The Numbers Are In
-
I'm not sure how many of you have had to work on budgets or calculations for the cost of software, but if you have then you know it's not much fun. I know that when trying to figure out what licenses you need for a particular piece of software and then crunching numbers for upgrades, it's a full-time job for a few days or more, depending on the size of your environment.
I had to do a little of this back in the early days of SQL Server 2000 and at the time Software Assurance (SA) was available for the platform. I had to try and crunch the numbers for the upgrade alone v including software assurance and I think I decided against it. Of course I had no idea that SQL Server 2005 would take 5 years, but I thought that it was probably more than 2 years. I'm not sure what the timeframe for SA was, but it seemed too short to me.
A few people sent me this article on users dumping software assurance and it shows what I've thought for a long time. Others haven't seen value in Software Assurance either.
Trying to keep customers happy in between upgrades it a tough battle. Heck, trying to sell upgrades is also a tough battle. I would love to know what percentage of users want upgrades, with new features, and what percentage wish they could continue running the old software.
Most of us that have written software know it's extremely hard to work on a timetable and predict when you'll be done with a project. Once it's any appreciable size to require more than 2 or 3 people, it's just too hard to guess when it will be done. I know software companies want to generate revenue on a steady scale, and I think Software Assurance could work if we were ASSURED to get something. The R2 releases help, but realistically, I'd like to see an "R" release every year. And regular support offerings that fix things that are broken. If I'm going to pay for your yearly staff, then deliver me something every year, not just when you're ready for a big conference and release event.
And don't have "launch events" if the product isn't released. I'm surprised I haven't seen many comments on that editorial.
PS - Support Katie and her family : Purchase a raffle ticket today for a chance to win some great prizes.
PPS: Apologies to our members and other authors. The article published yesterday contained plagiarized content and has been removed. I appreciate a reader sending me a note (along with research) and I am sorry that I did not catch this sooner.
-
Budgets are easy (lol), power and heat capacity planning for hosting centres, now thats hard
Our system grows by around 35GB each week (but with a significant upward trent), and we need new NLB web servers and supporting servers every few months, try working out the requirements 2 years down the road and you too will go into meltdown.
I'm not complaining by the way, i love this stuff
-
Our company has taken a strong position to stay on 'supported' releases of software, which puts us in a never-ending cycle of upgrades and, in the case of Oracle, patches for security vulnerabilities. (Oracle only produces their security patches for supported versions.) So the issue to buy or not to buy SA, has been decided. We buy it and we upgrade. We are caught between wanting the new features and not having time to use them because we are always upgrading!
-
Our problem is two fold. We have alot of bundled software packages on SQL Server. About 70% of them do not support SQL Server 2005 yet so we are stuck. The other problem is that we are in a budget crunch so getting money approved for an upgrade knowing that SQL Server 2008 is just around the corner is a tough sell.
-
Software assurance is something I want to like. Ideally you fight that decision through the budgeting process once and from then on you can upgrade as you want to, not have to go back and give the bean counters 10 good reasons to upgrade to SQL 2019. Arguably it's good for the business as they no longer have the direct financial disincentive to apply the latest technologies; it's just a matter of when and where to apply them. It's good on the people side because we know that we won't be stuck on SQL 7 until we retire, support from MS or not, and so we can stay employable.
But, alas, unlike magazine subscriptions where you know you'll get something of value in the mail each month, with SA you might get something or might not. Doesn't sound like a business friendly policy does it? I suppose extending the term to 5 years might help, or even 10 years, but if it any kind of gamble why would a business do it? Easier to just pay as you go and like most companies, skip a version to keep the costs manageable.
MS wants the steady cash flow, we want access to all the new releases. Anyone have a plan that really satisfies that? My try; change SA to reduce the cost of licenses by 30%, charge us 10% of the initial price per year after the first year to maintain access to all new versions. After 3 years they've collected about the same amount of cash as just a raw sale, but they will probably keep cashing that 10% check for a few years. No, my idea probably not any better than theirs!
Or...they could just include service packs as part of SA - then we'd be getting value (if we didnt move to MySQL).
-
Software Assurance is pretty lame. The only reason to have it for some products is, as one poster previously noted, so that you don't have to go back to the bean counters to get money released to get to the next version. If your bean counters are enlightened and forward thinking, you can easily dispense with SA.
Commonly software maintenance for an application is 15-18% per year and this usually includes free help desk calls. SA is 25-33% per year and help desk calls are not always free. If Microsoft were charging 15-18% no one would get upset and virtually everyone in business would be participating.
Here are some pretty good guidelines on software assurance:
Desktop/Laptop OS: NEVER buy SA; buy the OS w/ hardware and run it into the ground for three-four years, replacing both at the same time on a budgeted 1/3 or 1/4 replacement plan.
Server OS: Never buy SA; don't buy license w/ hardware. Buy a license and then keep on transferring it from production to development to lower end uses from server to server; you should be able to get 5-7 years out of a license before having to trash it due to age and non-support
Exchange, SMS, ISA: Keep these under SA; generally you want to keep your email and other infrastructure systems moving forward every 2-3 years
SQL Server, Sharepoint: Same as Server OS above
CALS: Keep these under SA; If you run a CAL model (as opposed to per processor), buy these in bundles of products (Server, SQL, Exchange, Sharepoint) and keep them under software assurance; need to determine whether /machine or /user or mix is best for you; by keeping them under SA you dramatically reduce the administrative overhead
-
Regarding my previous posting, here are a few more:
Office products: Same as desktop/laptop OS
Programming Products: Buy MSDN subscriptions for your IT staff
-
drnetwork,
I kind of like that model and it makes some sense. Personally I want to like SA as well. But it needs to have a tangible return, not some "iffy, maybe you'll get an upgrade IF we release one". Release every 2 years, regardless of features that are finished. The marketing folks will work around whatever the technical guys deliver.
-
I would concur with drnetworks comments.
We are a small shop (<50 users) so volume licensing is not a possibility or financially realistic.
Another topic to go with this is the purchase of licenses and where? Do you use M$, a M$ reseller or some other third party vendor?
Earlier in a current project, I also learned that my W2K3 server licenses were down gradeable to W2K. This was important to learn as I have to have a mixed mode network for awhile before I could bring those servers online. I did not want to purchase two sets of licenses. The W2K3 licenses I purchased (including TS Cals) were less expensive then W2K licenses. Once you are ready to upgrade, you call India, er, M$ and upgrade your license back to W2K3.
I learned all this from talking to M$ directly. An apparently undocumented feature that costs nothing except the nearly 4 hours of phone time for the whole process and the need to compress my English into high speed verbiage to keep up with the offshore support.
Viewing 9 posts - 1 through 8 (of 8 total)
You must be logged in to reply to this topic. Login to reply