“Actionable” is one of those new words that has sprung up in the digital economy, typically referring to data or events that can be the foundation for a resulting beneficial effort on the part of the possessor or recipient of the “actionable” item. In the world of measurement and web analytics, it often is used to define an insight gained from understanding visitor behavior: “Understanding that the intra-page behavior in our download funnel was indicating a long load time after clicking certain choices, gave us an action item.” Meaning: “That insight was actionable.”
It implies an imperative to the recipient: now you must act. Do something. Make use of this. And in today’s rapid spin cycles, that can mean “right now.” In the above example, it meant “figure out what is causing longer load time after those clicks, because it is probably depressing your scenario completion rate.”
But how hard is it to take action in the real world?
With all the “actionable insight” that can be gained from digital measurement, web analytics, and mobile analytics, what is behind the glacial pace of change within web properties even in the face of actionable imperatives?
There are two chief reasons.
Politics, Interpersonal Dynamics, and Fate
First and most difficult to overcome are the combined political and interpersonal factors surrounding change. It is important to note that there is no folly in attempting to keep team spirit high – or even one person’s spirit high – while leading the charge on “action” that may affect them. Often enough, action means displacement. Something old is going to give way to something new. It’s really part of the cycle of living and dying; and without delving too far into metaphysics, it’s often the case that someone’s fate (at the job, for instance) may at some point, in some manner, be tied to the fate of the item needing change. For it must be the case that the item needing change was, at some point, the new-new thing, and it was probably promoted by someone, and perhaps it was even successful at one point. Or, it was never a successful effort and, until measured carefully, its ineffectiveness was never known.
There is no simple way to break the news that “the baby is ugly.” The creators of a site feature or the underlying code often have developed a fondness for their own creation that trumps any rational discourse about its effectiveness. It pays to understand this dynamic; and that to bring change and effective results, one would rather not start internecine battles and dark rumors. Empathy is an ingredient often missing in the medicine-bag of the typical change-agent. There comes a coldness with the numerical imperatives brought about by quantitative analysis. But creators and developers are not numbers – they’re people with bills to pay and psychological needs that ought not be toyed with in the interest of immediate success. The change-agent must take on a persuasive tone and demonstrate the availability of a judgment-free zone to those that may be affected by change. In an atmosphere of trust, even those who took ownership of what now must be changed, can now be made to feel understood and will likely be more free in offering help in bringing about change – because they are thus given an opportunity to be the change, rather than just witness the destruction and replacement of their pet item that perhaps tested poorly.
The second (and awfully widespread) difficulty that keeps action from following actionability is simply a lack of consensus around what the change process ought to look like.
eBusiness 5-Step Optimization
Here I will mildly suggest that we can have a process for change that pulls in all the necessary parts of the measurement structure as well as the content contribution teams that ought to react when change is indicated. I have called it e5o or “eBusiness 5 Step Optimization,” but you may recognize some of it as the type of change cycle found in many organizations – except it is specifically tailored to the business of making successive rounds of change to digital assets. It will work for both web and mobile environments. Social media, chiefly a “reach” technology, may be appropriate for this model upon agreement that the point of social is to drive traffic to desired actions – but that is another discussion.
Here are the steps:
- Define drivers. There are many “pages” on the site and lots of content. Your backend can serve up content dynamically, targeted, and specialized per device even. But add all those up and it’s just costly bandwidth without specific goals. Time to decide what it is you want your users to do – for each section of your site, for each page. Often the drivers, sometimes called key performance indicators, are better understood when we have defined the type of site we have. But every page should have a reason to exist – either to drive toward conversion or to convert directly on the page.
- Build metrics. This is the core technology piece that makes web analytics work. Using tools like Google Analytics, Omniture, Webtrends, and others, targeted expertise is needed to construct an analytics environment (tagging, configuration, report-building, data integration) that responds to the questions posed in part one – how well are pages/sections/content performing against the desired results defined as “drivers”?
- Plan actions. We have referred to this part above – we have received reporting that indicates, say, underperformance in a key sector of the site. How to break the news without causing more trouble than it’s worth? How to align the organization and its vendors such that the tasks needed for improvement are well understood, and delivered in a timely manner?
- Create changes. Here we have taken the “actionability” imperatives to the “actors.” These are your content editors, copywriters, site-designers, architects, and coders; one or some or all of them need to take ownership of change. If the “plan actions” stage has been properly managed, these changes should come back with a minimum of fuss (much easier said than done). Responsive change is perhaps the most critical factor in site improvement.
- Measure success. Did the new changes make an improvement? How much? Or, did the changes make it worse? Be prepared for tough decisions based on quantitative analysis. Make sure everyone understands this is an iterative process – part of life in the digital world. If this were a shampoo bottle, here is where the label would say: “Lather. Rinse. Repeat.”
It would have been easy enough to post the five steps to success as a standalone “best practices” module for web analytics; except it isn’t just web analytics, and even if it were, it would not exist in a vacuum. For there are always people involved, and people are far more complex than the algorithms we have yet created. Success in “acting upon actionability” depends heavily on understanding not just the process, but the element of humanity wound tightly into all our communications.
This column is dedicated to the memory of Steve Jobs.
Emily Ma, product director of Tencent’s advertising platform products department, was a keynote speaker at ClickZ Live Shanghai where she discussed the ... read more
The terms that customers type into your site search function can help you to gain an understanding of user behaviour and can be used to optimise ... read more
Google Analytics comes with lots of standard reports and settings, but with a little customisation you can extract much more value. One way is ... read more