Pattern Finder

Why Google Apps Hasn't Taken Off in Large Enterprises

Last week, Fortune magazine published an article ("Google's tough sell to Corporate America") and blog entry ("Gen Y takes to Google Apps") on Google Apps, noting that it was popular with Gen Y but not with large enterprises. To put it in perspective, Fortune noted that Microsoft made 3,000 times more revenue than Google off of its Office software: Microsoft sold $12.2 billion of Office software in 2007; Google sold $4 million of Google Apps. Large enterprise sales have been few and far between. Fortune notes:

So far, though, the largest Fortune 500 company to use Google Apps is Sanmina-SCI (SANM, Fortune 500), an electronics manufacturing company that has 900 of its 45,000 employees using the full package of applications. Another paying customer is Valeo, a publicly-traded French automotive supplier. And General Electric (GE, Fortune 500) is running Postini, an anti-spam technology that Google acquired last year, for its 300,000 employees.

I'd predicted this would be the case when I published my first report on Google Apps a year ago (Burton Group clients can access the April 2008 revision here). So why is that? Google Apps has done well within small- and medium-size businesses (SMBs); large universities such as Arizona State and Northwestern have also adopted it.

It comes down to four reasons. First, Gen Y is not yet in charge of IT strategy at large corporations. Google is banking on Gen Y employees demanding Google Apps--"'We’re not waiting 15 to 20 years. The generation of new [office apps] users is already here,' says Glotzbach"--but there's a difference between a minority of employees (without seniority) pushing for them and the older generation of CIOs mandating them. While small companies headed up by Gen Y workers are happily using Google Apps, large enterprises headed by 50- and 60-year olds who are used to Microsoft Office are taking a wait-and-see attitude.

Second, Google's consumer-centricity has done it a disservice here. Google basically took a consumer service, slapped an Enterprise name on it, and assumed companies would lap it up. Unfortunately, Google neglected to add features (so far) that competing packages have offered for more than a decade, such as tables of contents, footnotes, sophisticated e-mail distribution lists, and records management capabilities. (Although Fortune quoted me as saying that the lack of footnoting was a major deficit [I'm sure because it was an example everyone would understand without needing an explanation], the lack of records management capabilities for documents and the need for better distribution lists are the bigger problems.)

Third, the incremental benefits of using Google Apps are not that great. True, you can edit documents concurrently--which you can't do in Microsoft Office--but otherwise Google Apps is pretty much a straight copy of base Microsoft Office functionality. In other words, Google Apps makes do by copying the market leader in a rudimentary way, rather than taking the gloves off and pounding away at the things Office doesn't do: keeping track of who's reading what, suggesting other documents it might be useful to refer to, recognizing that a document similar to this one has already been written and suggesting that it might be easier just to modify the earlier document. Software as a Service (SaaS) office productivity suites can do so much more than software because they can watch the community working, and thereby help the community leverage what it already knows. Amazingly, Google doesn't take advantage of that architectural difference.

Fourth, large enterprises are behemoths--enterprises aren't inclined to change and there's nothing compelling in Google Apps at the moment to make them change (and the attractive low price isn't enough). Word processors replaced typewriters not because they did the same thing, but because storing digital documents made revisions so much easier--word processors offered a quantum leap in productivity. Google Apps doesn't offer that productivity leap yet.

When it does--or a Google competitor does--I think large enterprises will do well to adopt the new technology. Until then, more nimble SMBs and cash-crunched universities will be the sweet spot for Google Apps.

Run an Accessible Web Site or Pay the $6 Million Penalty

A cautionary tale for companies that neglect the look, feel, and functionality of their web site. On August 28, the retailer Target agreed to pay $6 million in damages for running a web site inaccessible to the blind. 

Uh, Has Anyone Thought About Licensing?

It's been an interesting week or so in terms of movement towards, "I don't care where the system is, I just want it to run." Microsoft announced it was relaxing its virtual machine licensing requirements (CCS commentary here) and Cisco announced it was acquiring PostPath (CCS commentary here) as a way to offer Outlook-friendly SaaS-based e-mail.

 

However, as we move towards a world where IT departments deliver solutions that can run anywhere--in-house or in a condo (on a virtual machine) or in the cloud--one thing that all vendors quickly need to figure out is how to license all of this in a coherent fashion.

 

Back when the dominant model was software upgrades that came out every 9 or 18 months, enterprises and vendors could afford the luxury of sitting down and spending days or even months hashing out the licensing details. (I had one client recently tell me that he's spent nine months in his latest licensing go round with Microsoft and it's still not finished.) However, virtualization means that virtual servers can come and go within minutes, and workers can sign up for SaaS solutions within minutes as well. Enterprises also want to move towards an instant provisioning model, where they may have n servers in house but then contract with the cloud to augment those servers for a month or so for a specific project.

 

In other words, when it comes to licensing, rigid paper contracts have to give way to online checkboxes. The business decides, "OK, we need 250 SaaS-based e-mail mailboxes for the company we acquired yesterday," it ticks off the required checkboxes, the online mailboxes get provisioned within the hour, the licensing dashboard is updated automatically, and a correct bill goes out the next month. In less than a day, the system and affiliated licensing is taken care of.

 

It may seem like a dream now, but this is where companies such as Cisco, Google, IBM, and Microsoft all need to get to. And I would argue the company that does it first and well will gain a huge competitive advantage.

Ruminations on Netflix

As noted in technical publications and the general press, Netflix was "shipment challenged" last week. The company's shipping system went down, so it was unable to mail out DVDs on Tuesday the 12th, but was able to send out some DVDs on Wednesday and Thursday. The system was finally up and running at full capacity on Friday the 15th.

With an episode like this, two questions come to mind: First, what happened, and second, how well did the company handle it?

As to what happened, one rumor is that it was an Oracle upgrade gone bad--that after upgrading its Oracle cluster and everything seeming to go well for awhile, Netflix found problems. This is a data processing problem from hell, because sometimes you can't easily roll back. Let's say you implemented a new record format during the upgrade. If you roll the database back to what it looked like before the upgrade, you've now lost the new transactions in the system. At the very least, you have to figure out how to convert the new record format back to the old format (and perhaps lose some data in the process). Since Netflix seems to be having problems reconciling the status of recent DVDs--saying a DVD will be shipped when the customer already has it--I suspect that some sort of rollback issue occurred.

I went through a nightmare similar to this thirty years ago, when I was a programmer at John Hancock. In 1978 we implemented our first major individual life insurance upgrade in 15 years. I was responsible for a system called Pending Status. After installing 78 New Products, the system breezed through our comprehensive QA tests, and we all celebrated. However, there was a bug lurking that became visible only after the system had been running for a week. I got the call from Operations at 2 AM; I didn't figure out what was wrong until about 12 hours later; and it took me another 12 hours to rerun all of the week's transactions with the fix in place. So the system was down for 24 hours, fixing a system with a relatively low transaction rate. (Hancock approved approximately 1,200 policies a day at the time). Although I had multiple Senior Vice Presidents screaming at me, customers never knew of the problem--we just saved up the day's transactions and ran them when the system was fixed.

Unfortunately, Netflix didn't have this luxury--demanding customers notice when the DVD they were expecting doesn't show up. Which explains why stories about the snafu appeared in a wide range of publications such as Computerworld, FOXBusiness, Reuters, and The Wall Street Journal. The blogosphere also went wild, first debating what was wrong and then debating whether the 15% credit Netflix was giving affected customers was enough. Some said no, others said yes.

I'm a Netflix customer, I'm getting the credit, and I say yes, that was enough. I think Netflix did an excellent job of handling the problem. I can imagine the high-level discussions that went on within Netflix--"Should we do a refund, how much should it be?". (Disclosure: Visualization is especially easy for me since Barry McCarthy, the CFO, is a college classmate of mine. Barry's a straightforward guy--his Linked In entry for the time he was at CSFB notes, "Made every investing mistake imaginable.")

Netflix's proactive approach is one that many companies should adopt. Despite your best efforts, problems happen. Companies need to: (1) acknowledge the problem, (2) let customers know what's going on, and (3) proactively make it right in some way. Having worked in Customer Support in a previous life, I've done it the right way and the wrong way, and the right way is much more preferable for all concerned.

So I'm a happy Netflix customer. Netflix needs to check their operational procedures and make the necessary adjustments, but I'm not moving over to Blockbuster.