How Freshly Uses Customer Data to Influence Meal Development — in Real Time
At Freshly, we’re customer obsessed.
Our consumers are our lifeline — the heart of our growing business — and shape both our company culture and product development process. And, as the leader of our meal and menu development team, I am constantly monitoring their feedback and exploring new ways to offer both healthy and delicious meals that our customers love and crave.
Freshly is a health and wellness company that offers fully prepared meals for delivery. Health is at our core; we have unwavering nutritional standards, but our recipes also have to taste great. Because food is a subjective and emotional experience, it’s inherently tricky to gauge how customers will respond prior to getting that food into their hands. So, as a direct-to-consumer business that gets over 15,000 unique ratings on our meals each week, we have the unique ability to leverage live data to optimize every meal available, even when those meals are already live on our menu.
Data as Direction
To start, our meal development team was built to include people with strategic, brand and product development experience to manage the complexities of our menu and product portfolio, as well as two teams of chefs — one set is in our test kitchen, developing recipes, and the other is on the production floor at our facilities, helping make the food and monitoring quality and culinary techniques. So when Freshly customers use our ratings tool to give meals between one and five stars, our team is able to mine and dissect that data to understand what’s resonating and what’s not.
By aligning with our customer feedback in this way, our product development doesn’t stop when a meal launches. It continues into the first several weeks during which we get our meals into people’s homes, collect enough data to find patterns (that are statistically significant of course), and perfect meals based on customer feedback — in real time.
Striking a Balance
Freshly has one menu that spans the entire United States, which involves the mammoth undertaking of juggling millions of dietary and taste preferences. (There’s a reason that companies like Whole Foods and Costco buy products regionally — people’s choices can deviate drastically based on where they live and what they’re accustomed to eating.)
Food companies, in an effort to build a so-called menu for the masses, often shy away from making food that has too strong of flavors and/or ingredients that can be considered polarizing. There’s nothing wrong with that, but Freshly is not in the business of selling bland TV dinners, so we try to push ourselves to be more exploratory — in terms of cuisines, ingredients and flavors, and being comfortable testing concepts in the market.
Turns out, though, we didn’t push the envelope far enough on our Chicken Tikka Masala, and it was called out for being mute and inauthentic. In response, our chefs bumped up the spices, and, within weeks, the ratings skyrocketed. (To those who have been asking for the Chicken Tikka Masala to make a reappearance on our menu, we heard you and will have exciting news soon. Stay tuned!)
The ability and willingness to pivot both conceptually and operationally has helped Freshly’s business grow. We’re constantly making adjustments to make our food taste better because we believe there’s always an opportunity to refine our product from an executional standpoint.
Reacting Quickly
It’s common for food companies to conduct extensive product testing pre-launch. There’s nothing wrong with this approach — we use some traditional testing methodologies, too — but it has the potential to slow down the product development process and compromise speed to market. And that, in turn, can make it harder to be a leader in the space. If you want to be a disrupter and catch trends when they are ripe, you sometimes have to make bets.
The beauty of how we built Freshly’s product development model is that we can do just that, knowing that we will be able to measure success almost immediately and react with speed — adjusting products on the fly or even pulling them if the response is not up to our standards.
Additionally, conventional pre-launch product tests, like distributing concept surveys, may not result in accurate data. To take the Steve Jobs approach, your customers may not know what they want. We are the food and wellness experts and we are responsible for helping them figure that out. People may not think they want a mac and cheese with butternut squash mixed in it, but then they try it, love it, and are thrilled that they’re eating a “better for you” version of a nostalgic classic.
All this to say that data is important, but so is the pace at which you respond to that data. Freshly makes sure to take into account what our customers actually want to eat, not just what they think they want. This is why we built our team to support a nimble product development cycle. We are not afraid to try new things and innovate with our menu, with a keen eye on what people want as part of their diet. Ultimately, we’re quick to not only grasp that customer data, but also react to it quickly, so we can customize the menu to our customers’ needs and improve user retention and satisfaction.
The Full Picture
Today, it’s common for companies to rely on customer reviews. But I remain leery of relying exclusively on ratings. That’s just a piece of the pie.
In favor of data diversification, Freshly combines reviews with information gleaned from sales, site behavior, surveys, focus groups, user experiences, customer profiles and more. We’re also working with our data team to explore artificial intelligence as a way to better understand customer profiles and how that correlates to behavioral patterns and taste preferences.
It’s not just about the product itself; it’s about our customers’ experience with the product. Are they exploratory eaters or do they want to keep dishes simple? Do customers gravitate towards different types of meals? How do they feel when eating Freshly meals? Are we helping them achieve their dietary and nutritional goals?
Ultimately, sales reflects a demand for concept; repeat sales and ratings indicate whether we are delivering on that demand; and qualitative research reveals customers’ holistic experience with our products. By contrast, ratings can be from a vocal minority so if you lean on them too much, you can make product decisions at the expense of your other customers.
None of us would be in business without our customers so let’s be conscientious in exploring the full picture painted by their preferences, histories and behaviors.
