Continuing a theme from the previous post, I want to briefly discuss the notion of hard and soft costs in a business case or Total Cost of Ownership (TCO) study.
Hard costs are those costs associated with hardware, software, maintenance, operational expenses, capital costs, and so on. They are specific and identifiable costs to the business. Soft costs, on the other hand, are not so easy. They are often highly variable, based on intangibles, or things that are very difficult to quantify. They include things like the cost of downtime, the value of data, the benefit of higher availability, and sometimes, people and labor costs.
(Although, as an aside, I would note that people/labor costs can sometimes be soft costs, and sometimes be hard costs. If we are talking about the cost of personnel to administrate a backup system, this is a hard cost. If we are talking about lost productivity due to system/data unavailability, this is a soft cost.)
Generally speaking, most businesses will accept the distinction, and will accept hard costs and savings in a business case. It is pretty tough to argue that a tape library with tape drives carries a real and quantifiable cost: a cost of acquisition, and a cost of maintenance.
On the other hand, some businesses will accept soft costs, but many will not. Do you really save $1,200,000 if your mission critical server can be restored 4 hours faster than with a competitive solution? Do you really lose $10,000,000 if you lose (and cannot restore) 200 MB of data? Maybe. It probably depends on your industry. A financial institution may lose more, and have a better understanding of the magnitude of those losses, than some other businesses. Other businesses may not lose anything except people time. And, for better or worse, people time may not be valued with a specific dollar amount. So, for many, a business case based (solely) on soft costs is a much tougher sell.
Therefore, when we build a business case, we strongly prefer to use hard costs to justify the decision, and explain the Return On Investment (ROI).
Again, lets borrow from the prior post. First we have an example of hard costs:
Tape Backup | Double Take | Avamar | |
Hardware, Software, Setup | $120,000 | $31,750 | $800 |
Maintenance | $18,000 | $6,314 | $200 |
Media | $36,000 | $0 | $0 |
Salary | $150,000 | $15,000 | $7,500 |
Off-site pickup and storage | $36,000 | $0 | $0 |
Year 1 Total | $360,000 | $52,884 | $8,500 |
Everything here is a hard cost. Everything has a specific, quantifiable, identifiable cost to the business. And as a result, when I conclude that the cost of an Avamar solution is 83% less than the Double Take and more than 97% less than the cost of tape I have a very powerful argument that Avamar should be strongly preferred to the other two solutions. Put another way: this is an argument I can take to the bean counters, and they have very few ways that they can dispute it (other than making sure that the numbers we have presented are accurate).
Now, how about soft costs? The last table I used is a fine example of soft costs:
Tape Backup | Double Take | Avamar | |
Cost of unrecoverable data ($10,000/MB) | $200,000 | $694.44 | $200,000 |
Cost of downtime ($42,000/hr) | $336,000 | $1,680,000 | $168,000 |
Year 1 Total | $536,000 | $1,680,694.44 | $368,000 |
Both costs here are soft costs: how much does unrecoverable data cost? How loud is loud? What is the sound of one hand clapping? Likewise, how much does downtime cost me? You may know the answer to these questions for your business, but more likely you do not. I could tell you (and I sort of have, by assigning values to the categories!) but your answer is probably going to be more valid than any I supply. And even if you know the answer, it certainly doesn't mean that an accountant--or a manager or CFO--will accept it as part of a business case.
So when I say that Avamar costs approximately 40% less than tape, and 80% less than Double Take, that is interesting, but it is very likely not relevant to my business case, TCO, or ROI. Even though, on paper, I can show savings of $1,312,000 in soft costs, most businesses will be far more interested in the $44,384 of hard cost savings.
What is more--even if the soft cost savings were more realistic, this would still be the case. That is to say, assume the savings really are about $1.3m; but that instead of having to use a totally ridiculous number like $10,000/MB, I have more data, but a more realistic $/MB, for the same net savings. At least I am not insulting your intelligence now by telling you that you are carrying around $20,000,000 worth of data on that 2 GB USB key. Even with that said--even if I can justify my soft costs with sound logic and reasoning--most businesses will still be more interested in hard cost savings. And most will find it far easier to justify a decision on the basis of hard savings.
At the end of the day, the real question here is: do you know how much your data protection (backup and recovery) environment costs your business? Do you know the components that contribute to that cost, and how much each one of them costs? These are critical pieces of information as we seek to improve our backup environment--especially in these tough times. The worse the economy is (and hopefully it doesn't get much worse) the better we need to know our costs and have a sound business case in order to make change.
Do you know how much backup costs you?
Final thought: I don't think the first table is at all comprehensive when it comes to understanding the hard costs associated with backup and recovery. In a future post, I will attempt a more thorough review of what all the cost components of backup and recovery are.
I again say the numbers that this post is based on are completely bogus.
Posted by: W. Curtis Preston | July 08, 2009 at 05:53 PM
I'm trying to figure out these numbers...
Let's start with media: $36,000 buys 800 LTO-4 tapes with 640 TB of raw capacity. $36,000 for offsite storage at $1 per tape per month is 3,000 tapes. This is a huge environment.
Then we have Avamar backing up 40 GB. This is a tiny environment.
What am I missing here?
Posted by: Stephen Foskett | July 08, 2009 at 06:33 PM
Curtis;
Again--focus on the process and the reasoning, not the conclusion. And if you want something that isn't completely bogus (Mr. Diplomacy) then use $5k for the Avamar price. I won't attempt to defend the other two colummns as they aren't my numbers. The whole point of the two posts was to point out that they were suspect. So I guess we actually agree. Except that you are missing the forest for the trees if you think the post is worthless.
Posted by: Scott Waterhouse | July 09, 2009 at 07:38 AM
Stephen;
We are discussing 20 remote offices. I guess I could envision a rotation policy with daily off sites that only used 40 tapes per office. Remember--it doesn't matter if they are LTO1 or LTO4, they aren't going to get filled if they need to go off site. As I said to Curtis, the numbers are Double Take's and I wasn't trying to say their numbers were flawed, more that their process is flawed. So I am not terrible inclined to defend them.
The 40 GB comes from trying to extrapolate based on the cost of lost data how much data is at stake (assuming a .1% daily change rate). Since they don't really explain how much data is in their test case, this is the best I could come up with.
Posted by: Scott Waterhouse | July 09, 2009 at 07:42 AM
Scott,
I just don't buy the numbers you're talking about. EMC continues to push the Avamar barrow pretty hard, but how can anyone read those numbers without thinking they're made up? $800 for hard setup costs for Avamar? How can you justify that number?
Curtis may have been blunt, but I have to agree with him.
Please justify how $800 is all that's required in a comparative Avamar setup for hard costs.
Cheers,
Preston
Posted by: Preston de Guise | July 11, 2009 at 03:18 PM
Preston;
See comments in previous post for my clarification. If you are not willing to accept an incremental cost basis ($800) then I suggest that $5000 would suffice for a complete solution.
Posted by: Scott Waterhouse | July 13, 2009 at 09:21 AM
Can I assume then that you're doing comparative costs to bolting on each of those 3 systems to a central already configured/setup system to support remote office backup?
This isn't properly spelled out in your post.
If that is your intent, it's perhaps the only way that the numbers come any way close to being justifiable, and even then I still think they're way out. (In particular in that scenario I think your costs on tape setup for a remote office is unrealistic in comparison.)
I.e., the only way I see the Avamar cost as being justified is if it has already been purchased, setup and configured. Is this what your post is meant to be about? Extension of the backup system to support remote offices?
Cheers.
Posted by: Preston de Guise | July 13, 2009 at 02:33 PM
Preston;
The first two sets of numbers are from Double Take (for tape and their software). There is not much I can say in defense of them, as their white paper does a poor job of outlining the assumptions used to generate them.
As I commented in the previous post to Curtis Preston, it is possible to imagine a "new" Avamar system for $5k in some circumstances.
As for the tape being way out--maybe. I think the math comes to 40 tapes per site, with a robot, and a server at each site too. That is only $18,000 per year per site. That doesn't strike me as wildly unrealistic.
Posted by: Scott Waterhouse | July 14, 2009 at 08:18 AM
I think it is pretty clear in the footnotes the sources that were used to calculate the ROI. Whether you agree with them or not is certainly a valid debate since some of them are a few years old. Maybe it would be best to provide a link to the disputed charts in the white paper and let readers decide for themselves. Either way I think we all agree that there are more cost effective solutions than tape alone.
http://www.doubletake.com/english/resources/whitepapers/pages/default.aspx?ResourceID=16&SiteType=Global
1 See Double-Take whitepaper "Reducing Costs and Risks of Branch Office Data Protection"
2 Regan, Keith. "Concerns Raised on Tape Backup Methods." SearchSecurity.com 15 April 2004
3 Smith, David M. "The Cost of Lost Data." Graziadio Business Report Vol. 6 No. 3
4 Meta Group 2000 data
5 Smith, David M. "The Cost of Lost Data." Graziadio Business Report Vol. 6 No. 3
6 www.nwfusion.com/careers/2004/0105man.html
7 Zaffos, Stanley and Phillip R. Sargeant. "Designing to Restore from Disk: Backup Futures." www.gartner.com. 19 March 2001
8 Goldworm, Barb. "Alternatives to Tapes on Trucks." DM Review. 28 October 2005
Posted by: Brace Rennels | July 22, 2009 at 06:33 AM
Thanks for providing the references. I should have made it clear that you were citing 3rd party data to substantiate your claims, not just making stuff up as you went!
I think we can both agree the numbers are a little out of data, and in the case of the value of lost data, perhaps a little suspect.
And we can definitely agree that tape alone is not the most cost effective, reliable or secure way to protect data.
Posted by: Scott Waterhouse | July 22, 2009 at 09:32 AM
I can see how the loss of data can really hurt a company. Imagine losing all the contact info for all your clients, or losing track of what services each client signed up for. It can be a real headache to try and re-gather all that information ; not to mention it could make you look really unprofessional to have to ask for information again. You could waste payroll and lose customers. Easy solution: hire someone else to look after backing up your files regularly and even recover them if you need.
Posted by: Rhinebeck IT Support | July 28, 2009 at 10:00 AM