I want to have a bit of a chat in this post about cheap. In particular, I want to get two key points across.
- Cheaper is not always better.
- Cheaper is not always cheapest.
I’m going to start with a simple bit of family history for me: my father-in-law (soon to be formalised after almost 22 years) is a retired mechanic. Now, I’m not talking a mechanic who puts your car on a hoist and plugs a console into the on-board systems of your car to download diagnostics; I’m talking the sort of mechanic who can lay his hands on your bonnet while the engine is running, listen for a few minutes, and tell you with a fair degree of accuracy what, mechanically at least, is going wrong.
Before I met my father-in-law, my general approach to buying anything mechanical was ‘cheaper is better’. If it was a choice between a $100 lawn-mower and a $400 lawn-mower, I’d get the $100 lawn-mower. If it was a choice between a $99 vacuum cleaner and a $400 vacuum cleaner, I’d get the $99 vacuum cleaner every time.
My reasoning always went along the lines of: sure, the $99 vacuum cleaner might break down sooner than the $400 vacuum cleaner, but it was cheap and could be thrown out and replaced. Over time I learnt the value of buying something that was mechanically sound and readily repairable, even if it cost me a little more initially. That $99 vacuum cleaner is technically cheaper, until you take into consideration that it runs at a quarter of the wattage as the $400 vacuum cleaner – and you’ll buy 3 in 3 years and throw each one out when it breaks down, whereas the $400 one might last you a decade or more, and be repairable.
So here’s the first lesson: cheaper isn’t better. This isn’t a spectacularly new lesson; consider for instance the old adage, “the bitterness of poor quality remains long after the sweetness of low price is forgotten”. Yet a digital, consumerism economy can sometimes obscure this fact. We have to keep it in mind when we’re looking at options, and consider the bigger picture. (That sometimes means even considering the intangibles: that ‘bargain’ may be half the price, but if using it costs you double the time, have you achieved anything?)
Now, the second lesson is more something I’ve come to appreciate working with deduplication: cheaper isn’t necessarily cheapest. This is essentially a recommendation to look not at a simplistic price comparison, but a value comparison: do you establish the cost of a solution by the sum of its constituent pricing, or the result it delivers to you?
Let’s say you need a solution that will provide protection for your environment, and you get two responses:
- “I can sell you 100TB of usable storage for $100.”
- “I can sell you 100TB of usable storage for $150.”
(Of course, I’m just using simplified numbers there to keep things straight forward.)
Now, on the face of it, the first one is cheaper, right? But let’s qualify a little bit more:
- The first option will offer 8:1 deduplication
- The second option will offer 25:1 deduplication
This is where we look for value to define cheapest. If I get 100 TB of usable storage for $100 and that has an 8:1 deduplication ratio, that means I’m getting 800 TB of logical storage for $100. But, if I get 100 TB of usable storage for $100 and that has a 25:1 deduplication ratio, that means I’m getting 2.5PB of logical storage for $150. The net result? In terms of backups that can be stored, the cheaper solution will cost you 12.5c per TB stored; the more expensive solution will cost you 6c per TB stored. The cheaper solution is not the cheapest.
You know that because you can see the answer in the question, ‘which gives you greater value?’
Yes, again, I’ve made the prices up, but this isn’t about the specific numbers I’ve used: it’s about the comparison you need to make when comparing two different deduplication systems. It’s one on delivered value. When you’re looking at data protection solutions, it’s always important to remember: cheaper doesn’t mean better, and equally, cheaper doesn’t mean cheapest.
As soon as you start to get into something that you’re buying for a function, to solve a problem, then the cost should become just one factor to consider, and you need to ensure you understand what value the solution gives you.