Jun 09 2009

Help! We have major issues with our analytics!

Published by at 11:26 pm under analytics,conferences,Google

On Monday, Joshua Ellis (a co-worker and Google Analytics guru — and someone I’m hoping to get to guest blog here soon!) and I presented at the Penn State Web Conference. Our presentation was called Actionable Web Anaytics for Higher Education.

As attendees started asking question during and after the session, a theme developed – a theme that we are all familar with and is consistently brought up (especially in higher education).

Common Theme - I/we have issues with our analytics set up/implementation/limitations – what do we do? Help!

I am confident, especially in higher education, that *every* web site has issues with 1 or both of the following:

  1. The web analytics implementation.
  2. Web site or system limitations that, in turn, limit the potential of the web analytics tool.

I want to reiterate that we *all* have issues with 1 or both of the above (hopefully more of #2 than of #1).  :)

First and foremost, take care of #1 above. Make sure your implementation is as sound as possible.

  1. Make sure all your pages are tagged (use the WASP Firefox plug-in to help).
  2. Make sure your Google Analytics account number is unique (from other colleges or units within your university). This might seem obvious, but we’ve found instances of duplicate account numbers on college or unit web sites within the same university. This obviously depends on your school.
  3. Create some basic filters to make your reports as clean as possible.
  4. In the words of Linda Bustos, Don’t do stupid things.

Secondly, data collection on the web is imperfect. Numbers will never, ever tie. In higher education, we have web sites that are glorified portals, use archaeic application systems, and, in the case of large universities, are fragmented and siloed. We have many limitations.

There are limitations that every web site deals with as well. When IT folks tell you the data is crap because of  deleted cookies, users using different browsers/computers, blocking javascript, etc., etc. … smile and say, “get over it.” There is nothing we can do about these and the data will never be perfect, but that’s ok. That’s why trending is important.

Take a look at this recent post by Avinash Kaushik about being comfortable with incomplete and imperfect data. Read. Heed.

Remember some basic facts …

Fact 1. Web analytics is in its infancy. In higher education, it isn’t even born yet. I’m not a biologist, but I’d say it’s about a zygote.

Fact 2. Web analytics is not cookie-cutter. Each site implementation/set up and site analysis will (and should!) be different.

Fact 3. Especially in higher education, measuring through to the actual conversion (enrollment, registration, membership, etc.) is very hard and, in some instances, impossible.

In higher education, we need to focus on what we *can* do … and there is so much we can do.

Use micro-conversions:

  • clicking on “register”
  • using the knowledgebase
  • watching a video

Look at the data that is screaming at you:

  • top landing pages with high bounce rates (find out why!!)
  • 404 page referrers (clean them up!)
  • top internal search keywords (terminology) you’re not using (or not using prominently!) on your web site

I also think that there are some (yes, rare) instances when we realize that an implementation might be so hosed that we need to blow it up and start over.

There may be many reasons for this, but whatever the reason, the data that we’re getting now is really useless.

Obviously there are many instances where this is *not* a good solution, but depending on your circumstances, it might be time to start over … as painful as that seems. This is my opinion and I do realize that a lot of people disagree with me here, but there you go. I said it.

Whatever the issues (and we *all* have them!), get over it. Find something that *can* be done. You’ll be surprised at how many actionable insights you can get out of your web analytics, even with all our limitations.

The greatest thing about all this? The simple fact that you care about this stuff means you’re way ahead of the game in higher education. That’s fantastic and makes data folks like me smile on the inside. Baby steps, baby steps. Do what you *can* do.

Comments Off