Humans Behind Search: Hadas, software engineer and trends expert

Share
  • April 20, 2022

Hadas Jacobi is a software engineer working in Search on Google Trends. Hadas has been with Google, based in Tel Aviv since February 2019. She spoke to The Keyword about how Trends has developed over the years and what may lie ahead for the popular Search feature.

Can you tell us a bit about how Google Trends works?

What Google Trends does is take a sample of the searches people make on Google to figure out how high the search interest for a given keyword is at any given time — relative to the total amount of searches. For example, after the now infamous Will Smith Oscars incident, we saw searches for “Will Smith Oscars” increased by 500 times.

We also make an educated guess about where the searches are coming from, and that’s how we are able to display a map of search interest in different places. We do this based on the data flowing through Google Search day in and day out — so we’re able to see the search interest from 2004 all the way up to just a few minutes ago. This is really helpful in understanding how the interest in different topics changes over time and in different places.

A lot of researchers and journalists use Trends as sort of an anthropological tool to give them a view into what interests people in private, but may not be in the collective consciousness. For example, you can use Google Trends to help detect local disease outbreaks long before disease control centers detect them, simply by looking at where people were using Google to look up their symptoms.

What excites you about the future of Trends?

We’re working a lot on how Google Trends can help the world of journalism. As I mentioned earlier, it gives us this unique view into what interests people. We know journalists use Trends to research their stories, but the tool also gives them insight into what topics people are currently interested in, which allows them to tailor their content to their readers’ interests.

We’re really working on figuring out ways to help journalists enhance their storytelling by making the data more relevant to their needs, exploring ways to make things better, faster and more useful for local journalists.

What were people searching most for last year?

First up, “feta pasta.” It was a huge trend last year — and I even made some myself at one point — but seeing that it made the list of top trending five food searches globally did shock me. It can be surprising to see some of your own searches there. There were also a lot of cryptocurrency terms among the top trending searches — things like “ethereum” and “dogecoin.”

What are the challenges of running a product like Trends?

Trends is huge in terms of the scale of data that runs through it. We process a sample of the billions of searches that are made every day around the world, and the amount of daily searches just keeps on growing. We also have to be able to make all of this available to Trends users so that they can look up whatever keywords they want for any time and any place.

It’s pretty crazy the amount of engineering work that has gone into making all this run smoothly over the years — the amount of data grows faster and faster because people are searching more on Google every year. It always amazes me that of the trillions of searches we see every year, 15% are brand new every day.

Looking at Search as a whole, what direction do you see the product taking in the future?

I can certainly talk about what I’d like to see. I love it when it feels like the technology just gets me and knows what I’d like to do next. For example, when searching for a TV show, Search brings up the cast because it just knows that it’s something I might instinctually be interested in.

I’d like to see this mature into other areas. For example, Search could suggest related topics or other things I might be interested in based on my habits and what I just searched. Essentially anything that helps me to continue exploring my curiosities around a particular search I would love.

So if I searched for “weekend hike” it would suggest “local wildlife field guide” and then “baby deer season” because it’s possible I might be interested in that without even knowing that I was. So for me, I’d love to see a continuation of timely, topical suggestions as you go.

Source : Humans Behind Search: Hadas, software engineer and trends expert