Let’s ask a simple question. If your servers were down for a week, would your life change? Could you meet payroll? Would you have to cancel that family vacation? Could you pay your server bill? If you are unsure how to answer any of these questions, then you really need to take a hard look at how much you spend to assure your IT infrastructure continues to earn you money.
Too often, I see clients who started with a small project that later became a stable source of revenue. Along the way, they’ve never re-evaluated their hosting infrastructure. Consider the businesses who had to endure the Valueweb/Gate/Hostway merger. More than 500 servers dead for nearly a week. I wonder how many of these businesses are now either out of business or in a compromised financial position?
IT will Fail
Just like the stock market, past performance does not indicated future results. Just because your server has operated flawlessly for 2 years does not mean it will continue to do so. The hardware, network or human error can cause an outage at any time. Google did a huge study on hard drive failure. Their conclusion? SMART and similar items like temperature are poor indicators of drive failures. In other words, there’s no easy way to predict a drive failure. The point is outages will happen, and you need to be prepared.
A key component in looking at how much to spend on your hosting solutions is to ask how much downtime can you afford. So what do I mean by “affording downtime?”. To figure out how much downtime you can afford, you need to get a grasp on how much an outage event would cost your business. The more sensitive your business operations are to an outage the less downtime you can afford. Trying to identify exactly how much downtime you can afford is not an exact science, but there are some useful starting points.
Costs of Downtime
Most of the information you find on the web for calculating the cost of downtime is geared to large businesses, especially the manufacturing sector. Many small online businesses don’t fit these models and have many indirect costs associated with a downtime event. I like to break cost into direct and indirect costs. With direct costs being those immediate expenses and losses related to the outage and indirect costs being less tangible or potential future losses. If you’ve ever dealt with a server compromise or crash, you are likely aware of the more common direct costs:
These only represent a starting point in looking at costs. There are many indirect costs you need to consider as well:
If you server is down for an extended period, then your search engine results may be impacted. Also, consider the client who bears with you through the outage but then 3 months later cancels once they’ve moved their operations elsewhere. Also, with many server providers, the restored system may not be an exact match to the original, so new support issues may pop-up days or weeks after the recovery.
Trying to get a clear accounting of these costs is critical to understanding how much downtime you can afford.
Crunching the Numbers
I like to look at downtime from the standpoint of the number of hours your operations are impacted. Set benchmarks for 30 minutes, 60 minutes, and the in reasonable increments for 72 hours.
If you are a web designer for small businesses with no e-commerce operations, you may be able to “afford” 24 hours of downtime with negligible business impact. Provided you have off site backups and a support solution in place, you could likely move your operations elsewhere within 24 hours with very little lost. So 24 hours of downtime cost you very little. Now what if you did not have backups or a support team in place? Say your host was down for a week? How many of your clients would stick-around? Now the downtime is becoming costly. That extra $150/month for a management service with emergency support and offsite backups does not seem so costly compared to losing a $10,000/year account because you did not have the infrastructure in place to manage your downtime.
I am always amazed that someone will pay $350/month for the car they drive to work but not spend more than $200/month for the infrastructure that allows them to pay for the car.
Infrastructure as Insurance
Downtime costs money and can cause a loss of future revenues. So just like your office catching on fire, for which you likely have insurance, investing in the proper infrastructure insures you against downtime events.
Insuring yourself can be something as simple as spending an extra $20/month for RAID and $50/month for a reliable offsite backup system. That is only $840/year to protect you against drive failure and provide a recovery method in case of facility failure. Typically, we see charges for recovering from a drive failure fall between $300-600 per event. This does not even include any loss revenues. RAID and backups can help mitigate the risk of a drive failure, and for most operations, the cost to put it in place is trivial compared to the cost of the downtime.
Last Words (Probably not …)
I recently learned that a client’s operations had grown from a few thousand per year in online revenue to over a million. At no point did anyone stop and ask, “Have we beefed up our IT infrastructure to assure we don’t have a outage?”
If you take away anything from this rambling, let it be:
Take a careful look at your IT operations every year (or sooner if growing quickly) and ask yourself, “Can I afford to be down?” If so, for how long and how much will it cost? Then make sure your IT infrastructure does not have more downtime than you can afford.