Tuesday, October 25, 2011

Dogfooding IT with Office 365

Have you heard the term, "eating your own dog food?"  Microsoft manager Paul Maritz wrote an e-mail in 1988 that created the term and loosely it means... use what you sell.

I took the challenge to heart a while back and decided that if I was going to sell cloud services I needed to use them myself.  This idea really took off for me with Office 365 this year.  As I started talking to more and more clients about the benefits of cloud-based e-mail I first signed up for Business Productivity Online Services (BPOS, the precursor to Office 365) and then the beta of Office 365.  I used BPOS and the beta as test beds to learn the products and to assess their viability for my clients.  I wasn't ready to migrate to them internally yet however.

Eventually, just as Office 365 was releasing my first client wanted to get moving.  I had been talking to them about it for months.  My experience with the beta had been good so we went ahead.  I signed up for the full product and started moving over e-mail at the same time I was installing it for this client and actively selling the product to my other clients.  I figured it was time to eat my own dog food.

So, here's what happened: my client had some growing pains but Office 365 has turned out to be a good solution for them; I've had several other clients express interest and installed it for them successfully; I love using it internally and am recommending it to anyone that listens.  I did learn some lessons though:
  1. Doing your research really pays off.  It did for me!  I avoided several land mines by doing a beta assessment first.
  2. The entry-level Office 365 SKU (P1) doesn't come with phone support.  It isn't worth the $4/mo savings... get the E1 SKU, you won't regret it.  There's also a 50-user cap for the P1 SKU.
  3. You cannot switch from the P1 to E1 SKU.  Tell me it ain't so!  This was a real bummer for me.  I still don't understand why.  I'm sure there's a good technical reason.  You have to do a full migration to move between these SKUs.
  4. I know Office 365 is supposed to be easier than self-hosted Exchange, but it's no cakewalk and still requires significant technical knowledge and capability.  Your average tech-savvy client isn't going to do a migration without assistance.  Some of the migration stuff gets pretty complex.  Besides, most businesses just don't *want* to do the administration.  They have other stuff to worry about... like running their business and making money.
  5. The amount of money you'll make on Office 365 (aside from ancillary services) is negligible until you get in to hundreds of  deployed licenses.  Don't plan to make your money on monthly recurring revenue.  It's a 6% share per year plus 12% if you're the initial partner of record that signs up the client.  That's a maximum of 18% the first year plus 6% per every year after.  Make your money on services, that's all I can say about it.  It's not nearly as profitable as providing a hosting service if you do that.  If you do provide hosting services plan on this being a big competitor!  Stress the customization options you offer that Office 365 does not.
  6. A properly run and well maintained Exchange server doesn't take much more work to service than Office 365.  You'll still get the bulk of your work on user adds/deletes, adding new domain names and aliases and on the user-side support of Outlook.  You just don't need to monitor the hardware any more.  Oh, and you have a new administrative interface to learn.
  7. Including Microsoft Office in your Office 365 subscription seems like a good idea and it may be for some clients that have very seasonal workforces.  Do the math before you decide on this course though.  You have other options for subscription based Microsoft Office, like Open Value Subscription and sometimes Services Provider License Agreement (SPLA).
  8. There's no private branding or resale of Office 365.  It's 100% direct bill from Microsoft.  You don't get to mark it up.  This makes it hard to include in a fully managed service agreement by the way.  Microsoft really needs to fix this to increase adoption by Managed Service Providers (MSPs).
  9. Just like any service IT WILL GO DOWN.  Make sure your clients understand that 99.9% of the possible 8,760 hours in a year equals about 9 hours per year of down time.  That's a full day and then some and it could happen at any time.  Just the same, it's likely to be much more reliable and secure than self-hosted Exchange for most small businesses and may make sense for larger ones as well.
  10. It comes with some nifty additional products you may or may not use.  Bear in mind that you don't get as much of each product with each Office 365 SKU so learn your product before you sell it.  Also, don't try to sell Lync to people using Office Communication Server as an onsite VoIP unified communications system.  There's no integration with the public communications infrastructure yet.  That means for right now it's an internal-only system.  Cool, but not blow-me-outta-the-water awesome.
I like Office 365.  I've tried Google Apps and while that worked okay, Office 365 just feels more like a business solution to me.  It's been reliable and pretty easy to set up.  Most days I don't even notice I'm not hosting my own Exchange server any more.  And isn't that the idea?  Apparently, dog food ain't so bad!

Are there any best practices with regard to Office 365 that you've found and I've missed?  How many seats do you have on Office 365 and how has it been as a profit center for your business?  Let us know with your comments!


  1. Dear Scott,

    you might want to validate your statement on the SLA. It is measuered on a monthly base which does limit the maximum downtime window significantly. Also SLA measurement is much more of a mystery to most clients. That surely is something to dwell on. I mean they neither do understand their current SLA if they have any at all but do not spend time to really understand the service provider one either.

    Here is a SLA comparison between Google Apps and O365:

  2. I like the comparison in your blog. I'll leave it to you to go in to depth on the SLA. I had not really looked at the Google SLA before reading your article. It's interesting that Google doesn't publish anything with regards to spam and viruses. Also, with Microsoft's fondness for lengthy documentation I'm not surprised that they do. Let Google go through a couple antitrust suits and that may change though.

    My discussion here is merely meant as a warning that even when you have a highly reliable and redundant hosting environment provided by one of the top cloud services vendors in the world that downtime is likely to happen at some point.

    If people go in to cloud services with their eyes open and a realistic view of what the service level could be I think that they'll be more likely to be successful with cloud services in general.

  3. Great insite to the whole Office 365 offering. One thing I would like to add is Internet bandwidth. Make sure the client has a solid Internet connection and adequate bandwidth. An office with 10 user will not be happy with a 3x1Mb connection if they are storing documents in sharepoint.

  4. I couldn't agree more. If you're planning on moving critical infrastructure to the cloud it's incredibly important to make sure you have a highly reliable link to the internet.

    Offices with DSL, cable and such will certainly want to look in to a T1 or better. Larger businesses will want to consider redundant links from multiple providers.

    While bandwidth is important the reliability is much more important. While it is annoying, most users can handle documents taking 30 seconds to upload. Having your e-mail service unavailable because you have to reboot your cable modem every day is unacceptable though. Yes, you pay more for a T1, but you get a service level agreement (SLA) that guarantees a certain amount of reliability. If I'm going to run a business that depends on the availability of the cloud I'm going to want all the guarantees I can get!

  5. My hardest deal as a MS Partner for 365 is being profitable on a migration. Even if I charge say $100 per user for a migration, you can lose BIG if it takes 8 hours for 1 PC. I have experience cases where it flat out will not install SSO and Lync on a PC w/o reinstalling Windows XP (and this is what latest SPs, btw). Also, drawing the line with your customers between your 365 support and general PC support gets difficult after you 'touch' their device for your migration. Thoughts from anyone on this?

  6. When I do migrations I charge hourly for the service work up front precisely because of the difficulties you mention. Especially for clients that I don't have managed service agreements with, new clients, and where the workstations are in bad shape.

    Here are a couple tricks:

    1. Distribute a minimum standards and responsibilities document with your proposal. It should talk about O/S service packs and patching, Office version (2003 won't work) and patching. It will spell out what is your responsibility to fix and provide and what is not included.

    2. Don't give a hard cost for migrations but rather an estimate based on the average historical cost per user. Be sure to include everything... labor, hardware upgrades (memory to handle upgraded Office 2010), software upgrade costs, etc.

    3. Be realistic with your clients up front. Let them know that every migration is different and that the technology is new and has some kinks. If they are willing to accept some inconvenience the rewards are exceptional!

    4. The cutoff between migration support and general support is muddy. It always is. This is where good project management comes in to play. Plan out the length of the project before hand and include the number of post-implementation support hours that you'll be providing. Make sure they understand that there will be a change request that will alter the price if they require more time.

    This is Information Technology, not plumbing. There is as much intuition and art to expert IT services as there is logical thought process. You don't just "buy" Office 365 or anything else and have it go in without issues. Every system is unique and your clients should understand that getting past that is a shared cost, not yours alone.

  7. Great information, Scott. My understanding was that the P1 plan was month-to-month, while the E plans require a one-year commitment. This would remove some of the advantage for using the E3 & E4 plans to get short-term Office 2010 licenses for temp employees, interns and seasonal staff. Do you know more about this? I've been unable to find a clear answer.

    Thanks, Armen

  8. The P1 plan is month to month. The E plans are all one year subscriptions but may be paid month to month. I found the following answer online:
    (i) One month (“month-to-month”) Subscription. A one month Subscription may be cancelled anytime without any fee.

    (ii) One year Subscription (including prepaid). If you cancel a one year Subscription within 30 days of the date on which the Subscription is ordered or renewed, you must pay for the initial 30 days of the Subscription. No payments will be due for the remainder of the Subscription. If you cancel a Subscription at any other time during the Term, you must pay 25% of the Subscription fee otherwise due for the remainder of the one year Term.

    It's kind of like a cell phone plan in that you have a term (one year with Enterprise) but you pay month to month. You get 30 days after the subscription is ordered to cancel with only owing the first 30 days, or else you are liable for the 25% of the remaining yearly costs. The 30-day trial doesn't obligate you into purchasing anything.

    Thank you,

    Todd Douglass - MSFT Support Engineer
    See it yourself at

  9. Thanks Todd. So the benefit we sometimes hear about using Office 365 to hire temps and summer interns isn't quite as good as simply turning it on and off. The worst case scenario would be if you need a subscription for say 8 weeks in the summer. You would end up paying for the normal first two months, then an additional 2.5 months (25% of the remaining 10 months) for the cancellation, more than doubling the monthly cost for that 8-week time.

    It's unfortunate that the E plans aren't offered month to month, as this cost flexibility would be a nice selling point.



Due to excessive spam, only registered users may post comments. Comments are unmoderated and post immediately but they are monitored. Inappropriate content will be removed promptly and will get you banned.

If you wish to communicate with me outside of this blog please e-mail me at

Related Posts Plugin for WordPress, Blogger...