Many marketers have worked hard in 2011 to develop appropriately customized ROI measures for social media. I have dedicated a few previous posts on how to approach these measures.
As we move into 2012, I would like to raise a key question of measurement that isn’t consistently addressed but is of critical importance. That is: How are you measuring the interplay and overall performance of your combined paid, earned, and owned digital media channels? How do you know which elements are driving the most value?
Most marketers are starting to understand that the most effective digital communication plans seamlessly integrate content across paid, earned, and owned media channels. As a quick reminder, paid media includes display ads, sponsorships, and paid search. Earned media is largely social in nature. Owned media is represented by content such as your brand website and native apps.
As marketers continue to build out ever more complex digital ecosystems, understanding how all elements, individually and in unison, are contributing to overall success is a cost of entry for effective optimization.
It is encouraging to see marketers recognize the need to create campaigns with content that is tailored for each of these unique channels. Creating a campaign concept that has “legs” for the diverse channels of paid, earned, and owned is no easy task. And neither is the process of effectively coordinating efforts across these channels. Where the trouble lies is that holistic performance measurement has been left behind.
We are measuring vast quantities of discrete elements within these channels. However, these measures typically exist within their own silo and can’t be compared to other parts of the ecosystem.
For instance, in order to isolate social performance, it’s necessary to include the influence of other media within each category: paid (display ads, owned, and email) and earned (SEO). There are a multitude of products on the market to measure earned metrics such as Radian 6, Lithium, and Sysomos.
Owned channel metrics are more siloed and are provided by the individual channel platform like your brand Facebook or Twitter account. Reporting of paid metrics can be pulled from ad-serving logs like Double Click’s DFA for display and/or can be inferred from your latest Google Search index. But these are also fairly isolated.
In other words, it can quickly become a spaghetti of custom or off-the-shelf reports, dashboards, and owner pathways that require labor and additional cost to gain a clear and complete picture.
Even aggregating all of these metrics would still ignore the hidden, incremental value of these channels working in concert. The weighted value of positive blog sentiment in the earned space is different from blog network sponsorship within the paid space. But their coexistence could influence their individual weights further. The assumption is that 1+1≠2, but it’s something else.
Let’s talk about solutions. Over the course of this last year we have employed advanced modeling and a lot of elbow grease to create what we refer to as the Connection Index. The goal was to create a single, holistic, measurement index that links all these discrete channels within the ecosystem together.
This approach creates a “heat-map” view of the ecosystem elements and allows for easy comparison of which channels are driving the most value. With this information in hand it is easy to make optimization recommendations about which channels should receive more funding and which should receive less or be eliminated altogether.
Below is a five-step process that you can employ to create a holistic, cross-channel score for your own ecosystems.
1.Define what success is.
a.Improved customer retention
b.Causes of demand generation
2.Collect all of your paid, earned, and owned metrics into a single data repository.
|Visits||OLA View Through||SEO|
|Page Views||OLA Click Through||Google+|
3.Develop a statistical modeling framework that distills multiple channel metrics into single measurement scores for paid, earned, and owned by doing something like the following:
a.Rotate and orthogonalize interrelated data streams within an ecosystem channel.
b.Utilize data reduction techniques to determine the underlying movement within the channel.
c.Further reduce the dimensions of the data to determine the cross ecosystem channel impact on consumer connections.
4.Choose or develop a technology platform that facilitates the following:
a.Frequent data extraction from channel sources
b.Interfaces easily with known earned analytics providers (Radian 6).
c.A transparent database system for storage of cross-channel data with easy access for QA, ad hoc analysis and modeling.
d.A dashboard UI customizable to enable your “definition of success.”
5.Dig into your new ecosystem’s connection scores to determine what touch points are working for consumers and where your growth opportunities lie.
Mapping these indices across time, product-use stages and/or other client-driven dimensions provides additional context and allows brand managers to monitor how connectivity to the brand varies over the consumer journey. An example of the results would look something like the following:
|Total||Trial Purchasers||Repeat Purchasers||Loyalty Members|
+ scores represent high scores for the media within a consumer group.
++ scores represent high aggregate scores.
~ scores indicate potential problem areas.
*This example uses stacked index limit of 150
To begin with, it’s apparent that all channels in unison have the most influence on trial purchasers, at 57.91, and that earned media has the highest influence overall at 58.54. Going a level deeper, we can see that trial purchasers, possibly induced by digital couponing, are influenced most by paid media, at 68.38. Repeat purchasers are most influenced by familiarity with the product and may shop via owned channels at 55.33. Loyalists, who may be playing an active role in marketing your product via blogs and Twitter, are most influenced by earned media at 69.08. Finally, areas needing additional investment or message adjustment can be identified as in the case with paid media’s relatively pale effect on repeat purchasers at 36.95.
Taking this example one step further, let’s say our definition of success is the influence of the brand’s digital ecosystem on offline sales. A powerful aspect of this model is its ability to establish casual relationships between the index and lower funnel, online and offline conversion activities.
For example, by applying the Granger Causality method to a CPG client’s transactional data, we were able to determine how index levels could forecast purchasing behavior. With this approach we identified causation between the Connection Index and product trials, repeat purchases and even product shipments. Causality would most likely be different across verticals but we strongly believe this may be an opportunity to demonstrate, with rigor, the link between discrete digital activities (i.e. social) and offline transactions that can eventually lead to ROI.
There is significant business value to be gained by stepping through a thoughtful integration process of your paid, earned, and owned digital channel categories. This is exciting territory, and it provides a range of opportunities to help advertisers realize the true value of each channel.
So let’s toast to all the great marketing accomplishments of 2011 and put our heads together to solve the challenges awaiting us in 2012.