Anyleaf—Putting an End to the Old Supermarket Circular
Having a computer-science background, an appetite for risk, and an eye for solvable problems put you in a decent position be an Internet entrepreneur in Silicon Valley. But there’s also a downside to these traits. You wind up seeing all the flaws and inefficiencies in everyday things and wondering how you could use Web tools to fix them.
That’s what happened to Jeff Hunter and his college friends Jason Marr and Josh Staiger, all members of the Case Western Reserve University class of 2004. For the last three years, the trio of software engineers from the university in Cleveland, OH, have been renting a house together in Sunnyvale, CA. Hunter and Marr worked at Apple, while Staiger worked at Google—up until last summer, anyway. That’s when they decided to quit their jobs to go on a quest to fix, of all things, the venerable supermarket circular.
“We all like to cook, and we would go to the grocery store fairly frequently,” recounts Hunter. “We’re also all frugal Midwestern boys. My mom grew up clipping coupons and she never buys anything that’s not on sale. She’s given me that frugality gene, and Jason and Josh are the same way.”
But these transplanted Midwesterners have a geometry problem: their house in Sunnyvale is exactly midway between a Safeway supermarket and a Lucky store. “It’s not faster to go to one or the other,” says Hunter. “So we pick the store based on the sales, and the sales are all in the weekly circulars. We’d have them sitting in our dining room table, and we’d flip through them to get ideas for what we should make that week.”
The more time Hunter and his pals spent doing that, though, the more frustrated they got. “Many of the deals are not relevant to you,” he says. “I don’t have a kid, so I don’t care about diapers. And sometimes the deals aren’t that great. You might not know that this $5-a-pound beef isn’t such a great deal, because it was on sale for $2 a pound two weeks ago. As technical guys, we were thinking, ‘This is a very interesting data set—there’s a lot of potentially useful information—but it’s not being presented in a way that makes it useful for us.’ We thought that if we could collect this data, organize it, curate it, and think about all the different products we could make with that, there would have to be some inherent value in that.”
That’s the business concept Hunter, Marr, and Staiger pitched to Y Combinator in early 2010, and it apparently clicked with the venture incubator’s penchant for helping entrepreneurs solve what founder Paul Graham calls “hair on fire” problems. The trio were accepted to the summer 2010 Y Combinator class, and they launched Anyleaf in January. The service lets you sign up for a personalized weekly e-mail newsletter listing the best deals advertised in circulars for stores near your geographic location. If the point of circulars is to get bargain hunters into stores and help them save money, the Anyleaf team reasons, then anything that makes deals easier to discover is good for both stores and customers.
Last week, Anyleaf released its first mobile app, for the Apple iPhone, so now deal hunters can search and browse for the deepest discounts and add their favorite ones to shopping lists on the go. Right now, while the Anyleaf guys study how people are using the app and working out the early kinks, the startup is offering discount data only for stores in the San Francisco Bay Area. But their plan is to go national eventually, and to make money by selling sponsorship opportunities in its newsletters and apps and partnering with stores to offer coupons.
So why do grocery stores use circulars, anyway? Lacking the patience to be a bargain hunter myself, I tend to throw them directly into the recycle bin. But Hunter says they plan an important role in the grocery business.
“Stores use one of two models,” he explains. “It’s the ‘everyday low price’ model versus the ‘high-low’ model.” Wal-Mart is an example of an everyday-low-price chain, offering the same prices all the time. Safeway, by contrast, is a high-low chain, offering different items at a discount every week as loss leaders while raising prices on the other items the store knows you’ll buy if you visit. For high-low stores, circulars are the main way of spreading the word about the discounts. “Safeway spends a lot of money every year sending off those weekly circulars, and it clearly drives people into the stores and improves sales, or they wouldn’t do it,” says Hunter.
All of that said, Hunter and his co-founders are convinced that “the traditional circular is on the way out.” They realized that if they built scraper software to visit the stores’ websites and gather all of the weekly discount data in one place, they’d be in position to apply search and filtering algorithms to help people compare the deals at nearby stores, or see just the deals for the grocery items they buy regularly. (Vegetarians, for example, can opt to screen out deals on meat items.) With all the data in one place, Anyleaf would also be able to give some context for each deal—showing how this week’s circular price for steak, say, compares to the average price, the historical low, and the historical high. “Only the most hard-core bargain hunter is willing to do that kind of exhaustive research,” Hunter says.
But the first version of Anyleaf was more like Google than Groupon. “We were thinking of it as Google Finance for groceries,” Hunter says. “You could see price fluctuations in chicken breasts over the last three months. It was kinda cool from a geeky, data-analysis point of view, but what we realized early on was that that just doesn’t match how people shop. The one thing that’s not broken about circulars is that they just come to you. So we decided to focus more heavily on e-mail and having the deals pushed to you every week.” Two-thirds of the way through their Y Combinator session last summer, Hunter and his co-founders scrapped their first prototype and started building a new one focused on reassembling the scraped deals data into location-specific newsletters. “Think of it as your personalized weekly circular,” Hunter says.
The Anyleaf iPhone app introduces a new spin. “If you didn’t have time to plan before you went to the store, you can just use that to see the top deals” once you get there, Hunter says. “Right now we’re thinking of it as an ancillary thing. But for all we know, people will focus on the app, and maybe that will become our complete product.”
Even if the app and the e-mail newsletters find a big following in the Bay Area, however, growing beyond NoCal will be a challenge for the startup. That’s partly because the grocery business is so fragmented. The top 10 grocery chains in the U.S. account for only 30 percent of sales. To offer fairly complete coverage, Anyleaf would need to scrape weekly deals data for more than 200 chains—which would not only be a computing challenge but bring its own problems in areas like quality assurance and “data cleanliness.”
“All the stores post [their deals] in different formats,” says Hunter. “We have to parse the product name, the quantity, and the price, and then we have to logically identify different products—we have to say this Lucky chicken breast is the same as that Safeway chicken breast, and they should show up on the same page when you’re comparing.”
Hunter says the startup is considering seeking freelance help screening all this data from deals bloggers and mommy bloggers. “But we don’t want to spend a bunch of time and money expanding until we know for sure that we have the right product and the right business model in place,” Hunter says.
So far, Anyleaf is largely a bootstrapped operation, aside from Y Combinator’s investment last summer (the incubator usually puts $15,000 to $20,000 into its startups). “If we hit on a sustainable business model and we want to expand nationally, we will definitely have to do some fundraising then,” says Hunter. “For right now, we are not in desperate need of funding.” Which isn’t too surprising for a group of frugal Midwestern boys.