Google Insights for Search

» see google.com/insights/search as it lived, on the wayback machine

~ the goodbye ~

I used Google Trends. Everyone who was online and curious used Google Trends, because it was free and it drew a graph and the graph felt like it meant something about the world. What I did not know at the time is that Google was running two of these, and the one I had been using was the demo.

Insights for Search went live on 5 August 2008. Same dataset underneath, same normalized index where the peak week reads 100 and everything else is a fraction of it, but with the controls bolted on. Category filters pulled apart a query that meant two different things, so that a car brand and an animal with the same name stopped being one line. Geography went down past the country into region and city, and the date range was yours to choose instead of whatever window Trends had decided to show you.

By 2008 I was at the end of university with TTNet broadband at home, which is to say I had stopped being impressed that a page loaded and started being interested in what sat behind it. Trends was the first tool I can remember that let an ordinary person look at aggregate human behavior rather than at a document. It never once occurred to me to go looking for a better version of it.

~ the chronicle ~

~ what the second one could do ~

The feature that justified running two products was the rising-searches list. Top searches told you what was big, and what is big is mostly last year's news carrying a larger number. Rising told you what had gained the most against its own recent baseline, which is a different question and a far more useful one. A term going from a hundred searches to a thousand is worth more to somebody buying keywords than a term that has sat at a million for three years.

It also let you put two regions or two time ranges in the same chart and compare them directly, and it lived in the tool list next to AdWords rather than on a friendly public address. That placement was the honest signal about who Google thought it was for. Trends was for the press. Insights was for the people spending money.

~ why one of them had to go ~

Two products, one dataset, four years. That is not a strategy, it is a leftover, and Google eventually gets around to leftovers.

On 27 September 2012 Insights for Search closed and its filters moved into Trends. The merged tool kept the Trends name and the Trends address, and google.com/insights/search redirected to it. This is one of the very few Google shutdowns where nothing actually died. The category filter survived. The city-level breakdown survived. Only the second door closed, and the people who noticed were the ones who had bookmarked it.

~ what the data outlived it as ~

The more interesting afterlife belongs to the numbers rather than the interface. Somewhere in this window, aggregate search volume stopped being a marketing input and turned into a research instrument. Economists started using it to estimate unemployment and retail activity ahead of the official statistics, on the reasoning that people search for a thing before they do it. Google's own team built Flu Trends on exactly that premise, had a famous run of accurate years and then a famous inaccurate one that retired it.

Trends is still there in 2026 and still free. Insights for Search is a redirect that stopped redirecting at some point nobody recorded. The tool that got absorbed was the one that could answer a real question, and it was absorbed so cleanly that almost nobody now remembers there had been two of them.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »