Let’s take a moment and consider the early web. Before we Instagrammed our food, Snapped our moments, and Whispered our secrets, we tapped into the expansive World Wide Web using a once-famous directory known as Yahoo!. A team of curators (or maybe just Jerry and David) filtered their favorite websites into categories we could peruse. (Image
Something magical happened. Google crept out of nowhere, superseding a whole slew of search engines (Alta Vista, Excite, Ask Jeeves, etc), crawling the web and using some new-fangled algorithm called PageRank to find us the best results.
It was amazing. As fast as they could, Google scoured the web grabbing information, places, academic research papers, and more. They built translators and dictionaries and calculators, all into a simple search field with a seemingly infinite knowledge base. Information was attainable that Yahoo! never would have delivered.
But Google was built for a PC-era world. This was 1998, 17 years ago, long before the advent of the iPhone. We had slow processors, big keyboards, text-based input, and all the time in the world as we sat in the comfort of our living rooms.
Fast forward to today. We’re using our smartphones on the go, often while driving or waiting in line at Starbucks. We have small keyboards and location-aware devices, yet search is powered the same way.
The search box doesn't know what you want, so you have to type or talk a query, which is cumbersome on mobile. The time it takes to search is longer than it needs to be (can take up to 60 seconds to open your browser, type your query, scroll through the results, tap on the website you think has the best result, then look for details. Mobile search doesn't use your precise location, the weather, time or day or time of year, or personal preferences when serving results. The results are largely text-driven, which isn't optimal for quick decisions, and the whole experience isn't social.
This is a huge opportunity. More than 24% of all search queries are for local inquiries, with that number growing to 40% on mobile devices (link
So what would local search on mobile look like if done right?
For starters, it wouldn't require typing. Local search done right would utilize your context such as time of day, time of year, location, weather, what's trending nearby, etc. The system would learn from your behavior, enabling smart recommendations personalized to you. The display would be highly visual, reducing the amount of text and making it quick and easy to make a decision. Finally, local search done right wouldn't feel like a chore, it would feel like a game, it would delight. Local search done right would be social.
This is the vision for Yeti. Where Google showed us that a search engine is more powerful than a directory, we believe a recommendation engine is the answer to nailing local search.
We’ve imagined a self-learning, semantic platform that aims to answer your search queries without a single keystroke. By understanding your environment, your personality, and what similar people have found valuable, Yeti makes recommendations for your offline world.
By creating a binary decision matrix, Yeti learns with every swipe about your interests as well as the content, so we can target the best information for the right users.
Today Yeti is still in a very early stage of its lifecycle. With members in more than 100 countries who have swiped over 2,000,000 times, we're off to a great start, but there's still much to do.
This simple but elegant UI could be the future of local search. We're pretty excited about what's possible by connecting people based on location, making nearby information available, and decreasing cognitive load for everyday tasks that answer the question "where to go and what to do".