The Marketing Strategist:
From Raw Data to Raw Power: Why You Need a Marketing Analytics Engine
In this interview with Judah Phillips, partner at Knowledgent and author of Building a Digital Analytics Organization, Digital Analytics Primer, and Ecommerce Analytics, he gives us a preview of his talk at ITSMA’s upcoming Marketing Vision 2015 annual conference, November 2−4 in Cambridge, MA.
ITSMA: Thanks for taking the time to speak with us. Tell us what you’re planning to discuss in your conference talk.
Judah Phillips: My talk is about tips for building a marketing analytics organization. I’ll provide some guidance and insight I’ve gained from working with executive and line-of-business leadership at a number of Fortune 500 companies and their analytics teams. These ideas can accelerate the creation and evolution of an analytics organization. Expect to hear helpful suggestions around people, process, technology, and outcomes that you can action when you return to work.
ITSMA: What do you consider to be the key foundations of a good analytics function?
Phillips: Alignment with business goals and outcomes. Where a lot of companies go wrong is not focusing on the business questions they want to answer and instead concentrating on deploying technology to collect, store, or process data. Technology is critical and necessary, but it’s only overhead unless you’re working to answer important business questions that help guide decisions and inform business actions.
When I begin an analytics engagement, one of the first steps is interviewing key stakeholders to understand their requirements, frame the data required, and create an analytics plan that aligns to their goals. That necessary precursor work, where people talk to people to contextualize analytics, helps us put the right focus on what we’re trying to do with data and analytics.
ITSMA: So it’s not just a question of hiring a data scientist and implementing some tools?
Phillips: A lot of hype continues around big data, data science, and the ability to use advanced analytical techniques to understand and drive performance. Companies have made investments but many of their programs are still aspirational. Other companies may recognize the potential value of model-based or algorithmic approaches to analysis, but they don’t know where to begin. Some haven’t made that critical connection to why they’re doing this or what it’s going to achieve.
Another difficulty is finding people with the right combination of skills for your business. Just having someone who has a background in statistics, math, or applied methods isn’t enough. They also need to understand the business context to inform their analysis. They must have the right level of seniority and experience. Companies may not benefit as much from delegating this work to a junior person who just completed a new data science program as they would from training existing business analytics staff on newer approaches to analysis.
Analytics isn’t something that just happens because you hire some analysts and put them in a line of business and tell them to go. It happens based on deliberate action around a business question—and requires a multidisciplinary team in most cases. You may need a data engineer, a data scientist, a data visualization person, and an analyst working together under solid analytical leadership in way that aligns with the business and is supported by technology.
ITSMA: Are you saying that the analytics function should go beyond marketing to be a separate function?
Phillips: Ultimately, yes. I am a proponent of centralized, autonomous analytics teams run out of the business led by a business-savvy chief analytics officer or chief data officer. Marketing can be a great place for that function to start. An effective analytics function is often a centralized one that unifies data from across the company. Whether the data is big or small, the goal can be all data and unified data to power analysis and outcomes. Centralized analytics teams can work with (and even act as) a data steward and participate in or lead data governance as well as, of course, performing the analysis on the combined data sets. The analytics team should be able to look at the entire business and give an opinion based on data that the business leaders can use to make decisions with confidence. Omnichannel analytics is by default omni-stakeholder.
Marketing collects and has access to a lot of data and spends a budget where the effectiveness of that spend can be tracked, measured, and understood. The objective of the analytics is to improve business outcomes and using analytics to understand marketing effectiveness is a good idea.
ITSMA: So you think marketing is in a good position to lead the charge in establishing a credible analytics function?
Phillips: It depends on the company, but it can be a good place to start because improving marketing effectiveness is tied to financial performance. The best way to do that is to ensure the financial impact of marketing is identified in analytics. It’s not just conversion rate, it’s also the net profit per marketing-generated conversion.
ITSMA: Thanks, Judah. You’ve given us a lust for the analytics life. We can’t wait to hear more at the conference!