In anticipation of his upcoming conference presentation at the eMetrics Summit Boston, “Analytics Strategy & Roadmap,” we asked Tony Stagg, Digital Marketing Strategist and Web Analytics Consultant at thinkDigital, a few questions about his work in digital analytics.
Q. Where does digital analytics sit in your organization and how do you interface with your business units?
A. Digital analytics reports in to the SEO department, or in to the Paid Search department in most digital agency organizational structures. In some cases, it might report directly in to the President or CEO. In client organizations, it reports in to e-commerce, web operations, or marketing. It is rarely a department unto itself reporting directly in to senior or executive management. This is a problem in itself for reasons I can discuss later, if necessary.
Q. What’s the most valuable thing you’ve learned so far in 2013?
A. That there is a lot of confusion in both client, service provider, and agency organizations regarding the role and purpose of digital analytics. This is manifested in their organizational structures, their perception of what it requires to be an effective digital analyst, and how they are compensating these positions. Having worked for a variety of different organizations as a consultant, I have found that, by and large, these challenges are much more common than uncommon regardless of the industry vertical, size of company, and maturation level.
Q. What’s the latest analytics method/process/tool set that you have implemented and what advice would you give others?
A. I am experimenting with Tableau for data visualization. While Tableau is not designed to replace common data sources, such as Excel, or SQL, it allows you to represent very quickly different views of the data, and decide what visual format works most effectively, i.e. chart, line graph, word cloud, etc. It also allows you to do certain things, such as calculations or formulas directly in Tableau, without having to go back in to the underlying data source, such as Excel, to do this, and then exporting the data again. Tableau won’t fix bad data problems—that you’ll have to use other tools for, but it will help you improve the visual presentation of your data, which is half the battle in gaining acceptance for strategic insights and recommendations.
Q. What sort of attribution model are you using to allocate marketing funds?
A. Testing several different ones, because this is not a “one size fits all” proposition in my experience. Have recently been testing Google DoubleClick’s attribution tool, which allows you to run 5 basic attribution models—first click, last click, linear, time decay, and position- based. Running each model shows different patterns of channel interaction along the conversion path, which is important for understanding how channels complement or do not complement each other.
Q. Sneak preview: Please tell us a take-away that you will provide during your talk at the eMetrics Summit.
A. One of the take-aways is this: Digital analytics needs to deliver value to the business. It is virtually impossible for digital analytics departments to accomplish this in a consistent, repeatable way, if they don’t have a strategic plan for the identification, development, and delivery of analytics capabilities. This is a very important undertaking and should not be left to the business to define. Analytics departments need to step up, and demonstrate that they need to be taken seriously. Analytics is a whole lot more than dropping a batch of Excel reports every Monday, but all too often that is what it becomes.
Q. How relevant is the concept of Data Scientist?
A. In some ways, not relevant at all, nor even necessary. The mistake that companies are making is in thinking that this is a data issue and that the competencies required are advanced statistics, data manipulation, and advanced modeling. Consequently, they are hiring incredibly gifted individuals in these areas with PhDs in Computer Science or Statistics, who lack the requisite domain experience, knowledge of consumer behavior, and understanding of how channels work. Knowing what questions to ask comes from years of exposure to marketing and business situations, and if you don’t know what questions to ask to improve marketing optimization, all the data, tools, and technology in the world are not going to help you. Organizations really need to examine more clearly exactly what they’re trying to do, and what their existing culture and organization structure, and internal processes are first, before they assume that hiring a Data Scientist is the silver bullet.