Menu Highlights

Food search is the second highest query volume across Google’s products, at when it comes to these dining related searches 80% of users report seeking menu data of some sort. When it comes to manual refinements, adding the term “menu” to a query is the most common clarification after location. And these insights are quite concerning since often times menu coverage for places is low and often there is no lightweight snackable way to get the gist of what is good to eat at a place - what are people mentioning the most in reviews, what are people photographing the most? What if there were a way to see on a business profile all the highlights of a menu that people are buzzing about?

“Menu Highlights” is a new dining discovery feature found in both Maps and Local Search that caters to the 230M users who search for food everyday and enables people worldwide to learn and share the best dishes at any given place.

____________________

MY ROLE

The Food Discovery team caters to the primary user need for menu & dish data in Search by adapting existing surfaces and building new food-forward experiences. As UX lead on Geo Food, I initiated the product effort to design a cohesive and scalable framework for how we display varied forms of dish & menu data, created new contribution flow for attaining user dish data, and championed a dish-first discovery experience - efforts which have since increased the perceived usefulness of Geo surfaces for dining as seen through overall increased engagement.

Mission

Our mission heading into this project was to “organize the world dishes and highlight what’s amazing at a place” - this incapsulates building a rich corpus of dish data using machine learning to scrub existing UGC for dish data, collecting dish inslights from users via new UGC flows for dish tagging, and lastly rank and scaffold the best dishes at a place.

CUJ - Critical User Journey

And the CUJ we kept top of mind at all times was “I want to know what’s offered and what not to miss at a place” - this speaks to a user need for access to both a comprehensive menu (seeing everything a place has to offer, before making a confident decision) and more importantly to learn what the best scaffolded menu items are at any given place.

Strategy

However, one big obstacle loomed – we only had menu coverage for a meager 12% of places. Unfazed, we pivoted strategically, capitalizing on our abundance of user-generated content (UGC) about dishes. We then created what we called "dish clusters," utilizing machine learning to combine photos of the same dish, identify the dish depicted, and subsequently pair it with relevant reviews. This innovative approach transformed our menu coverage, allowing us to overcome the initial hurdle and leverage the power of UGC.

Introducing “Menu Highlights”