What it takes to compete (and win!) on analytics

Things have come a long way since the days of analyzing Web server logs (raise your hand if you were at THAT party.) Web analytics tools like Google Analytics, SiteCatalyst, and WebtrendsAnalytics have brought basic site tracking and reporting to the masses, often with little more than dropping some javascript on the page.  Closer to home, Business Intelligence (BI) and data warehousing tools such as MicroStrategy and Microsoft’s suite of tools provide a structured view into in-house data.  Google Site Optimizer, SiteSpect, and tools like Omniture Recommendations go so far as to provide intelligent testing, targeting, and recommendation abilities to companies that even a few years ago would have met serious challenges developing these competencies in-house.

But these tools are not how you compete on analytics.

What these tools do provide is time — time for your analytics team to be strategic, to answer questions that aren’t being asked, to be knowledge workers — instead of being 100% allocated to producing static reports. (And let’s be honest: most of those reports are never read anyway.) So absolutely, your company should have a third-party Web analytics package.  And management should be using it a lot more than your Analytics team does.

But let’s take a step back for a minute.  Does your company need to compete on analytics?  My own opinion is that if you’re not out-innovating your competition, they’re out-innovating you.

You need six things to compete — and win — on analytics:

1. Your analytics team must be specialized.  Hire experimental psychologists, sociologists, economists, econometricians. Hire skillsets in experimental design, predictive modeling, SQL, and SAS.
2. The analytics team must be organized as a strategic function within the organization — at least 50% of the group’s time must be self-directed.  If the team is spending all its time answering questions from the organization, you’re created a top-down organization — not a data-driven organization.
3. All significant behavioral and transactional data must be available on a reporting tier in a fully relational format – including session and page-view data. Asking an analyst to provide strategic direction based on data from something like Google Analytics is like asking a mechanic to fix your car just by looking at the speedometer.
4. Competing on analytics requires specific data-analysis tools.  If your analytics group doesn’t have world-class skills in SAS, SQL, or SPSS, you don’t really have an analytics group — you have a reporting group… which is OK, just recognize the difference.
5. A strategic analytic group requires a small amount of development, database engineering, and database administration resources available to them on an ad-hoc basis.  If you can’t make this level of commitment, no sweat — let your competition do it instead.
6. Lastly, you need a reporting solution that allows management, product, and marketing to self-service their reporting needs.  Whatever meets their needs is the best choice, but that choice has little to do with the Analytics function.

To sum it all up: The competitive landscape has changed. Web companies are gearing up to out-innovate their competition and they’re snapping up the best and brightest.  Masters and PhD graduates in the social sciences have the skills you need to take the next step, and in the current market, they’re cheaper than MBAs.  If you’re going to beat the competition through analytics, build a strategic group focused on asking questions rather than on answering them.