WW is in the business of wellness. The company, previously known as Weight Watchers, is perhaps best known for its affiliation with celebrities like Oprah Winfrey, Kate Hudson and DJ Khaled.
What the company is less known for is their tech stack, but it’s hard to imagine that will be the case for long — the fact is, WW has developed one of the most competitive data and software operations in town.
We sat down with the company’s busy data science team to learn more about the people behind the WW product, which has its own social media platform, Connect. Here’s what we learned.
EMPLOYEES: 18,000; around 600 locally
WHAT THEY DO: Grounded in science, enabled by technology, and powered by community, WW partners with individuals on their wellness and weight loss journey.
WHERE THEY DO IT: New York City, with offices in San Francisco and Toronto
AN APP FOR THAT: The WW app allows members to track their food, weight and activity, as well as chat with a WW Coach, scan food labels and connect with other WW member via Connect, a social media community that connects WW members to share their stories and support one another.
Carl Anderson, Director of Data Science
As WW’s director of data science, it’s Carl’s job to grow the team and deliver impactful data products, with everything from advanced analytics to recommender systems. Carl’s team works closely with the rest of the business to understand what projects and products to tackle and to implement, roll out, test and refine.
BEYOND WORK: Carl is an avid cyclist.
You’ve had a really interesting career. Talk about your work experience leading into your current role and how that’s informed your work at WW.
I’ve found that my breadth of experience working across different domains — for example, healthcare, e-commerce, data compression — helps me bring a new perspective to my job at WW. I think my ability to jump into new domains, read, absorb and ask a lot of “dumb” questions is actually one of my superpowers.
I’m passionate about data strategy and helping organizations to extract value, impact and insights from their data. My research and experiments at other organizations have informed the principles and best practices we bring to all our data users at WW.
What was important to you as you helped groom the culture of your team at WW?
Data science is fairly ill-defined and broad — the truth is, you simply can’t know everything about data science. This often leaves young data scientists with a feeling of imposter syndrome. To make sure this didn’t happen within my team, I worked to foster a culture of openness, where no questions were dumb questions, and where we were inclusive in the ways we share knowledge and give people opportunities to learn and develop their skills.
Our department has a huge scope, giving employees the opportunity to learn any number of new technical skills and domains.”
When you look at your to-do list, what are a few regular things that you love and are excited to get to in your daily work?
In the last month, it’s been great to see the whole team rallying around our central code repository — making it better, more powerful and easier to develop and deploy our machine learning models. This is a Python framework we find useful for ourselves, but that we also think others will find useful. For this reason, we will soon be open-sourcing it under the name Primrose on GitHub. It’s been a fun activity for the whole team but also a great opportunity to give back to the data science community.
What does career growth look like in your department? What resources and opportunities are available to employees?
Our department has a huge scope, giving employees the opportunity to learn any number of new technical skills and domains, whether embeddings, natural language processing, time series modeling or beefing up their skills and experience on the software engineering side. As the team grows, there is also the opportunity to learn to manage and mentor, whether that means managing interns or other team members.
We have a budget for conferences and every member of the team also has access to O’Reilly’s online library and online courses through services like LifeLabs.
Michael Skarlinski, Data Science Manager
Michael manages the tactical and day-to-day aspects of the WW data science team. He plans project roadmaps, chooses technologies and refines the company’s approach to using data science to solve problems.
BEYOND WORK: Michael relaxes by playing the guitar.
What have you learned from your time as a manager and team leader? What surprised you?
I’ve learned that there is a lot of R&D involved in our data science projects, which needs to be separated from the more deterministic deployment and final leg of project development. This makes scoping projects difficult for data scientists. I’ve found that it’s often helpful to start with the most minimal approach so you can see where you hit snags later down the line. I’ve been surprised that even the simplest of approaches can run into major challenges when building data products.
Tell us about a project or challenge you’re working on that excites you most. How are you solving that challenge?
We have a lot of interesting projects that use both models and recommendation systems to best serve our WW members. We’ve identified similarities across projects and have been working on a unified framework to solve our common machine learning use cases.
We’re open-sourcing that framework — which we’re calling Primrose — as a team, and it’s really helped everyone to think about the way that our code can impact others.
The teams at WW are really willing to innovate and push the bar forward in the sphere of wellness technology.”
Did this job turn out to be what you expected it to be? How or how not?
I had some preconceived notions about WW as a non-technical legacy brand, which turned out to be totally unfounded. The WW tech and product organization has been fantastic, and I’ve had the freedom to complete projects without red tape or organizational push back.
The teams at WW are really willing to innovate and push the bar forward in the sphere of wellness technology.
Reka Daniel-Weiner, Senior Data Scientist
Reka is a senior data scientist on the growth team at WW. In her role, Reka aggregates data and builds models that predict what keeps WW members engaged with the program and helps them succeed.
BEYOND WORK: Reka enjoys making simple pantry items like bread and cheese from scratch.
How has your academic background — you have a Ph.D. — influenced the way that you tackle challenges or the processes that you use to do and verify your work?
In academia, you tend to take your time to make sure that absolutely every calculation is performed correctly without violating any statistical assumptions and explore all options under which the validity of conclusions might be impacted. This time away from academia has taught me both about rabbit holes that one might be tempted to disappear into but also about when it matters to look twice.
In your role, how do you work with other teams across the company? Is it easy to get feedback from coworkers who aren’t on your immediate team?
Everyone within WW is super excited to put our data to the best use. Whether it’s the marketing, product or finance teams, people have a strong need for data products, and we work very closely with them to define and fulfill their specific needs.
How often does your work intersect with the work being done by other colleagues in your department?
While most of our projects tend to be owned by a specific person, we closely collaborate and check in with one another. Since we use the same data and often encounter similar challenges, it is very helpful to have a team of people to exchange learnings with and bounce ideas off. Once a week we get together and review each others’ code — and even in the short time that I’ve been with WW, these sessions have taught me a lot.
Once a week we get together and review each others’ code — and even in the short time that I’ve been with WW, these sessions have taught me a lot.”
How do team members support one another to successfully hit both team and individual goals?
We have an extensive process in place that involves both the people team (human resources) and our direct managers. The process helps us define our own goals for the upcoming year, makes sure that they are aligned with the team’s goals and the company’s goals, and then gives us the opportunity to check back in and update them if need be throughout the year.