Instacart: Grocery Startup with No Groceries
Instacart offers a very good shopping experience and a predictable delivery window.
When the email came from Instacart, advertising that the site is now letting people order food from local alt grocery chain Trader Joe’s, I sat up and noticed. I had heard about the service a few months before, but I’ve never used any grocery delivery service. And if I was to, I thought, chances are I’d go with the plain vanilla Safeway.com, which delivers that mega chain’s food.
Instacart also delivers Safeway food, so it gives San Franciscans a good choice. But Safeway’s own delivery service was built, in part, on the back of the failure of Webvan: In addition to learning from Webvan’s hard lessons, Safeway acquired some of its assets, like trucks.
So what has Instacart, a third-generation Internet grocery delivery service, learned from its predecessors? I asked CEO Apoorva Mehta to walk me through the business.
Instacart, he tells me, is not a grocery business. Mehta calls Instacart a “marketplace,” but not for food. For labor. The business is built on hiring contractors that go to local grocery stores, pick items that customers order off the shelves, and then drive them, in their own cars, to customers’ houses. It is more TaskRabbit than Webvan.
Instacart’s workers are thoroughly vetted (like Lyft‘s drivers are, Mehta says), and have the use of a special worker-based smartphone app that walks them through stores, aisle by aisle, telling them what to pick up.
Although the core of the business is logistics and QA, Mehta says that it wouldn’t work without the human touch. He says that his smart shoppers know when to make a substitution if an item that the customer wants isn’t on the shelf, or when they need to call the customer for more info. “Our shoppers understand the intent of customers,” he says.
Instacart’s technology and information architecture includes, as I said, store maps for shoppers, and a proprietary system to get and categorize data about food items for display on the customer-facing site and mobile app, but Mehta says, “The hardest thing is the logistics of people, and managing quality of service on a collaborative consumption model.”
Instacart’s business model is still being written. In addition to per-item markup and a per-delivery fee, Instacart recently rolled out an Amazon Prime-like $99 annual fee option for all-you-can-eat deliveries. It’s all in beta, I gather. But Mehta is convinced the model works.
One thing that is certain: Webvan could never have done it this way. It was too early. Instacart, which I think will work, is built on the back of the smartphone industry. It’s only because entrepreneurs like Mehta have access to piecework or hourly contract laborers with their own information terminals that the model functions at all.
-Rafe
See also:
Peapod, Lyft, Uber, Exec, TaskRabbit, Zaarly, GigWalk
Read:
What Webvan Could Have Learned from Tesco
A message from Evernote
Build knowledge. Share ideas. Get things done. Evernote Business.
