4 ms·

I wonder if Ozzie considers Groove to be one of the products he expended energy on and failed. From my perspective it seemed to actually be succeeding as a deep

I wonder if Ozzie considers Groove to be one of the products he expended energy on and failed. From my perspective it seemed to actually be succeeding as a deeply network-aware thick client for collaborative knowledge work when Microsoft acquired it and killed it off. Or if not outright killed it, then managed it to death by strapping it to the SharePoint boat anchor. I wouldn't count that as a failure per-se. It didn't succeed, but wasn't really given the chance to do so, which seemed to be what Microsoft wanted to get out of the transaction.

gerikson 6y ago
I think OneDrive had its origins in Groove back in the day.
- Arainach 6y ago
  No, SkyDrive was always centralized server-side without any of the conflict resolution coauth concepts that Groove had. It was a completely separate project.
  - Exmoor 6y ago
    What about OneDrive For Business? I recall the executable was, confusingly, Groove.exe for quite a long time.
    - Arainach 6y ago
      As I recall, OneDrive for Business had absolutely nothing in common with OneDrive except that marketing in all their wisdom decided to throw that name on it for maximum confusion (a very popular technique at Microsoft - see also Skype for Business). I don't know its history, but assuming it evolved out of some Sharepoint product it wouldn't surprise me if it had some Groove DNA.
      - Exmoor 6y ago
        > a very popular technique at Microsoft Indeed. See also: Surface...everything.
        lukeck 6y ago
        And slapping .net on everything.
    - gerikson 6y ago
      That's where I got the impression from.
rozzie 6y ago
Undeniably the product failed commercially. By the time we decided to sell the company, the product was doing ~20M+ of revenue but on a low growth rate, after having gone through ~140M of capital with (not unreasonable) investor expectations of unicorn-like returns. That said, I absolutely loved Groove the product and Groove the team. It was a dapp with a wonderfully-usable interface, built on a pure P2P blockchain-like infrastructure that achieved distributed consensus for generalized transactions. It had an extremely loyal following in the NGO and GOV space. They "got it" as to the unique benefits of a secure end-to-end encrypted system for small team collaboration. But the enterprise and consumers chose centralization as the winning architecture, and the rest is history.
- webmaven 6y ago
  > Undeniably the product failed commercially. By the time we decided to sell the company, the product was doing ~20M+ of revenue but on a low growth rate, after having gone through ~140M of capital with (not unreasonable) investor expectations of unicorn-like returns. Groove's revenue exceeded expenditures, though, right? Or could have? I'm not sure it was reasonable to expect success at that scale for an entirely novel class of application in an era where startup investments were still expected to 'mature' only after about a decade or more. Groove was only ~8 years old, and 1.0 had been released ~6 years prior. > But the enterprise and consumers chose centralization as the winning architecture, and the rest is history. As I recall, around 2.x you were more-or-less switching from a distributed architecture to a decentralized one, but... your point is taken. I suppose I should admire that you saw the writing on the wall as soon as you did. The wheel keeps turning though, and I wonder if another stab at the idea could succeed today. Certainly the various risks of centralization are much more apparent, and to a broader audience, now.
  - rozzie 6y ago
    Perhaps. Hard to say. The dev team was large by necessity, but the GTM was far too large because we hadn't come to terms with giving up on enterprise and individual sales. Had we focused exclusively on public sector and hunkered down we MAY have been able to be profitable on an operational basis. But we would have had to recap and get existing investors to write it off at a time when the board had two offers on the table that returned some value to investors. Not straightforward. Today it could almost certainly be built MUCH more economically - likely on the not-dissimilar and rich/mature matrix.org infra at the bottom, and using Electron on top. And I would sell not to enterprises nor to decentralization zealots: I'd build a small focused GTM org focused on public sector and those who will need to collaborate when trying to deal with an APT.

gerikson 6y ago

I think OneDrive had its origins in Groove back in the day.

Arainach 6y ago
No, SkyDrive was always centralized server-side without any of the conflict resolution coauth concepts that Groove had. It was a completely separate project.
- Exmoor 6y ago
  What about OneDrive For Business? I recall the executable was, confusingly, Groove.exe for quite a long time.
  - Arainach 6y ago
    As I recall, OneDrive for Business had absolutely nothing in common with OneDrive except that marketing in all their wisdom decided to throw that name on it for maximum confusion (a very popular technique at Microsoft - see also Skype for Business). I don't know its history, but assuming it evolved out of some Sharepoint product it wouldn't surprise me if it had some Groove DNA.
    - Exmoor 6y ago
      > a very popular technique at Microsoft Indeed. See also: Surface...everything.
      - lukeck 6y ago
        And slapping .net on everything.
  - gerikson 6y ago
    That's where I got the impression from.

rozzie 6y ago

Undeniably the product failed commercially. By the time we decided to sell the company, the product was doing ~20M+ of revenue but on a low growth rate, after having gone through ~140M of capital with (not unreasonable) investor expectations of unicorn-like returns. That said, I absolutely loved Groove the product and Groove the team. It was a dapp with a wonderfully-usable interface, built on a pure P2P blockchain-like infrastructure that achieved distributed consensus for generalized transactions. It had an extremely loyal following in the NGO and GOV space. They "got it" as to the unique benefits of a secure end-to-end encrypted system for small team collaboration. But the enterprise and consumers chose centralization as the winning architecture, and the rest is history.

webmaven 6y ago
> Undeniably the product failed commercially. By the time we decided to sell the company, the product was doing ~20M+ of revenue but on a low growth rate, after having gone through ~140M of capital with (not unreasonable) investor expectations of unicorn-like returns. Groove's revenue exceeded expenditures, though, right? Or could have? I'm not sure it was reasonable to expect success at that scale for an entirely novel class of application in an era where startup investments were still expected to 'mature' only after about a decade or more. Groove was only ~8 years old, and 1.0 had been released ~6 years prior. > But the enterprise and consumers chose centralization as the winning architecture, and the rest is history. As I recall, around 2.x you were more-or-less switching from a distributed architecture to a decentralized one, but... your point is taken. I suppose I should admire that you saw the writing on the wall as soon as you did. The wheel keeps turning though, and I wonder if another stab at the idea could succeed today. Certainly the various risks of centralization are much more apparent, and to a broader audience, now.
- rozzie 6y ago
  Perhaps. Hard to say. The dev team was large by necessity, but the GTM was far too large because we hadn't come to terms with giving up on enterprise and individual sales. Had we focused exclusively on public sector and hunkered down we MAY have been able to be profitable on an operational basis. But we would have had to recap and get existing investors to write it off at a time when the board had two offers on the table that returned some value to investors. Not straightforward. Today it could almost certainly be built MUCH more economically - likely on the not-dissimilar and rich/mature matrix.org infra at the bottom, and using Electron on top. And I would sell not to enterprises nor to decentralization zealots: I'd build a small focused GTM org focused on public sector and those who will need to collaborate when trying to deal with an APT.