Standard Deviation and Goal Setting

It seems to be that time of year, a few of our clients are asking us to help set goals for the upcoming year. Typically my experience with goal setting has always been the edict of the management, and arbitrarily increasing goals by 5% or how ever much they wanted to grow. This has always seemed odd to me and I wanted a way to back up the goals to see if a) they were achievable b) if they made sense for the coming year and c) if they were too achievable and we'll always be going above the goal.

To do this, we've started using Standard Deviation based on prior months data to get this new goal. Ironically, its a pretty simple way to tell where you want to be next year, and back it up.

A little background..
Businesses today are all about six-sigma, a system originally created by Motorola to improve processes by eliminating defects. You'll hear the term mentioned in nearly every corporate boardroom about the nation. Standard Deviation is the core to this whole process because it shows what a 'significant change' actually is, in whatever youre measuring. Pretty cool stuff.

Avinash Kaushik had a great post on using Control Limits to judge swings in traffic. The method he discusses uses Six-sigma and standard deviation to work it all out. Its a great way to show clients if the change is indeed good, bad, or just doing what it normally does.

So how does this apply to goal setting ? Because based on prior performance, you can say with a certain amount of confidence that your can actually get to your goal.

I'll grab a sample on how to do this soon, but you basically find the average of your data points, find the standard devitation of the data points and add them together. This will give you one standard deviation above the mean. Add the average plus twice the standard deviation, you have two... and so on. The more data points you have the better but overall thought is to get an idea of where we want to be going forward. You can then present options of what a 'good', 'better', and 'best' goal is.

To go a little deeper into this, a goal in the 'Good' range will be hit 34.1% f the time, a goal hit in the 'Better' range will be hit 13.6% of the time, and a goal hit in the 'Best' range will be hit 2.1% of the time. This is based on the typical 'bell curve' . Ideally, everything would be 'Best' but 2.1% of the time isn't always realistic (and depending on your data, ever achievable). If the difference between the whole deviations is too large, you could pare it down to half of a deviation.

Overall- its a pretty neat way to show what a good goal for the upcoming year should actually be, and you can have a bigger drive to reach it if you know that it might actually happen.

Web Analytics Wednesday @ ZAAZ

ZAAZ hosted Web Analytics Wednesday (WAW) again last evening. It was the fifth or sixth time I hosted the event for the group, and they keep getting better. We had people attend from ZAAZ, Microsoft, Razorfish, Farecast, Neutrino, and Whitepages.com. Its interesting now that after having done this a few times, we are seeing recurring people which is great because you get the feeling of community now among the group instead of everyone not knowing each other and having to play the get-to-know-you dance. The main topic I heard going around a lot last night was Google, Google, & Google. Why the heck not? They're hot right now. Snapping up Doubleclick just to Microsoft couldn't have them and the latest release of Urchin there is a lot to talk about in the space. Mostly the talk is generally about the industry and whats going on, rather than best practices- which to some extent is okay. Best practices is what gives you a slight lead among the competition, which essentially is what all these people are anyway. Keep your friends close, but your enemies closer I suppose right? Regardless of all of it- WAW is really panning out to be a good event for Web Analytics professionals to develop community among themselves and finally feel like they belong. Most companies right now just have one or two web analytic FTE's, and its hard to play in a game like that when you have all this creativity but no outlet. This definintely provides that.

ta-da

finally thought of a fun name for my blog that represented a fun creative side, combined with the nuts and bolts of what I do for a living. Hmm, sounds an awful lot like my employer =)

More to come... stay tuned, we'll be covering a lot of thoughts on Web Analytics and how it applies to the ever expanding reach of the internet