Stitch Fix, now a public company, proved that its customers love its data-powered styling service when it disclosed it was on track to earn nearly $1 billion in revenue in 2017. With new investors to please, the company has to show it has the ability to get more customers on board.

On Friday, 8 million Stitch Fix shares started trading at $15 each, a lower-than-anticipated opening that raised the company $120 million. The IPO signals wariness around the limits of e-commerce companies, following the plateaus of companies like Blue Apron. According to Eric Kim, managing partner at VC firm Goodwater Capital, however, the modestly priced IPO fits in line with how Stitch Fix has paced its growth so far. The company — which combines the intelligence of data algorithms with the curation of personal stylists to send subscribers monthly shipments of clothing that are then either purchased or sent back — has mainly built up its 2.2 million-person active customer base through word of mouth.

“Data is really Stitch Fix’s key differentiator,” said Kim. “Traditional retail has no clue as to what goes on in the minds of their customers or their potential customers, other than when they make a sale or get a return. Stitch Fix knows who’s buying what, who’s trying and not buying what, what they like and what they don’t like.”

Now that Stitch Fix has gone public, the company will have to continue growing both revenue and members each year, meaning that it will have to convince more and more people that its algorithms and stylists are better shoppers for customers than customers are for themselves. Eric Colson, chief algorithms officer at Stitch Fix, said the company has barely scratched the surface when it comes to data. As the company celebrated its IPO at Nasdaq in Times Square on Friday, Colson laid out the company’s three areas for potential growth, and how the company’s vast data profiles can promote each.

Entering new markets
Stitch Fix has already expanded its service to include plus-size fashion and men’s clothing, but Colson said there’s room for growth in places the company hasn’t yet extended its reach (say, kids’ clothes or international customers).

When entering a new market, Stitch Fix has to build up new sets of data profiles, and it can’t look to its existing database to identify existing trends and patterns. But, Colson said, it still doesn’t need to reinvent the wheel. The algorithms are the same, regardless of demographic, and to start collecting and applying customer data, Stitch Fix just needs to open the floodgates.

“When we launch in new categories, we don’t need to build new algorithms. They’re there and ready for the data to start streaming through, and then they start training themselves,” said Colson.

Bringing on new clients that are similar to the existing customer
Stitch Fix’s core demographic is still women, and there’s plenty of room to grow in that space. Colson said the first priority is to increase personalized marketing efforts across channels including Facebook, Google, Pinterest, podcasts and TV, with the goal that no one channel accounts for more than 50 percent of marketing spend.

Stitch Fix needs to prove to its new customers that its service is worth it, but that becomes tricky without the data history that builds up after a customer spends time with the service. To make the styles as personalized as possible for a new customer, Stitch Fix takes what it learns from the initial customer survey, then draws on its existing customer database to match up lifestyle and body type to make suggestions. The company also uses information around how the customer found them — like through a search for jeans, for instance — to guide its first communication with that customer.

“In retail, personalization is the Holy Grail,” said Kim. “This in an advantageous position for Stitch Fix to be in, and it will continue to be their main differentiator. The sooner they can suck in a customer by nailing their style, the easier they’ll be able to succeed.”

Improving the relationship with the existing customer
Colson said that, for its first few years, Stitch Fix’s marketing budget was untapped because the company was too busy aligning its supply chain, stylist network, data profiles and warehouses with demand to make a concentrated effort in bringing on new clients. As the company prepares for its first marketing blasts, it still plans to improve the experience for existing clients by adding new ways to collect information and bringing in more interesting inventory.

When customers sign up for Stitch Fix, the company collects 85 initial data points to learn more about a client’s body type, style preferences and lifestyle. As the customer tries out their monthly shipments, or “Fixes,” there’s an 85 percent review rate on each individual item. But Colson said his priority is creating more ways that customers can share more about themselves. The company has integrated Pinterest boards into customer profiles so they can help customers communicate something that’s hard to put into words, like personal style. Recently, the company also launched a Facebook Messenger game that serves up styles to users and lets them thumbs-up or -down what they do and don’t like in rapid succession.

“We’re always looking for new ways to get our hands on that crucial information,” said Colson. “An easy place is in the onboarding, but what about beyond that? On Pinterest and through gamified experiences, we can collect information while the client is having fun. We want to keep creating things that let us collect data like that.”

Finally, Stitch Fix improves the overall customer experience by using data to improve its inventory. Stitch Fix shares reviews with its partner brands so they can act upon negative reviews around fit or style, works directly with designers like Jason Wu to design exclusive collections based on customer information, and designs an in-house collection called “Hybrid Designs” made up of new pieces based on white spaces found by the data algorithms.

“Working with our own data, and with designers and brands, means that we can identify gaps in inventory, and there are a lot of gaps,” said Colson.