Ibotta is a free mobile app that gives users cash back on brand purchases at a range of stores (e.g. grocery, convenience, pharmacy, liquor, etc.). The company has 40 million users in the United States and partners with 2,700 brands. An Ibotta use case would be, a user goes to a store to buy beer. While in-store, she opens Ibotta and sees that, if she gets Modelo, she will receive $1.00 cash back. She usually drinks Heineken but will try Modelo given the discount. Ibotta is uniquely positioned to influence in-store user buying intent. The company generates revenue through brand partnership commissions and in-app advertisements (i.e. watch a 10-second advertisement to unlock offers).
Data Generating & Renewing Processes
The typical Ibotta user experience involves the users going to a store, opening the app, identifying the store, viewing offers, buying those brands, and submitting a picture of their receipt for cash back. Ibotta converts each receipt picture into a row of data per item tied to the user, store, location, price, and brand. Ibotta ingests these data using an AI algorithm and incentivizes its user to submit as many receipts as possible to grow their database.
Big Data Value Creation
Ibotta’s uses big data to offer: 1. high confidence offer ROI analytics, 2. hyper-targeted in-app advertising, 3. market share and trend insights, and 4. shelf placement verification.
High Confidence Offer ROI Analytics: Ibotta leverages receipt data to develop user profiles including store visit frequency, average spend per visit, and brand loyalty. These profiles allow Ibotta to estimate offer ROI with high confidence. One method they use is next store visit loyalty. This metric measures whether a user buys a brand’s product the store visit after they receive an offer. When combined with store visit frequency, Ibotta estimates offer ROI over time, assuming continued brand loyalty. This analysis increases brand partner’s confidence in their investment.
Hyper-Targeted In-App Advertising: Secondly, the user profiles enable targeted advertising to users most likely to convert. Two applications for this are new product launches and market share capture from competitors. An example of new product launches would be if Modelo were launching a margarita product to complement its beers. Modelo can target Ibotta users who have a favorable view of Modelo based on past Modelo purchases or are redeeming Modelo offers. The second type of campaign would be targeting users who consume competitor products to gain market share (i.e. targeting Corona buyers).
Market Share and Trend Insights: Ibotta’s data gives it early access to industry trends which it can share with brand partners to inform new products and targeting. For instance, the data may indicate a brand is gaining market share (i.e. Corona gaining) or that a new product is becoming popular (i.e. bottled margaritas). Traditionally firms gather these data through focus groups and surveys which lag trends. With Ibotta, brands get these data months earlier and more cheaply than traditional methods.
Shelf Placement Verification: Lastly, brands face a problem verifying if or where their product is placed in-store. Ibotta can use data to identify stocking issues. By combining store and offers added but not redeemed data, Ibotta can identify the likelihood a product is out of stock or poorly placed in-store. The Ibotta app can prompt users to verify the product is out of stock in-store or take a picture of its placement in-store for $0.25 cash back. This functionality verifies placement much more cheaply than brands-to-store outreach.
Investments & Processes / Challenges Overcome to Get Here
The three major challenges Ibotta overcame to curate this massive database were: 1. generating receipt data without brand partnerships, 2. incentivizing receipt upload frequency, and 3. processing the receipt data. For the first two challenges, Ibotta employed a loss-leader strategy. They offered $0.25 per receipt uploaded even if no offer was redeemed. Neither of these programs generated revenue; they generated data. The third challenge was turning receipt photos into uniform data. Ibotta developed and trained a machine learning AI to scan receipts. This AI has improved over time but initially required humans to code massive training data sets from user submissions.
Challenges & Opportunities Looking Ahead
The biggest challenge for Ibotta will be the move towards online shopping and away from in-person shopping. Ibotta will need to invest in embedded browser tracking and cashback online to ensure it can capture item level data from online purchases. This is also the company’s biggest opportunity.