Reflecting on things I failed to get done at Google
And what I can learn going forward about pitching new things
I worked at Google sometime 2012-2014 and I got nostalgic for the projects I never managed to get off the ground.
I wanted to preserve these ideas on my blog but while I was thinking through them I realized there’s also some interesting lessons to be learned about how not to pitch innovative ideas.
All of these were ideated and pitched sometime 2012-2014 while I was working at Google Creative Lab or as part of the Google Jamboard team.
1 - Work Teams
The idea: a workspace for Google teams, combining Google+ streams, Gmail and Drive into one unified experience. This was before Slack was “a thing” and frankly ahead of it’s time.
How I pitched it: hallway conversations with execs (I recall one SVP pulling me aside and telling me to work on this instead of everything else, but.. not to tell anyone).
Why it failed: I think two reasons it failed. Firstly I was pitching it inside the Google Creative Lab and the lab hated working on anything that wasn’t consumer facing (this was mostly before the consumerization of the B2B landscape). Secondly, I wish I’d had a working prototype - I pitched it by mocking up the UI instead of mocking up the work-flow. I think showing a sample project in-flow would have been more powerful than presenting static mocks of UI that frankly were hard to understand (I’m not a designer!)
2 - Google Map Search 2.0
The idea: let you search for “things” on Google maps, not just places. Still today in 2019 when you search on Google Maps for something like “HDMI cable” or “Lamb Tagine” or “Comedy night” you get businesses and places as results - not the objects, events or information you were looking for. This is a self-fulfilling prophecy because if those searches don’t work then people don’t search for them and so on.
How I pitched it: I pitched it in two ways. Firstly by building an actual MVP. I’m still surprised I pulled this off but I built a working prototype in Google Sheets using scripts to basically first run a Google Maps search, then find all the websites for the listings and then search each of the websites for the keyword used and then filter the results. It actually worked pretty well which I was amazed by and the underlying technology was simple - just combining the Google Maps places info with the Google Search crawl info of what was available. The second way I pitched it was by a very rough market sizing of the ad opportunity on Google Maps and how that would be dramatically expanded if we could shift consumer behaviour from searching for places (where relevance for ads is low) to searching for things (where relevance for ads is high).
Why it failed: I actually thing this entirely failed because of the Google Sheets prototype. It was a functional working data model but had zero GUI and when Andy Berndt looked at it he completely shut it down.
3 - Google Illustrator
A vector based canvas model. This was before the collaborative design tools like invision, figma etc.
How I pitched it: using a blog post! I built a
4 - Google Cube
Ok ok, this one might not have been the most practical based on Lamps http://berglondon.com/blog/2012/12/19/lamps/ it was a little cute cube that would sit on your kitchen / living counter in your home and would do two things. Scan paperwork that you put next to it and also project information onto a surface. Basically a short throw projector combined with camera.
The core utility of auto-scanning paperwork to Google Drive actually got some people excited but in hindsight I was so close to the idea of alexa/google home but I just barely missed the idea of voice+screen in your kitchen. I think if I’d pushed this a little harder I could have built some prototypes but the team I was on at the time was focused on building what would become Google Jamboard and they didn’t have the bandwidth for it.