Brightloom raises $15 million to help ecommerce brands predict what customers want to buy next

Brightloom, a company that’s setting out to help online businesses use transaction data to predict what a customer may want to buy next and how much they may spend, has raised $15 million and officially launched its platform out of stealth.

The launch comes as companies across the spectrum invest in digital and look to capitalize on the massive push to online sales ushered in by the pandemic. Back in December, Bloomberg acquired Second Measure, a startup that uses transaction data to glean insights into consumer behavior and company performance, though the target market is slightly different.

Formerly a restaurant technology provider called Eatsa, the company rebranded as Brightloom in 2019 and announced a partnership with Starbucks to create an “integrated digital platform for restaurant brands,” alongside a $30 million round of funding. The company has since pivoted to become what is known as a “customer growth platform” (CGP) aimed at restaurants, retailers, and consumer brands, and now licenses its old Eatsa technology to a company called Apex which continues to market and develop it.

Data

The Brightloom platform ingests transactional data from its customers, alongside catalog or menu data, and marketing campaign data to predict what a customer will buy next, and then creates personalized product recommendations and promotions for each customer.

“In order to keep things simple for brands, we limit the range of data we ingest,” Brightloom CEO Adam Brotman told VentureBeat. “The fewer the data sources and the narrower the range of data we need, the easier it is for brands.”

Companies can channel this data into the Brightloom CGP in one of two ways: regular updates via a SFTP server, or through setting up a direct connection between its data warehouse and the Brightloom platform. Then, brands can deliver personalized marketing campaigns to encourage them to spend more.

“Our models use the data we get from brands to identify patterns in customer behavior relating to the products they buy, how often they visit, and the amount they spend over a set period of time,” Brotman said. “These identified patterns in customer behavior are used to predict for each customer the product(s) they are likely to purchase next, and the amount they’re likely to spend in the next few days, weeks, or months. These predictions then drive personalized product recommendations and promotions that incentivize customers to act in predictable and repeatable ways.”

Above: Brightloom: Personalized offer

Under the hood, Brightloom uses machine learning to analyze each customer’s transaction history, and compares it to similar customers to predict future behavior and maximize spend. This is kind of similar to what Spotify does with music recommendations, leveraging vast data sets to build a genome and suggest new music to subscribers based on the tastes of other like-minded users. Indeed, as Brightloom proudly proclaims on its website, “they’ll think you’re reading their minds, you’ll know it’s the CGP.”

Using what it calls SmartSegments, Brightloom enables brands to test and experiment with various offers, seeing which ones are having the most impact and tweak the kinds of discounts or offers they serve up at the checkout, for example.

Above: Brightloom: Statistical experimentation

Moreover, Brightloom offers various reporting and analytical tools to see what impact their campaigns are having on their bottom line.

“When a brand connects their transaction and product data source with the Brightloom CGP, a few things happen immediately,” Brotman said. “Each time the brand refreshes that data, Brightloom ingests the update and transforms it into a standardized dataset. Once the data has been canonicalized, it is fed through proprietary models that, in addition to generating personalized product recommendations and promotions, produce two different analytical data products.”

These products include “actionable intelligence reports,” which provide visualizations and insights into customer behavior and the impact on business results. This also includes recommended strategies to “move groups of customers into a new activity tier,” as Brotman puts it.

Then there are marketing campaign results, which may show correlations between campaign tests and revenue growth, transaction frequency, and average order value between different control groups.

Above: Brightloom: Results

Prior to now, Brightloom had raised $30.5 million, and with its latest investment, from backers including  Valor Siren Ventures and Tao Capital Partners, the company said that it plans to accelerate its R&D and grow what is effectively a “data science-as-a-service” platform. The company has already secured clients such as Evergreens, Jamba, and Kickee.

This is a key point worth picking up on. The kind of technology that Brightloom is pitching effectively replaces data scientists and analysts, which may allow smaller companies to compete on a more level playing round to the giants of the ecommerce world.

“We see growing demand from brands that can’t afford and / or don’t have the internal resources to manage a complex, long data project,” Brotman said. “Our customers value that the CGP is simple, fast, and effective — they can get up and running with personalized marketing in weeks, not months.”

VentureBeat

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform
  • networking features, and more

Source: Read Full Article