An entrepreneur we met in DC, Mili Mittal, told us about her site, mor.sl. Unfortunately, she didn’t describe it fully, leading me to believe it was far less cool than it really is. That failure has been averted thanks to a quick browse and subsequent salivation.
Mor.sl is, in short, a recipe recommendation engine. You like you some cheese? They got you some cheese. You like you some pasta? Bang. Pasta. The site is arguably beautifully designed – it looks a bit like Pinterest – and it looks nothing like the ad- and text-heavy recipe sites like Epicurous.com. Mili describes it as “Pandora + Amazon for cooking.”
“We just launched the site in mid February and I’m estimating 25% growth in unique visitors for month two,” she said. She founded mor.sl because she noticed she wasn’t eating enough healthy food.
“My mom worked night shifts when I was growing up. She left the house at 4pm, came home at 8pm on her break, whipped up a fresh-from-scratch, healthy Indian meal, fed the family, and left to go back to work at 8:30. I was privileged to eat healthy fresh food every single day growing up, but somehow, as an adult, I never managed to provide such meals for myself,” she said. “Nearly all my friends – busy professionals – faced the same problem. They wanted to cook at home, but they lacked the mental model required to plan and cook gourmet meals efficiently throughout the week. So, we decided to automate the mental model with mor.sl.”
To help you cook like Mili’s mom, mor.sl recommends recipes based on a few simple criteria. The system works via tags and hand curation and all of the recipes – at least the ones I’ve seen – look delicioso. The key to the recommendation engine is a simple quiz that assesses your skill level, allergies, and food preferences.
Mor.sl wasn’t always supposed to be a recommendation engine. In fact, Mili originally wanted to download users’ purchase history and offer recipes based on data stored by Safeway. They built it, but found some problems with the system.
“After testing that product with users, we realized it required far too much user input to stay accurate after the first set of suggestions. We learned that the user inputs had to be optional and minimal, so we flipped the model to more of a Pandora-style recommendations engine. This, in turn, forced a business model pivot which we’re testing now,” she said.
Future improvements will allow you to plan meals and shopping trips as you grab recipe recommendations. Mili told us that she loved all of the recipes on the site but said she really loved one recipe in particular.
“It’s difficult to choose amongst recipes from roughly 35 amazing food bloggers and publishers. I’ll have to be honest though – my favorite is one that comes from our earliest set of recipes – before we had any content partnerships: my mother’s recipe for Yellow Dahl (basically, Indian lentil soup). Comfort food is where it’s at,” she said.