KitchenPC is a new Seattle area recipe search site whose goal is to help you find the best bets on everything from chocolate chip cookie to lasagna. Developed by former Microsoftie Mike Christensen and launching today, KitchenPC faces a slew of rivals, from Seattle area companies like AllRecipes and BigOven to the Web sites of established media brands like The Food Network and Martha Stewart.
Here’s more from our chat with Christensen, who calls all of the competition “healthy,” for the latest installment of Startup Spotlight.
Explain what you do so our parents can understand it: “KitchenPC is the most powerful recipe search engine on the Internet.”
Inspiration hit us when: “I’d love to claim there was a moment where the light bulb lit up, but in reality KitchenPC has been an evolving idea. It started out as a hardware device that would assist you in the kitchen. From there, it moved towards an online meal planner. Today, it’s a powerful recipe search engine. Tomorrow, I imagine it as a platform to help bring the power of software into the kitchen. I try to constantly be inspired, which is perhaps also a weakness since eventually you have to focus on one thing.”
VC, Angel or Bootstrap: “Bootstrap. The only thing that matters is building a product people want to use. I say consult part time to pay the bills, and get some traction going on a product people love. After that, VCs will be trying to setup meetings with you!”
Our ‘secret sauce’ is: “We taught a computer how to really understand a recipe. What does it taste like? When would you most likely eat it? Is it healthy? What would you need to buy at the store to make it? Most search engines just know how to parse text. We can really dig into a recipe and look at it from a culinary point of view. If a search engine doesn’t really understand what it’s indexing, it can’t really help users find what they’re looking for.”
The smartest move we’ve made so far: “Blogging. When I first started KitchenPC, I made the commitment to blog about everything I did. I took the approach that “stealthiness” is not a good thing in the startup world, as your ideas are probably not as valuable as you think they are. I’d much rather be in a position where I’m sharing my thoughts and ideas, and allow strangers to participate in their development. You’ll win by outmaneuvering your competition, not by coming up with some billion-dollar idea on your own. Today, my blog gets almost as much traffic as KitchenPC, and has helped forge more business connections than would have ever been possible for an introvert like me.”
The biggest mistake we’ve made so far: “The first version of the site was an online meal planner. It’s been up for about two years, with only a moderate amount of success. I assumed people wanted to plan what to cook each day by dragging recipes onto a calendar and printing out a shopping list. In reality, very few people work this way – definitely not enough to build a business around. I spent way too long building a product without understanding potential customers first. When I took a step back and looked at how people were actually using the site, it became clear they were using it as a tool to find recipes. Luckily, I had all the pieces in place to pivot the site into an incredibly powerful recipe search engine, with a much lighter emphasis on meal planning.”
Would you rather have Gates, Jobs, Zuckerberg or Bezos in your corner: “I would never dare disturb Gates, as he’s working on much bigger problems than recipes. Since right now, we’re a consumer app, I’d have to go with Jobs. He just knew how to connect with consumers. He didn’t invent things, but he could take an idea that never really took off and turn it into a perfectly polished product that simply worked. However, if I eventually pivot towards a B2B company, such as licensing the data behind KitchenPC, I’d have to ditch Jobs and tag in Bezos for sure. I’d love a partnership with Amazon Fresh!”
Our world domination strategy starts when: “Hardware manufacturers start building appliances that bring the power of the PC into the kitchen. Powered by the KitchenPC back end, kitchen appliances can start organizing your meals, managing your shopping list, and letting you know when you’re low on milk. It’s time someone did for the kitchen what TiVo did for the living room. At least that’s what I’m led to believe from every futuristic sci-fi movie I’ve ever seen!”
Rivals should fear us because: “We spent years building a truly normalized recipe database. We have a consistent set of ingredients, complete with accurate metadata on each. We know how much a cup of shredded cheddar cheese weighs so we can deliver accurate shopping lists. It would be nearly impossible to convert an existing recipe database into our format, and pretty much every major feature of the site takes advantage of this.”
We are truly unique because: “We’re the first recipe search engine that really took the time to ask what people want. Turns out, when most people Google for recipes, they’re looking for more than just one. KitchenPC lets you group recipes you find into menus, allowing you to plan events and easily print out a shopping list for several recipes at once. We also found that there are two reasons people look for recipes. One, they want to first figure out what to cook, and then go buy the necessary ingredients. Two, they have a bunch of leftover ingredients and want to know what they can make. Most recipe search engines only handle the first scenario. KitchenPC allows you to type in what you’re trying to use up (as well as amounts), and it will come up with recipes that will try to efficiently use the ingredients given, while requiring the least number of new ingredients.”
The biggest hurdle we’ve overcome is: “Getting content! This is just something engineers don’t think about, as we express the solution to every problem through lines of computer code. If your startup depends on content before it can be valuable to anyone, go ask yourself where that content will come from. I wasn’t able to import it from other sources, as the normalized format I store recipes in made that extremely difficult. Only around one in a thousand users actually uploaded a recipe, so I couldn’t crowd source it. I tried outsourcing manual data entry on VWorker, and that was a giant disaster as well. The new version of the site uses natural language processing and an in-depth knowledge of ingredients to automatically parse and understand recipes, which took almost a year to get working accurately.”
What’s the one piece of advice you’d give to other entrepreneurs just starting out: “Don’t try to invent something brand new. I thought people wanted to plan meals online, and it hadn’t been done before. The reason it hadn’t been done before was because no one wanted to plan meals online. Find something that everyone uses, and build a better version of it. If I were starting again, I’d find an industry that was dominated by one piece of overpriced software that everybody hated. I’d talk to those people, find out their pain points, and try to deliver a better solution for half the price.”