Blog‎ > ‎

2009/12/24 - Do No Evil, A Case Study

posted Dec 23, 2009, 9:26 PM by Rick Anderson
It's the wee hours of the day before Christmas and I was thinking about a variety of things related to Google, their "do no evil" mantra, and the recent memo by Google's SVP in regards to what it means to be open.  That led me to running a Google query on "google acquires" which led to a Wikipedia article on Google's list of acquisitions.  Although the Google query results display reCAPTCHA as their latest purchase on December 16th, Wikipedia has their last acquisition as Appjet (Etherpad) on December 4th.

Following along the Etherpad links I nosed around a bit on their blog and found an interesting read that makes a great case study in how a big company such as Google can ingest a startup and the fear, uncertainty, and doubt that follows in it's user base.  There's a couple of interesting points to make - 

(a) this is a fascinating study in change management that's ongoing and something that I've spent my 17 years in business practicing - in this case they're moving a user base from Etherpad to Wave; 

(b) understanding the user base of a startup or any new technology - there's a tremendous amount of publication done by Geoffrey Moore in regards to early adopters of technology - they are, if nothing else, passionate about their decision to support the startup/new technology;

(c) the purchase time frame occurred at about (or just a bit before) Google's Senior Vice President of Product Management, Jonathan Rosenburg, internal posting of his memo (memo to his product managers) on the meaning of open and how does that leadership statement drive every day actions;

(d) "do no evil" - you be the judge and what does the perception of that statement mean to Google, developers, users, the market.

Etherpad started out as a simple collaborative tool to showcase the AppJet engine - the AppJet engine being a web development platform started by three MIT grads (2 of those being ex-googlers).  About eighteen months later, the work on AppJet was discontinued as a public project as Etherpad started to take off and be the focus of the team.  It appears that primary school teachers and students were early adopters of the technology as they could work on projects in really real time together.

Etherpad is a web based really real time collaborative editing tool.  Essentially a text editor / word processor on the web that allows several people to all work on the same document in "really real time".  Really real time means that the document is updated to everyone's browser every half second showing all the input happening at the same time.  This differs from Google's Docs as there is a 15 second delay between entry and presentation.

The trade industry originally expected Etherpad to compete with Google's Docs as can be seen in the Techcrunch review of July 23rd, 2009.  However, the folks at Etherpad were really competing with Google's shiny new real time collaborative technology called Wave as can be seen by Daniel Greenspan's blog posting comparing and contrasting Wave to Etherpad (and claiming that Google used Etherpad as an example in their initial presentation of Wave).

The rest as they say, is history... several months later Google acquires AppJet/Etherpad and assigns them to the Google Wave team.

That's where our case study starts...

December 4th - Google acquires Etherpad for an undisclosed sum and the company blogs the going forward terms and conditions to the user base - Etherpad will be shut down by the end of March, 2010 - however, no new pads (pads are the editable collaborative documents) will be allowed to be created - there is a download tool available to retrieve your data which needs to be done prior to end of March or else lose your data - all registered users will receive a Google Wave invite, and no further charges will be made for usage between December 4th and the end of March (essentially making the service free while the user base migrates).

Now remember, Etherpad's user base is in the early adopter tail of the curve - meaning they are radical followers of the new technology - yes it's nice that they can edit existing pads for four months, yes it's nice that they can download their data, yes it's nice that they'll receive a Google Wave invite, yes it's nice that they can use the service for free until it's shut down; BUT.... they just got locked out of their ability to create a new pad in what they consider their tool.  Understand that the words "their tool" in this sentence is critical as early adopters are cult-like in their following.  If you read through the blog comments, you'll see that there are numerous posts of how x user just showcased the product to their friends and co-workers - they convert others like religious zealots which is what every startup is looking for in order to get them across the chasm and into the mainstream.

Aaron Iba quickly posts a comment to the blog... "Hey guys, thanks for your passionate feedback. We hear you loud and clear, and are working with Google on a new transition plan."

December 5th - Aaron Iba makes a new blog posting announcing that Etherpad is back online until it can be open sourced.  Three points are made (a) pad creation is possible until the product is open sourced; (b) the technology is going to be open sourced; (c) Google Wave invites will go out to all Etherpad users within a couple of weeks.

While some users still aren't happy, in general the comments to this blog posting are positive or at least a let's wait and see if this open source action is really going to be taken.

December 17th - Aaron Iba releases a new blog posting announcing that Etherpad is now open source with a declaration of "we hereby release all the source code to Etherpad".  The entire world can now setup Etherpad servers simply by downloading the code from Google's site - - not only that, but they are allowed to dig into the code, make modifications, and improve upon the code.  Additionally, the draconian ceasing of new pads has been diminished to a we'll leave it open until it tapers off approach.

In our case study, where does that leave the user base - they're now given three choices (a) convert to Google Wave; (b) join a hosted Etherpad community; or (c) find a completely different solution.  It will be interesting to see what further actions occur but a couple of points:

* Rosenburg's directive of being open is put into play - essentially Google is saying this - we'll be open and we're not afraid because we're innovative - if you stuck with us (in this case Google Wave), we'll continue to innovate and deliver a superior package - however, if you don't believe in us, then you're free to use what we've developed in the past.

* The Etherpad community has sprung up...  Piratepad (although how are you a pirate of open source code? - but it sounds cool); iEtherpad (I think he'll have to put together a new front end instead of copying images from Etherpad); and Primary Pad (marketed towards primary schools and looks to be a positive group to carry the torch - and the funding to really keep their servers up long term).

Finally, going back to the original headline of this post - Do No Evil - this being a mantra carried forth by Google in their endeavors as a business.  Trolling through the comments it is obvious that a majority of the user base felt betrayed - a David and Goliath whereby David got swallowed.  But is that really evil?  Certainly, the closing of new pads was a bad move - but really, in my opinion, that was simply Etherpad misunderstanding the who of their customer base - and ultimately (and very quickly) a different plan was put into place whereby Google's openness directive is truly tested.  There are some commented opinions that selling out or Google devouring Etherpad was evil - but is it really?  

In it's simplistic form, is not the goal of every startup to be bought?  I think that is every startups wildest dream - work crazy hard - be innovative - conquer what can't be done - and then sell.  It is obvious that the founders had ties into some very big technology companies when you read Daniel Greenspan's Listening With Big Ears post - there he talks about how they innovate around user requests and follow the guidance of world-class software product €”guys like Paul Buchheit (creator of Gmail), Dustin Moskovitz (Founder of Facebook), David Jeske (engineering manager of eleven products at Google), Paul Graham (founder of ViaWeb and YCombinator), Barney Pell (Founder of Powerset), and Mitch Kapor (Founder of Lotus and Mozilla Foundation) - and remember, most initially came from Google.  The founders game plan was to create an innovative product and sell the product and its user base.

That's exactly what happened - the guys at Etherpad worked for two years developing an unique and advantageous technology that Google needed - "really real time collaboration" - and in the end they profited from their work.  Initially the transition wasn't handled in the best method, but the fumble was quickly resolved and the users have a legitimate path forward.

Two thumbs up from me - way to go guys and good luck with Wave!  My only request is that I'd love to see the "really real time collaboration" implemented in Google's docs, spreadsheets, and presentation prior to incorporation into Google Wave.  I believe that Wave's time will come, whereas Google Docs is already here.

As a side note - if anyone needs a Wave invite - please contact me!