A look back at Diane Greene's tenure as Google cloud boss: a struggle to catch up with AWS and Azure, and tension with Sundar Pichai on GitHub and Project Maven
> Google Git —> G Git for Domains —> Google Cloud Git —> “An Update on Google Cloud Git” drama play out over the next two years. http://twitter.com/... Joseph Jacks / @asynchio : .@GitHub will be a $100B+ asset for @Microsoft in < 10 years. $7.5B was a very good / well priced deal. http://twitter.com/... @markcallaghan : Google had a nice alternative to Github and shut it down in 2016. I lost some work history when that happened — wiki pages and repos from the Google MySQL effort are gone although I might have them archived somewhere. http://twitter.com/... @dhh : Google bid $6 billion for GitHub, but wouldn't up the offer to counter Microsoft. I wouldn't have said this a few years ago, but I'm glad Microsoft won the bidding war. Grave risk Google would have lost interest in 6 months after the next exec rotation. http://www.cnbc.com/... Thanks: @mattrosoff
Context & Ripple Effects
The arc starts with Google's 2015 appointment of Diane Greene — bought alongside her startup Bebop — to run all cloud businesses, followed by a 2016 retooling of every cloud service around business customers. The CNBC retrospective closes that chapter by naming what held the effort back: a persistent gap behind AWS and Azure, plus two flashpoints with Sundar Pichai — a reported $6B GitHub bid that Google declined to raise before Microsoft closed its $7.5B GitHub acquisition, and Project Maven, where employee outcry ended Google's Pentagon AI work.
The GitHub thread is the sharper wound for developers: Google had run its own internal Git service and shut it down in 2016, deleting repos and wikis some engineers never recovered, so when the acquisition went to Microsoft, Google was left with neither the asset nor the internal alternative. The Maven thread has a longer tail — three years after the outcry, Google is aggressively pursuing a major Pentagon cloud and AI contract, reversing the stance that defined Greene's tenure.
First-order effects
- Microsoft converts its $7.5B GitHub purchase into direct leverage over the 36M registered developers Google courted, while Google's lapsed $6B bid and shuttered internal Git service leave it with no equivalent developer-platform asset.
- Greene's exit from the Maven fight hands the defense-AI opening to rivals willing to take the work — a gap Google itself later moves to close with its renewed Pentagon contract push.
Second-order effects
- Microsoft's open-source positioning post-GitHub pressures Google and AWS to compete for developers through tooling and pricing rather than infrastructure alone, raising the cost of the catch-up game Greene was hired to win.
- Employee activism becomes a factor competitors must price into government bids: Google's Maven retreat shows that workforce opposition can force a hyperscaler out of a contract category entirely, reshaping who bids on defense work.
Third-order effects
- Hyperscaler competition structurally shifts from raw compute scale toward ecosystem ownership — developer platforms like GitHub and trusted enterprise/government relationships become the moats, which is why the third-place cloud can't close the gap on infrastructure spend alone.
- If the pattern holds, big-tech culture becomes a strategic variable in public-sector AI: leadership must weigh employee revolt against contract revenue, and reversals like Google's Pentagon return show those positions are negotiable under competitive pressure.
The trend: Cloud market position is being decided less by infrastructure scale than by control of developer ecosystems and the political cost of government AI work — the two fronts where Greene's Google stumbled.