Crawl-Walk-Run is a framework for digital analytics and performance optimization that I’m working to make approachable for people across experience and digital disciplines, and comprehensive enough to cover the spectrum of business goals involved in digital marketing and advertising.
At a high level, each stage differs in terms of:
- Questions answered at that stage (what, why, how)
- Time perspective (past, present, or future)
- Level of optimization applied (reactive, tactical, or strategic)
- Activities used (monitoring, testing, predicting)
Let’s see how this applies:I’m using the standard media category framework that both Forrester Research and Fleishman Hillard use: distinguishing Owned Media, Paid Media, and Earned Media. Here are the core definitions (as provided by Forrester’s Sean Corcoran):
- Owned Media: a channel that a brand controls, such as a Web site or Mobile site
- Paid Media: a channel that a Brand pays to leverage, like Display ads, Paid search, and Sponsorships
- Earned Media: when customers become the channel, like “Word of Mouth”, “Buzz”, or anything “Viral”.
Owned media is today’s focus – primarily websites and mobile sites. For owned media, here is how Crawl-Walk-Run applies:
- Crawl: Monitoring and Reporting – the goal of owned media analytics at this stage is to understand what happened in the past. In some cases the analyst or team will explain why something happened, but that’s not the primary objective. The tools used are often the standard reporting outputs from traditional web analytics tools like Google Analytics, Omniture SiteCatalyst, Webtrends, NetInsight.
- Walk: Test and Profile. When you move to this stage, all your actions are directed at performance improvement. Analytics moves from showing what happened in the past to identifying what can be changed in the present to improve. Online testing is often the biggest focus, but another way to improve in the present involves understanding a website audience through surveys that combine concepts like “Primary purpose”, “Task Completion”, market attitudes, off-site behavior, and other attitudinal elements. Tools including online A/B and MVT testing software like Google Website Optimizer or Omniture Test & Target, as well as online survey tools (appropriately integrated with your analytics package, obviously), like SurveyGizmo (my preference), Zoomerang, or something else.
- Run: Predict and Personalize. In this stage, you have moved past basic testing (and even most MVT testing) at the large scale. Now, the goal has moved to predicting the future, accurately. There are two categories of tools used here: dynamic content tools (like Monetate, x+1, and others) and sophisticated statistical modeling software like R, SPSS, and SAS.
How does this relate to how you think about analytics and performance improvement for owned media?