and community, rather than consumer, fully supporting our business model shift.

With Whaly, we now have a centralized view of our data, and can see the impact of one team on another. So, how the seller acquisition team is helping the development of new partnerships, and vice versa. There’s definitely more synergy happening between teams.

Q: Shifting briefly to the retail angle - why is it particularly important for retail and marketplace companies to track data on an ongoing basis? What makes retail unique in its need for data?

LV: That’s a great question, and this came up immediately in my onboarding period with Whaly. When I was working with the Whaly data expert during our POC, I mentioned that while we have a solid way of collecting and understanding our data when it comes to online activities, we also need to combine retail (or offline) data to see how they compare and model them against each other. This is more challenging because offline retail data is much more granular and is harder to collect and analyze.

When it comes to Marketplaces and e-Commerce businesses, it’s no surprise that data is super important. It’s the lifeline and bread and butter for these types of companies, and what makes one stand from another and defend itself. You could say, “look at our acquisition rates, sales, return rates, etc. - this is why we’re better.”

For offline and more traditional retail (brick & mortar or department stores), a lot of the data is focused on customers. For our model specifically, our data helps us figure out whether a partnership is successful or not, and what are the factors that play into it. Let’s say we have a new partnership deal with a high-street retailer. Out of our 2,000 vendors, we have 10 of them that pay that added fee to be distributed in that retailer. The sales in that shop can go well (or not) according to multiple factors: alignment of vendors and shop, seasonality, location, brand itself, and much more. The more data we have, the better, so we can understand which factors are going well and which aren’t. To take it a step further, the more data we have, the more we can understand WHY things are going well. Is it because the vendor and shop are a great match and fully aligned? Is it a really good retail season, or is the shop in a great location?

So basically, it gives us a quick grasp and an ability to compare multiple locations at different times in their history. We’re able to do interesting cohort analysis to see shops performance and we can make predictions based on trends we’ve seen in the past.

This is all really easy to set up in Whaly, and all you really need to do to consume and analyze this data is to use time filters - which is something anyone can do, even without a data background. Our supply team, or acquisition team, partnerships team - anyone can use it and learn from data. This enables any of our team leads to be incredibly reactive and provide feedback and business intelligence to our vendors, which is part of what we offer.

Just a point on the resources and our ability to give everyone the ability to consume data - we’re growing, but we’re not trying to hire left and right as quickly as possible. Hiring a data team can be super costly, especially pre-financing. So the ability to have this ease of use, without having to dish out all your money into a building a dedicated team allows you to have more flexibility. That’s super key in this economy. You want to stretch your resources out as much as possible.

Q: Are you able to combine online and offline data with Whaly?

LV: Yes, definitely. That was the first exercise that we did - connect our online Marketplace data with the Sales reports that we get from our offline retailers and distribution channels.

We do it mostly through the Modeling feature within Whaly, to get an overview of our business. I think it’s really beneficial because one of our priorities is to really automate the process and reduce the manual tasks when it comes to combining our data. This is a work in progress, so we’re not quite fully automated yet. Once models are created, we do still need to partially update the figures manually, but we spend a lot less time on it.

However, we know that our organization is becoming more structured and and clean about collecting data in our databases, as we’re re-organizing it now, we can ensure that we’re increasingly automated bit by bit over time.We’re able to adapt our models and dashboards as we’re adapting our company as well!

Q: Where would you say you currently are on your data journey, and what’s next for you to tackle?

LV: Continuing on the automation path - the dream is to be fully automated! I’d love to be able to put down my data science hat once and for all once I’ve really implemented an automated data foundation and a data-driven culture at CrushON.

It would be great to start having fun with it, and to start providing dashboards to our retailers. Our POC has been quite successful, so we’ve been able to develop or open new shops within the same distributors or retailers quite easily.

Right now, we’re focused on “zooming out” and getting that bird’s eye view of the company. What would be interesting, later on, is that once we connect our sources, automate them all, and we realize this is a self-run process - is to create dashboards that focus on a more “zoomed in” view of what’s going on in each of our partnerships at a super detailed level.

Q: What tips / advice do you have for the audience who are looking to get value out of their data quickly as an early-stage startup?

LV: This will be a brief takeaway, that I hope is memorable. Just do it earlier rather than later! Whenever you think you should do it, just do it as soon as you can. Ideally, even 6 months earlier. Having your data easily readible and accessible, and analyzed, makes it really easy to communicate it as well. Not only did our data help us confirm our convictions and know-how, but we’ve also been able to prove it externally and be more credible, which has been a game-changer.

To learn more about how Whaly can help your startup or small business put a proper Business Intelligence process in place for the first time, get in touch here!