Web Analytics?

It seems to me that a good place to start would be defining web analytics.  I think we’ve gotten a little out of whack with the expectations around this job, and not in the right ways.  What do these people do?  How much should they be paid?  What resumes should stand out when you’re hiring these people?  All very good questions.

So, what is web analytics?  It turns out that the answer is both simple and loaded: web analytics is pretty much everything.  It’s a multi-discipline practice that aims to quantify today’s success and uncover usability, design, architecture, copy, product, and marketing optimizations that will breed even more success tomorrow.

But no!  This is simply too much for one person to understand, never mind do well! There aren’t enough hours in the day to run all of these reports!  There is no way that a usability person could know HTML!  There is no way that a product person could know design!

Are you sure about that?

The Bigger Picture

Pick up a Wall Street Journal and look on the front page.  What’s the one word that appears more than almost all others?

“Analyst”

Watch CNBC, read the Motley Fool, or flip your radio to a business talk show on AM (do we even know how to do that any more?). Analysts, analysts, analysts.  These guys are paid big bucks, make CEOs shiver in their thousand-dollar shoes, and cause billions of dollars to move with the stroke of a pen.  Do these people sound like the Omniture ferrets we know?

These analysts research corporations very deeply with the aim of understanding a business inside and out: where it’s going, where it’s been, and what it’s capable of.  To do so, they leverage their knowledge of accounting, marketing, product development, competition, capital valuation, macroeconomic stimuli, political stimuli, and even weather! (click here for a more detailed explanation of these)

They have degrees from Ivies.  Sharp suits.  Expensive haircuts.  3 computer monitors.  2 telephones.  Now those are analysts.

And you know what?  We should demand the same from ourselves.  Especially the 3 monitors part.

What is “Web” Analytics, Then?

Well, let’s start with what web analytics isn’t:

Web analytics is not the measurement of something. If you’re paying someone decent money to tell you how much revenue you made yesterday, I have a bridge over the Pacific Ocean I’d like to sell you.

Web analytics is not defining success or determining if you’ve reached it. Sort of the opposite of the above.  If you’re asking someone in this role to tell you what “success” is, that is not necessarily good – although, they should be helping the business translate success terms into web terms and vice versa, and once earned, should have a seat at the strategic table.  But when it comes to defining success, the only right way is from the top.

Web analytics is not Omniture, WebTrends, Urchin, Google Analytics, Clicktracks, or anything where improvements are measured in version numbers. “Building” is not a hammer.  ”Driving” is not a Garmin GPS.  ”Cooking” is not an oven. Web analytics is not a tool.

So what is web analytics, then?  It’s more or less answering these five questions, not individually, but holistically:

  1. Who is coming to my web site?
  2. What are they doing (or better, what are they trying to do)?
  3. What is the gap between what they are doing and the ideal? (and do we mean ideal for the business or the customer? good question!)
  4. What are some concrete ways we can close the gaps?
  5. How can we get more of these people?

That’s not so bad, right?  Well, you show me an Omniture report that answers this group of questions and I’ll show you a pink velvet unicorn that can cook pancakes.

It takes the financial analyst months to answer a similar set of questions:

  • What is this company worth?
  • What are they doing to grow?
  • What does the stock market think they’re worth?
  • How does that compare with what it is worth: Is it a good buy/sell/hold?

Again, seemingly simple questions, but they have to leverage a lot of knowledge, experience, and research to answer them.

Think about this at your workplace.  No more reports that show page views, sales, or uniques.  Get real answers.  Drive real decisions.  Have real conversations.  You should be talking about the business, product plans, marketing plans, overhead, technology issues, human resources limitations, and rolling all of your web data into the context and limitations of the world around you. Because there is a world around you. Everyone will be better for it.

What It All Means

At the end of the day (don’t you hate it when people say that?), there are only two things that determine the success of a business over time: revenue growth and profitability.  Everything that a business does helps or hurts one or both of these things.  Everything.  So think twice about the Swingline.  The generic is just as good.

The effective web analytics practitioner is merely tasked with answering the above four questions in the context of growth and profitability.  The questions will span many disciplines in a business, including architecture, design, usability, marketing, advertising, pricing and product strategy, and more; which will require the analyst to collaborate with department specialists and work through the details.  But the effective analyst should never consider herself a specialist.  The strength of the analyst is in not being a specialist.  The moment the analyst becomes married to a particular discipline, they lose focus on the big picture, which ultimately results in clouded judgment and a sub-optimal outcome (thanks to Avinash for the ‘let you down easy’ phrase “sub-optimal”!).

The reason it’s so important to not be a specialist is that the analyst has the incredibly rare opportunity in an organization to be clear-headed about their recommendations.  Ask a marketing person what the best next step for the site is and you get a marketing answer. Ask IT, and you’ll get an IT answer.  HR? HR answer. Analytics is one of the only disciplines, with the exception of the COO and CEO (and a handful of others), that is truly central and can order their recommendations by greatest company impact, rather than greatest personal impact. This is incredibly valuable to the executive team, once you earn your seat at the table.

And would you believe that the greatest outcome of good analytics actually isn’t the betterment of the web site? When you come up with ideas of what needs changing, the most important part of the entire adventure is watching how the work actually gets done. Do some teams work poorly with others? Do some people defend poor decisions because it may affect them personally or their department’s bottom line? Are there knowledge gaps? Are there talent, skill, tool, or technology gaps? Again, the analyst has the very unique opportunity to judge process through un-tinted glasses, and help the executive team tighten up and improve efficiency company-wide, all without sounding like they’re complaining.  This is an amazing job, but it’s not one we have realized the potential of yet.

The effective analyst will piss people off.  They will tell people that their ideas didn’t work.  They will have days where people aren’t their friends.  They will argue with executives, whether directly or indirectly.  And they will make the whole company better at the same time.  If you can handle the ups and downs, get ready for the most interesting and rewarding job you can imagine.  Except for maybe being a travel writer. Or a photographer for National Geographic.  Or an astronaut.  Or Simon Cowell.  OK – maybe not the most interesting, but it’s still a great job.

Posts you might like:

Are you a web analytics beagle?