Web/Behavioral Analytics 2011 Wishful Resolutions

Well, it is that time of year when everyone starts making new resolutions or attempts to abide by ones from years past.

As with most resolutions, they sound great at first but not too soon after they are forgotten and the excuse making is not too far behind. Whether it be to go to the gym more often, quit smoking, spend more time with the family, etc… Our regular day-to-day lives eventually “get in the way” of our not so lofty hopes.

This is also often the case in our work life as well. People talk about what they are going to do so that they advance their career, surprise their boss or come up with the next great, earth shattering, idea. Once again, this ends up being an insurmountable goal or it gets put on the back-burner due to dealing with all of the challenges your regular job throws at you – the job they pay you to do.

So, for this new year I’ve decided that setting a number of goals for myself would be pointless. Rather I’d like to propose a couple (small) changes that could perhaps help you make an impact in your career development and if they are lucky, your company’s success. This is not meant to be me standing on a soapbox but rather a couple of things that I have seen past clients change that improved their business (and made executives happy).

1. Detailed / Summary reports are a good start to sharing data with-in your team but management doesn’t care
   • While these are often reports that help guide you to further discovery, they should not be used to drive specific change to site or specific marketing effort. If you think about what typically happens with a report; you generate a report (or series of reports), that report would then be handed off to someone who would need to interrupt what the report is telling them (error prone), that interruption would then be once again summarized to either a development or marketing manager that would then ask either another person or team to action on this chain of potentially misunderstood information. In this simple example, that one report (or associated information) crossed no less than five hands before any action is taken.
2. Juxtaposed to #1 – Scorecards provide great insight to management on change effectiveness
   • If you think about what most people in management often ask it is commonly “what is the most effective effort.” For the most case this can be boiled down to
     • which ‘anything’ is providing the most X
     • which ‘anything’ on the watch/concern list is not changing
     • which should we cut due to poor performance

     This is what I’ve started calling “intersection management” (Green, Yellow & Red – respectively).

3. You’ve got some of the data – start spinning the cube!
   • I understand that people have built their careers by focusing on web analytics and I applaud you. It has been a long uphill battle that started with some people back in the mid 1990’s and up until recently it has been a challenge to gain buy-in at many companies as was represented by John Lovett (@johnlovett) WAY back in 2009.
   • 
   • So now that you’re perhaps over the initial hump, how do you continue to bring value to your company and move beyond your current role? In many companies this means that you will need to become friends with people outside of your web analytics group. Why would you want to do such a thing? You have information…. and so do they! If you can merge your information together with their data, perhaps customer demographics, offline transactions, firmographic or other non-website details. Rather than building web analytics results you would have created behavioral analytics results.
   • Why do we do what we do? Is it to be the “hero”? Of course not, although if you think you are the company hero at least entertain people and purchase a Batmobile. We are all trying to influence what our company is doing online AND offline (if applicable) with more information to make better / smarter / more informed decisions.
4. Decisions, decisions, decisions: They aren’t your fault but they could be your success
   • If you were to stop delivering reports to people what would happen? You may get spoken to, replaced by someone who is willing to do the job as asked or they won’t even notice. There is the possibility of having the web analytics information fall off the table because in many cases there are no direct actions associated with the results. Not only should there be insight from the information but also actionable results, as alluded to in #1 & #2 above.
     Unfortunately for many people action means they look at doing some lightweight A/B, MVT, Product Recommendations or other non-structured interaction with the user. This is not to say that these approaches are flawed in any way at all but rather they are only part of what you could be doing to improve the user experience.
   • Do you know why sites like Amazon, eBay and Facebook succeed, aside from content? These sites make the experience about you.
     How do they achieve this? Not so simply, by first taking everything they know about you and secondarily about others like you (with similar behavioral traits) and delivering an experience that makes you feel at home. I don’t mean having your first name at the top of the page or maintaining the color scheme you configured on the portal during your last visit. It is a matter of bringing together what the user would likely look at (online), what is in the same realm of what they have looked at (online) and/or what their experience with you in the past has been like. There is a wealth of knowledge available – get off your hands and use it!
   • At this point you are likely thinking of a million reasons why you can’t do this; privacy, performance impact, resources, etc.
     I’m calling BS.
     These are all excuses that every other group in your company has to deal with on a daily basis. The fear is in the fact that you’ve finally got to the point that people are willing to listen to you and this could jeopardize that respect. This is where some risk taking and relationships you should have built play in. What you need to think about doing is finding a way to manage the volume of information in a way that allows for key targeted segments and influential factors. I give you pattern detection and predictive analytics which will allow all of the information from the business segments to start understanding which information is noise, what the important ‘bits’ are and what are the influencers for the important parts.
   • How do you make this magic happen? Go find the people in your company that have access to SAS, R, SPSS or something similar. These solutions will churn through anything you point it at and in processing the information using any number of models you can automatically deliver information that can be built into your content management solution, campaign management solution or any number of other platforms.

I know this is a lot to digest if you’ve never done these things. Have no fear, you aren’t the first and you won’t be the last but you can perfect the actions and results over time.

This post assumes that you would have access to many of the aforementioned solutions available to you or there is a limitless budget that would allow you to go get them. This is not often the case, unfortunately. So what could you do with limited resources?

There are a number of traditional ways that people have worked around a number of these problems whether it be using Excel, word clouds, demo versions of products or whatever other tricks you may have up your sleeves. Regardless what the challenge may be, your goal needs to become change – process change, result change and most importantly a change in the way you and the rest of the business think about the information/data that they have available.

If you have any suggestions or recommendations for other people on how to get around some of these hurdles, please share!

Many thanks to Michele (Jojoba) Hinojosa (aka @jojoba) & Jason Thompson (aka @usujason) for helping me edit my first blog post in a long time!