Quantcast

KNOW YOUR STATS: Find out what data is important and know how to read it


Guest Post by Daniel Elroy

No two stats are created equal.

Alas, it is a cruel truth that certain statisticians are just going to have to accept. These days, online marketers are able to collect an overwhelming amount of data from consumers. While it’s easy to get excited by the sheer volume of information available (“Wow, our product is extremely popular with divorced men who chew spearmint gum and are over six feet tall!”), the key is to zero in on the data that’s most relevant to you and optimize your campaign based on that analysis.

KNOW YOUR GOALS

We’ve all heard the overused maxim “Know thyself,” but there’s a valid reason it’s managed to survive all these centuries: it’s good advice. If your goal is brand awareness, then an email marketing campaign might be deemed successful based on open and click-through rates. On the other hand, if you’re striving for consumers to spend a fairly substantial amount of time on your website, then an enormous visitor rate simply isn’t going to cut it. An online marketing campaign can set itself up for great results by establishing from the outset certain criteria for which success will be judged, thereby allowing the analysis of future data collection to be as useful and streamlined as possible.

BE FLEXIBLE

So you’ve established your goals, and everything seems to be going swimmingly . . . except for the fact that you’ve totally misjudged the type of consumer you thought you’d attract. You thought your campaign was going to appeal to pimply male teenagers who wear braces, but it turns out that debonair older women who wear braces are the ones turning up in droves. Well, is this a bad thing? You can use the data you’re collecting to help nurture your newfound demographic and perhaps completely reassess your target audience, or you can stay the course and use the stats to determine why young men aren’t buying what you’re selling. A good way to collect useful data is the time-honored questionnaire, which you can use to pinpoint your audience’s precise likes and dislikes. Tactics & Techniques Video walkthrough demonstrating some tools & tactics which can be used to research your market online


(video courtesy of Web Analytics World)

KEEP YOUR HEAD

It’s easy to overdose on the abundance of information coming in during the course of a campaign. When a particularly striking anomaly occurs, it’s prudent to keep a level-headed approach to the data. While a sharp spike or drop can’t help but get your attention, it’s necessary to consider the root causes. Any number of things can cause this sort of “rogue data.” For example, a controversial comment on your website posted by a consumer or even a certain holiday approaching can cause certain consumers to seemingly disappear (or reappear) overnight.

FINAL THOUGHTS

The ability to collect data for the purposes of internet marketing will surely only increase in the coming years. By establishing goals, embracing flexibility, and maintaining composure, companies with successful campaigns can avoid getting drowned in the tidal wave of available information, and instead use the stats they’ve gleaned to reach their full potential.

What Happens to Reporting as Analytics Approaches Real Time?

Analytics has been about reporting for many years now. As more and more web technology moves toward real time, what does that mean for the future of reports? Let’s speculate!


Reports are generally used to confirm a prior decision or to provide support for an upcoming decision. What we really need is both confirmation on previous decisions and help making future decisions, which we currently use reports to accomplish. As analytics approaches real time, reporting will be too slow of a method for assisting in decision making. So what would future innovations look like?

Confirmation on decisions

All work starts off as tasks. Tasks can be correlated to onsite impact. Future systems will be able to map our tasks to impact and provide a conclusion on the activity. Some of those conclusions will be used by site optimization engines to make adjustments on the fly without need for humans to read and decide on reports. Some of the data will make it into quarterly board meetings and reviews with agencies and consultants.

Assisted decision making

While there are some unique scenarios, much of what happens online in marketing fits into patterns. In the not too distant future, we’ll see systems that use these patterns in a library to provide a dashboard of recommended actions. We will be able to complete simple actions with the push of a confirmation button. Other actions will be as easy to complete as clicking through a link. All of those actions will be tracked automagically for confirmation on the decision.

What do you think the future of reporting looks like for real time analytics?