Determining the value of network monitoring in the data centre
Mon 23 May 2016
Monitoring beyond numbers means more effective workplace operations for both business and staff, argues Dirk Paessler, CEO of Paessler, an IT monitoring specialist based in Nuremburg, Germany…
Anyone who is going to invest in network monitoring will at some point be asked to justify the costs, and possibly even be asked to provide a concrete ROI number. For data centre managers this can be a nightmare, as the true value of network monitoring over time can sometimes be difficult to demonstrate – a new system that doesn’t even reduce upfront labour costs is a hard sell. There are ways of making a convincing business case for IT monitoring, though you will need to include a number of “soft factors” in conjunction with the basic data.
Sitting down to make a calculation can be an arduous and thankless task, but it in the long term it will prove worthwhile. So many of the costs you will be saving emerge only after a long time of smoother operation, and will differ from location to location. With that in mind, the first step is to draw the distinction between those hard and soft factors.
The upfront costs of network monitoring are relatively straightforward to estimate. Licensing, hardware, and implementation costs should be provided by your provider, and can go straight into the “costs” ledger. You can follow that with an enquiry into costs for new modules, for upgrades, and for the maintenance fees – with some leeway for purchasing potential future technology that will require a new monitoring solution.
To work out the hard savings you might make from monitoring, you will first have to identify the costs that it will smooth out. These might stem from the inefficiency of a system lacking the latest software updates and versions (network monitoring identifies software versions across a complex IT system, allowing for quick replacement), the purchasing of unnecessary modules, and the slowdown from not purchasing the modules you actually need (network monitoring allows for easy planning on this front). And don’t forget the extra staff hours from the “on-call” manual monitoring that network monitoring eliminates.
Giving each of these costs a number may be difficult, but estimating them at least gives you a baseline ROI. Unfortunately the rest of the calculation is much more speculative, though no less essential. From here the challenge becomes how to calculate the cost to your company of downtime. Plenty of attempts have been made to put a figure on the money your company can expect to lose per minute of downtime – in 2014 Gartner put it at around £3,900, or £234,000 per hour.
You will have to make your own calculation depending on your business –if a data centre operating on behalf of a high-volume e-retailer goes down it will harm business more than in the case of an office where downtime just means employees might have to do other tasks for a while. But in either case the amount of money lost is going to make monitoring’s ability to reduce the likelihood of a failure far more attractive.
The soft costs
After that the calculation gets even more difficult to quantify. The “soft costs” take place mostly in the heads of both customers and employees. If the data centre goes down, then customer support systems fail, and repeat customers are likely to be lost in their droves the longer the frustration continues. If they then start complaining about their experience on social media, telling their friends or simply going to a competitor, then the consequences begin to cascade outwards. Since the departments that deal with that sort of brand reputation and customer service are usually far removed from the data centre managers, the connection between these events might not even be that explicit. It is only when the annual financial report is filed that any change will be noticed, by which time the effects of downtime have become so sunk in other factors they are impossible to isolate.
At the same time, employee morale will suffer if it becomes clear that the servers are failing regularly. The moment data loss occurs or servers slow down on a Wednesday morning, your average employee is looking at a morning’s lost work – and possibly even the careers pages.
You may have to create another column for these costs, making it clear they are based on your own opinion of what will happen in the event of data centre failure. You will also have to acknowledge that no monitoring solution is so fool-proof as to predict and head off any failure, so you will need to factor in the probability costs against potential losses. Your resulting ROI calculation will be highly speculative, but it can be made to work provided that you have solid reasoning to back up your expenditure on monitoring.
You may also need to make an addendum to your workings – long-term benefits. It is very easy to limit monitoring in the data centre to nothing more than heading off potential disasters and the costs that come with them. You should also be thinking of how much money you can save in non-crisis times.
By reducing IT staff workload you are making them more efficient. The automation and intelligence of network monitoring tools enables IT staff to spend less time searching for the source of smaller outages, and these time savings can add up over the course of a year.
A survey by Paessler polled 648 of its customers in 2015, and 64% reported saving “significant” to “exceptional” amounts of time by using network monitoring, while 24% reported saving “some” work time. The knock-on effect is that staff spend more time on long-term strategic projects, as opposed to routine troubleshooting, with positive impact on the business as a whole.
In addition, network monitoring allows you to do long-term data collection to analyse how and where hardware and bandwidth resources are being utilized, so in the future routers, switches, servers and other hardware can be acquired and used to meet exact requirements, helping to avoid unnecessary costs due to unused resources.
The case for network monitoring
What you eventually piece together is a rationalisation for monitoring beyond the numbers, which simply means a workplace that operates much more effectively for both business and staff. If you wanted to try and calculate it you could, but after laying out all the potential long-term benefits, it should be clear to any management executive that they are the kind that don’t fit neatly into a ledger, but are ultimately essential to the smooth operation of a data centre.