How to save your skin with sensible alerting

Have you ever been caught out by not knowing that something in your business was going off track until it was too late?

And have you ever been overwhelmed by the massive amount of data and KPIs that are continually produced by your business, so that you spend far too long checking on the figures rather than focussing on building?

Things happen all the time which hurt you till they’re spotted and fixed – some part of a complex system goes wrong, or one of your external service providers messes up, and it doesn’t get picked up and repaired until it’s already become major. This stuff costs businesses a lot of money, both in revenue loss from noticing problems too slowly, and in staff time lost anxiously checking on the figures.

Here’s how to do it right, so that you can set and forget metric monitoring and alerting which will keep your business safe from harm:

Identify the key metrics and their time-sensitivity. Any complex business has hundreds of numbers you could track. You need to prioritise – which ones have genuine business impact? It’s a good bet to start with what the CEO looks at (certainly daily revenue, and probably a couple of other industry-specific metrics) and track backwards from there – can you split that out into maybe 3-4 key numbers which drive your business results?

You also need to decide how time-sensitive you are, and how fast you’re able to react to problems when they’re spotted. If you’re a lightning-speed online business with engineers on call and massive hourly revenues, you’ll want to track these figures by minute – alternatively, if things move a little slower for you, it’ll be more realistic to count by day.

Analyse patterns and make a basic forecast. This is where you get analytical. To pick up any but the simplest business issues, you’ll need to not only be watching for where a metric goes beyond a fixed value, but also for where it does something different to its usual pattern – there’s a big difference between sales being £0 on Sunday when your shop is closed, and sales still being £0 on Monday when it’s open. You need an expectation for each data point based on its regular patterns, whether based on opening hours, consumer behaviour, or any other factors which affect you.

Understand each metric’s level of variability and set your thresholds. The analytical bit, part two – you also need to analyse how much a metric usually varies, to set tolerance for how far away it gets from forecast before you should be worried. See my previous post on standard deviation for more explanation.

Start monitoring! Now that you’ve got the machinery set up, you just need to set the alerts (email/text message?) and set it going! What a dream. Or at least, when things *do* break you can know about it and get it fixed, before anything gets to the running and shouting stage.

If baking all of this yourself sounds like hard work, Magic Dashboards’ first product is going to do exactly this – we’ll monitor and analyse any metric you give us, and alert you immediately if it does something unusual. It will be available in private beta this month – contact us here if you’re interested in signing up.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s