Does payback on email alone make Google Apps' $50/yr. worth it? (Docs, etc? That's icing)

Does payback on email alone make Google Apps' $50/yr. worth it? (Docs, etc? That's icing)

Summary: Whenever I talk about doing something crazy like using Google Apps instead of other office suites, a conversation that's usually driven by my beliefs about computing in the cloud, there are some number of ZDNet readers that chalk it off to insanity or a momentary lapse of reason. Or, maybe I forgot to take my meds that morning.


Whenever I talk about doing something crazy like using Google Apps instead of other office suites, a conversation that's usually driven by my beliefs about computing in the cloud, there are some number of ZDNet readers that chalk it off to insanity or a momentary lapse of reason. Or, maybe I forgot to take my meds that morning. But, on the heels of some unrelated events and discussions that took place over this past week, not only am I more convinced of going after a cloud computing approach like the one offered by Google Apps, I've got a different way of explaining the ROI that will probably resonate with many IT managers, small business people, and CFOs better than any other previously used rationale.

First, this is and isn't a story about Google Apps.

It is a story about Google Apps because of the way Google Apps is currently the best example of how a scale service provider can deliver certain IT functions at such a reduced cost ($50 per user per year) that it is almost impossible not to consider.

For small and medium businesses (SMBs) that can't afford basic IT fragility, costs, and headaches, Google has figured out what, since 1999, has been the Holy Grail for every company (not just IT) looking for an in with SMBs -- how to hook them online. Think about it. To the extent that FaceBook, MySpace, AIM (AOL Instant Messenger) have been outrageously successful at attracting a specific audience and making them stick, who has really had that kind of success with the SMB market? NetSuite, with its blend of CRM and Intuit like bookkeeping functionality probably rises to the top. Perhaps eBay as the electronic storefront of choice for many. But no single provider is the FaceBook, MySpace or AIM of SMBs. Meanwhile the SMB market in aggregate is widely recognized as being the most lucrative to those who can crack it. So far, despite thousands of attempts (a good many of which I was involved with back in the 1999-2000 timeframe), no one has.

This isn't a story about Google Apps because to the extent that Google finds a way to crack the SMB market the way no one else has before, Google Apps won't be the only offering of its kind from a provider that's in Google's league. The SMB market is simply too important to too many of Google's competitors for Google to be allowed to run away with the show.

So, bearing in mind that this is as much as a story about the future competitors to Google Apps as it is a story about Google Apps itself, consider the following.

As seems to happen every couple of weeks (or sooner if some person e-mails me a message with a 5 MB attachment), I get an automated message from our Microsoft Exchange server that tells me that my inbox is over its storage limit of 150MB. Given how close my inbox always is to that threshold, I just got one of those "purge your inbox or die!" messages last week. At first, the message is like a nudge. But if I ignore it long enough, the server simply goes into action and rejects all outgoing mail. It tells me it's doing this, but it's nice enough to let the inbound mail, including spam, flow in.

After not touching on the topic in at least a year, last week, I had a conversation that had to do with migrating from IBM Lotus Notes to Microsoft Exchange (strangely, in the last five years, I've never heard of anybody going the other way). That conversation morphed rather quickly into a pros/cons discussion of outsourcing e-mail.

Back in February of last year, my fellow blogger Phil Wainewright asked "What's the true cost of running e-mail in-house?" The truth is, I'm not sure anybody knows the answer to that question because so much of that cost is fuzzy. For example, whereas storage architecture is the life-blood of outfits like Google (not just for e-mail, but for search, YouTube, etc.), Amazon (with its S3 service), Yahoo and Microsoft and their scale affords them certain opportunities to cost effectively increase overall capacity, storage is a checklist item for most non-IT companies (I said "most." Not all.).

The typical thought pattern goes something like, "how much storage do we need to bring that application online?" A decision gets made, product is purchased, the application is set up and until its users get nearly mutinous about its limitations or something catastrophic happens, the IT staff attends to other tasks. Even with storage coming down so dramatically in cost per gigabyte the way it is, the idea of doubling that server's storage capacity is still painful enough to someone's budget that it's simply not an option until things really start to break.

Meanwhile, in comparing notes with other people who rely on some sort of corporate e-mail system, pretty much every person I talk to has a limitation on their inbox size. For some, it's less than the 150MB limit. For others, it's more. But one thing is for certain: now that e-mails are just as likely to have a multi-megabyte movie or PowerPoint attached to them as they are 10 words of text, everybody groans when it comes to the limitations on their inboxes. Yet somehow, for absolutely no cost, Google can provide you with the same inbox and e-mail address ( with a limit of 2GB. Pay $50 per user per year, and then, the ceiling for everyone in your company goes up to 25GB. For me who has been struggling with the 150MB limit for years now, that would be more than 177 times the storage I'm allowed now.

Going back to the fuzzy costs of e-mail, on the basis of storage alone, I'd posit that Google's charge of $50 per user per year is probably enough to cause any SMB or CFO in larger organizations to re-examine their current e-mail choice.

For example, when I hit the limit, which is about once per month, I spend around 30 minutes digging around my e-mail looking for stuff to delete. Junk mail is an obvious one but a block-delete there is a bad idea given how many legitimate mails are getting routed there. My sent mail folder is another no brainer, especially if it has a few outbound emails in it that have big attachments connected to them. Then, I dig through my inbox (sorting on the basis of attachment size) and where there are e-mails with really big attachments that I need to keep, I save those attachments to my local hard drive (a re-allocation that I wish my e-mail could easily keep track of) and delete the mail. Finally, I purge my Deleted Mail folder.

Let's say I spend 6 hours of my time on this exercise every year. I can guarantee not just my employer, but most employers, that the 6 hours of time their employees may be spending annually on this futile exercise is worth more than the $50 you'd pay for each one of them so this wouldn't be a problem. Or, let's try the math another way. There are some number of people in your company where there's no cost to keeping their inboxes clear. Either their hourly wages crash the break even analysis or their inboxes never come remotely close to reaching their limits. But, on the other hand, there is some number of highly paid people in your company who are worth $300 per hour. Over the course of the year, the 6 hours they spend purging their e-mail costs $1800. That's 36 users of Google Apps Premium.

Of course, we're still talking about one extremely fuzzy cost. Never mind all the others that would take us a while to list before we finally came around to the hard dollar costs. Last year, after some local weather-related event, the e-mail server on which my account resided became inaccessible. Several single points of failure between my PC and the server had failed. The local area network was down. Power was out. The virtual private network was down. Naturally, having once been a Lotus Notes user and having been able to recover from an IT disaster like that by simply dialing into another location and attaching to a different Notes server to which my e-mail was routinely replicated, I thought I'd be able to do the same thing between my Outlook client and some other Exchange server at another geographical location that was still accessible. But this wasn't the case.

So, I inquired with our IT people as to why things were the way they were. The answer was cost. In fact, it doesn't matter whether its Notes or Exchange (even though it seems as though Notes' replication architecture makes it better suited to this sort of fault tolerance). The expense (servers, software, networking, etc.) of having such redundancy was so outrageously prohibitive that basically, someone made the decision that the staff would have to grin and bear it in the event that the one server that hosted each person's mailbox went down. One issue, according to the IT people I spoke with, was scalability. There comes a threshold (in terms of number of users) where you simply have no choice but to add another e-mail server to your server room or data center.

But with Google Apps, adding users (starting with the first one) and giving them 25GB of personal storage never requires an additional server and only costs $50 per user per year. Alternatively (and completely forgetting any of the hardware costs for a minute), once you've navigated Microsoft's complicated labyrinth of server software, client software, client access licenses (CAL) , and software assurance, $50 per user per year for an e-mail and group calendaring system that requires minimal oversight from someone in the IT department (again, completely forgetting about the hardware), seems like a pretty good deal (not to mention how differently it hits the books because its a service).

For example, assuming you're an SMB using a Microsoft Volume License (which you would for 5 or more users) and you're getting one of those standard 30 percent discounts from some reseller of Exchange, you could expect to spend around $660 on a copy of Exchange Server with the kind of upgrade protection (what Microsoft calls Software Assurance) that ensures you'll be able to upgrade when a new version comes out (at no additional cost). The cost of Software Assurance (approximately 35 percent of the acquisition cost of your software) recurs once every two years. With that same 30 percent discount, for each CAL required to access that Exchange server (CALs are required regardless of client software used), you pay a one time perpetual licensing fee of $47 ($67 list price x .70) plus about $8 every year for Software Assurance (SA). In other words, $63 initially and then $8 every year after that (fyi: Microsoft only sells SA in two year chunks so it'd be $16 for 2 years of SA). According to the licensing specialist I spoke to at 800-426-9400, the reason a CAL needs SA is because CALs are not transferrable when you upgrade your server. Upgrades to servers require upgrades to CALs as well. SA assures you that you won't have to pay the full boat for new CALs.

Or, for the server side of e-mail and group calendaring (Microsoft has completely separate charges for client software like Outlook), you can pay Google a straight fee of $50 per user per year, never have to worry about hardware, backing up, etc. and, oh, by the way, when Google upgrades its e-mail service which seems to happen pretty often (in other words, you don't have to wait for major releases to get some cool functionality upgrade like what was added as a result of the Postini acquisition), the IT people barely have to lift a finger (oops, there are those fuzzy costs again!). All the end-users have to do in their browsers is press the refresh button. By the way, for organizations paying the $50 per user per year charge, Google promises 99.9 percent uptime.

Guess what? None of this analysis is meant to paint Exchange Server in a negative light. For some organizations, Exchange Server along with some of the other products that Microsoft is just now coming to market with in the area of unified communications are great solutions that hang together really well and for some organizations will provide terrific ROI. What I'm trying to point out is that when you factor in all the hard and soft costs of running an e-mail system -- the servers, the storage, the networking, the fault-tolerance, the accessibility, the software, the upgrades, the people, the spam, etc. -- $50 per user per year ain't such a bad deal to get the sort of e-mail system that Google Apps gets you.

Enter: All the people who poop on Google Apps because, relative to the cost of Microsoft Office, paying $50 per user per year isn't such a great deal (they have their break-even analyses and, to give them the benefit of the doubt, let's assume they're fair). OK, fair enough. If the only reason to look at Google Apps is for its ability to do the sort of stuff you'd do with Microsoft Office, then maybe Google Apps is no clear winner. But then again, although I'm sure they exist, I can't imagine many companies becoming clients of Google Apps without taking advantage of the e-mail functionality which, as I just proved, is probably worth more than $50 per user per year when you consider what it costs to run your own solution. In other words, all the Office-like functionality is simply icing on the cake. Not to mention the collaborative abilities built into it, the Start Page that all users get (the organizational equivalent of iGoogle), and the free Web hosting.

Finally, as a Google Apps user, I'll be the first one to admit that you must accept some compromises when you move into the so-called "cloud." Yes, your applications may not be accessible while you're disconnected from the Internet (that problem is temporary thanks to Google Gears). And yes, you're trusting Google with what could very well be incredibly sensitive information (the Googlefolk occasionally remind me that if they EVER screw that part up, it's game over for them... in other words, they take that trust very seriously). And, the apps aren't nearly as robust as their desktop competitors. I was reminded of this yesterday when a document I created in Google Docs neither translated the way I expected it to when I saved it as a PDF nor did it look very good when I cut and pasted its HTML into a regular Web page (it took a bit of hand editing, but I finally got it fixed). Google Apps is most definitely a work in progress and not nearly as mature as MS Office.

But all this said, there are a lot of organizations that might be willing to accept these sacrifices once they consider what they might be getting for their $50 per user per year and whether they really need anything more.

Finally, a reminder that this really isn't about Google Apps. It's about the model of Google Apps that will invariably be repeated by others -- a model where just one part of the offering (in Google Apps case, it's e-mail) makes for such a compelling ROI story that it makes it a lot easier to overlook the shortcomings of the rest of the package (provided there are any such shortcomings in the context of your organization's needs).

Topics: Apps, Collaboration, Google, Hardware, Microsoft, Servers, Software, Storage

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • As usual, great analysis!! And, considering you can have it for free if you

    only need 2 Gig per user!! That is still a lot more than most corporate systems give you, and then you can forget about all of the costs of running an email server. If you are a small company, that might even mean you need one less person on the payroll since it is so simple to administer Google email for your domain.
  • Just one nit to pick

    For a truly small business (less than 50 users) your analysis of the costs of Exchange and Software Assurance don't make much sense. Microsoft Small Business server is a pretty compelling sell. It includes Exchange, Sharepoint, AND Outlook licenses, and the premium edition also includes SQL Server 2005, and ISA 2004. It costs $600 retail for the standard edition and $1300 for the premium edition with 5 CALs. Additional CALs are about $1900 for a 20 pack which covers a license for each of the products. Everyone, especially small businesses, know that Software Assurance only makes sense if a new version is imminent, so few purchase it. So for 25 users, you have one time software cost of $2500($600 + $1900) or $100 per user. The last upgrade to Small Business Server was in 2003. An organization that purchased SBS 2003 in 2003 will have had no recurring software costs since that time. If the same organization used Google Apps for 5 years, the software costs would have been $1250 per year or $6250. That doesn't include the cost of potential yearly increases in price of the service.

    For a small business, I would also submit that quota size isn't that big an issue because you can grant each of your 25 users a 2GB quota and only use up about 50GB of disk. Not really a big deal at all in the era of 1TB desktop hard drives.

    So you have to weigh if the other benefits of the service model that you listed (not worrying about backup, uptime, etc.) outweigh the burden of maintaining the Small Business 2003 server.
    • But, what about the cost of the hardware, maintenance, administration,

      redundancy, the cost to move to another server when you run out of disk space, . . . .

      For the 2GB per user scenario you give, Google is FREE!!!
      • I thought I covered that

        when I said, "So you have to weigh if the other benefits of the service model that you listed (not worrying about backup, uptime, etc.) outweigh the burden of maintaining the Small Business 2003 server."

        I agree that there are significant maintenance/administrative headaches with running your own server. I just wanted to point out that some of the costs and scenarios that David described (Software Assurance, small quota size, software license costs) did not really apply to small businesses. The quota issue is a problem for medium/large organizations, not small ones.
        • Ok, I just think all of those other costs add up really fast, and, managers

          then have to spend part of their time hiring and managing email server administrators and other things that distract them from their real business. So, for anybody under 100 employees, I would definitely recommend hosted solutions, if not Google, some other provider. If you go with the free version of Google Apps, MyDomain, you pay only about $10 per year for the domain name. That gets rid of a whole lot of headaches and costs.
        • Also, you may have mentioned the other costs, but did not use them

          in the price comparison. You should have at LEAST estimated them for the price comparison. You also should have compared your solution to the free Google solution which has the same storage capacity as the solution you mentioned.
    • um, what about maintenance?

      Your costs are 100% acquisition. What are the costs to install it? What are the maintenance costs? What about the cost of servers (with maintenance contracts) to run your software on? How many people will you have to hire to keep the system running? How much will they cost?

      There's a lot more cost than just acquisition. With Google Apps, you can run it without any of those costs (OK you may need a part-time techie to set it up and add new accounts, etc). That's where the real savings comes in.
      • Maintenance?

        While I agree that maintenance is an important cost to consider, there is also an often overlooked benefit of Google Apps. That is the API that would allow you to tie it in with your HRMS apps or other systems.

        So what I'm saying, is that if the integration is done right, you wouldn't even need an admin fellow to do the adding/deleting of users. Just the one time investment for making a change to your HRMS system.
    • Software costs are the tip of the iceberg.

      As far as I can see, the main costs of IT have been largely ignored. So what if for a
      mere few thousand dollars I can buy some MS Software. Unless I've also paid for
      some hardware, to run it on, and more importantly a skilled (OK probably not that
      skilled - and therein lies another problem) IT technician to configure it, keep it
      running, secure, back it up (oh did I mention you'll need more hardware for that),
      fix it when it dies... the MS Exchange CD's will sit on the shelf. $50 dollars per
      user per year for a 99.9% available solution is a bargain. I'd already done the math
      and it was clear to me that paying $30 dollars per user <b>per month</b> to
      outsource email for my SME was far preferable than hosting email in house.
      Nick Maxwell
    • What's keeping me from switching

      So I actually run the IT infrastructure for a small 501(3)c with about 15 users in my spare time. That charity currently runs MS Small Business Server 2003. What Google Apps is missing is the ability to create web applications. From my understanding, they currently only allow for the creation of static html pages for websites hosted by them. If Google were to add the ability to create dynamic web pages using PHP, Java Servlet/JSP/JSF, or ASP.NET along with MySQL or some other database support I would definitely switch. But as it stands, I need to run my own infrastructure for hosting my web applications anyway. It doesn't cost any more money to host Exchange and Sharepoint on the same box.

      And I think some of the maintenance costs are being overblown. I spend about 1 hour a month applying the monthly security patches, and no more than 30 minutes a week doing offsite backup activities. We haven't spent a dime on hardware or software since 2003. In terms of the initial setup that I did in 2003, that took about 40 hours. The initial hardware costs were about $1700.
      • One correction...

        While you're busy touting your SBS prowess, lets just fix 501(3)c to 501(c)3...
  • $50/yr justified

    I think you've fairly rationalized the $50/yr cost of Google Apps.
    It would be good if you pilot it and write about 'Life in the Cloud' for a period of time to determine its viability.

    You can store and share in GA 'sensitive' documents in a secure manner using encryption, e.g., OpenPGP.

    Fair and balanced.
    D T Schmitz
  • RE: Does payback on email alone make Google Apps' $50/yr. worth it? (Docs,

    Excellent summary and analysis. Especially the part about fuzzy costs.

    My new startup (we're 10 people right now) chose GApps for our email and collaboration platform instead of a MS Small Business Server on a Dell server. I used to work as a VAR installing Exchange and other email platforms, so our decision wasn't based on lack of technical expertise. It was based on "all in costs" and the strong belief that the current state of IT is poised for a big shift from local apps to cloud based services. SMBs and mid-sized organizations are starting to see the overall economic benefit to IT services for their "essential" infrastructure needs.

    I simply love the idea of everything hosted and a "frictionless" IT infrastructure that is affordable, reliable and secure.

    It's 2007 - why is anyone managing their own email server anymore? There are better ways to use our collective IT budgets.
  • lazy way of doing things

    The whole computing cloud thing really reminds me of manufacturing in America in the 1960-70s. Why worry about this or that.... just subcontract it away and don't worry about it.

    Same thing with the "computing cloud". Sure you don't have to worry about your email server or whatever, but what happens when Google gets bought out by China's state ISP provider?

    The Cloud is an entity just like a DVD player or a plasma tv. It exists on real servers and real routers and real networks. To advocate moving everything onto the 'cloud' is foolishness driven by a shortsighted ROI.

    So you tell me. What happens when some entity takes ownership of the cloud after we've structured everything around cloud-based computing? Total stupidity.
    • But, should you also build your own cars? Your own microprocessors?

      Where do you draw the line? Just spouting off about the chance of one provider possibly being bought by a Chinese company does NOT mean the whole industry is a bunch of untrustworthy crooks.

      And, if the Martians take over the Internet as you are theorizing, locally hosted email will not work either.

      Do you keep your companies money under your mattress at home, or do you trust banks?
  • Then you are on a G-leash

    Based upon standard net service provisions they are free to change conditions or cancel the service at any time. Years ago this kind of service made sense but now, with cheap server appliances and open source software it is all rather pointless. Lack of control and business confidentiality requirements makes it a hard sell. The SMB market is localised service, support and administration, all with a personalised touch. People dealing with people, not hype and marketing illusions. I worked in small business and tech corporations really just don't get it. Google crack the SMB market, no way, but they might be able to convince their investors they can, at least for a time.
    • Already on Microsoft's leash. Why not Google?

      That's exactly what you got when you use Windows. They can even cripple Windows if they feel they need to. It's all pathetic.

      [i]Based upon standard net service provisions they are free to change conditions or cancel the service at any time. [/i]
      • Bottom line: it is somebody's leash

        Please recall in the original post that I said this story is as much NOT about Google as it is about Google. It really is about a model and no matter what model you pick, you're on someone's leash.

        The question is which leash to pick and what is your rationale?

        • Exactly.... Agreed.... That is correct.

          I would think that anyone willing to put up with the crap that Microsoft dishes out would be more than happy with Google. Enough said.

          [i]The question is which leash to pick and what is your rationale?[/i]
          • Well, to be fair, you could use a complete open source solution, and then

            you are not on anybodies leash. But, then you have to be a guru, or you are on some guru's leash!!