A little more than half of marketing decisions are influenced by marketing analytics. That means a little less than half of marketing decisions are influenced by something else.

Marketing decisions without data? Without analytics? Experience — and gut instinct — actually still matter here?

“I think that plays a large role,” said Lizzy Foo Kune, analyst at Gartner who shared those findings with CMSWire from the Gartner Marketing Data and Analytics Survey 2020 released in September. “In the drive to be more data-driven, I think we've also cast aside the value of experience. There is a certain element of using past knowledge to inform future action even if that isn't grounded in hard data. I don’t want to say that gut-feeling should be used in all decisions or that experience should overwhelmingly support the actions that an organization is taking.”

Poor Data Quality, Lack of Clarity on Recommendations

Foo Kune knows this all too well as a former marketing practitioner: you can likely count on some rolled eyes in the marketing Zoom meeting when you say, “This worked in the past.” Someone’s inevitably going to ask for the data behind that successful campaign and request some new data supporting the thesis why it will work again.

But here’s the problem: marketers’ mistrust for marketing analytics is growing, according to the Gartner findings. CMOs say they are prioritizing investments in marketing data and analytics, but others say poor data quality, inactionable results and nebulous recommendations make them leery of marketing analytics to drive decision-making.

“If your marketing analytics team isn't giving you really clear recommendations and a way to take action, then we're finding that marketing leaders are just saying there's no reason for me to do anything else,” Foo Kune said. “There's no reason for me to change course. I can't trust the data. I don't know what I'm supposed to do with this data, so I'm just going to keep doing what I was doing anyway.”

Related Article: Why Prescriptive Analytics Is the Future of Marketing

Communication Must Improve

Foo Kune cites a lack of communication as one big underlying problem. Analytics teams and senior stakeholders need to do better here. Senior leaders are the ones who have a greater mistrust of the marketing analytics: 54% of senior level marketers said marketing analytics hasn’t had the expected influence while only 37% of mid-level marketers said the same thing.

“We need to realize that some of these senior stakeholders have the responsibility to communicate better with their analytics teams,” she said. “And a lot of times the analytics teams lack the business context that they need in order to guide their analysis. If we're going to use this data to drive change, not only do we need to ensure that some of these senior stakeholders or decision makers are more confident in the data quality and receive the actionable recommendations that they need, but we need to make sure that the analytics practitioners have the context that they need to guide the analysis.”

Are Marketers Just Too Darn Proud?

The number one reason why marketing analytics is not used in informing decisions? The data findings conflict with the intended course of actions. In others, marketers had a hypothesis, a plan. The data told them it was a bad plan, so they ignored the data and went with their original plan of attack.

Marketing leaders often seek out data to “help them make the case for a desired course of action, cherry-picking data to make them look good,” according to Gartner researchers. How very narcissistic.

Seth Earley, author of “The AI-Powered Enterprise” and CEO of Earley Information Services, recalls an organization he worked with which began to adjust use cases to favor the preferred design and game the results. Initial testing showed that a new approach would not be as effective, and the organization ultimately decided against further testing since the data would not support their desired direction.

“What? Wait a minute. Isn’t the purpose of testing to take opinion out of the decision and base things on actual data?” Earley asked. “Apparently not. Analytics can yield insights but if the answers are not what stakeholders want, then there is nothing that the analytics can do.”

He calls this an example of "culture in action." Peter Drucker has said, “culture eats strategy for breakfast.” “My variation,” Earley said, “would be ‘culture eats analytical insights for breakfast.’ Assuming that the culture is actually one that appreciates data, facts and the truth about the situation, then we can move on to the tactical, technical and logistical constraints.”

OK, ego is not the only reason marketers may not turn to analytics. Gartner also found analytics are not used to inform marketing decisions — in addition to what we’ve already mentioned — because:

  • Decisions are driven by trading/promotional calendar.
  • Analysis does not incorporate different sources of data.
  • Analyzing data takes too long.
  • Analysis does not account for business context.
  • Lack of access to sales or conversion data.
  • Analysis is too difficult to understand.

Related Article: Is Your Marketing Truly Data-Driven?

Path to Meaningful Insights

The biggest challenge in marketing analytics is compiling a complete set of accurate data for all the different campaigns, whether they be online or offline, and compiling the data in such a way to give meaningful insights, according to Bonnie Crater, CEO of Full Circle Insights, which helps marketers measure campaign effectiveness.

“It's hard enough to analyze the information, but the time it takes marketers to manually pull the information and assemble it from all the different marketing systems is money down the drain,” Crater said. “The problem is, many companies have not fully digitized, preventing marketers from having access to all the proper tools they need to effectively track data. Marketers should demonstrate how digital transformation is improving a company’s bottom line to show the value of these analytics and the importance of implementing the right tools to thoroughly track metrics.”   

Lack of Baseline Measurables

Most decision analytics tells us something about the past. What happened? If we apply human knowledge and expertise, we can understand why something happened, make a change and then determine the impact of that change, according to Earley.

Learning Opportunities

However, one challenge is when there is no measurement of baselines. Baselines tell us the status quo, he said. How does this process function without any intervention? What are the “normal” outcomes? “Some baselines are straightforward because the metrics are continually being gathered such as online sales per region, product category, per product and product variation and so on,” Earley said. “But others are less clearly understood or instrumented.”

If we are trying to understand how online behaviors on the corporate site impacts revenue from distributors, we would run into a big problem if we don’t have clear line of site connections between those variables, Earley said.

“It will be hard to determine how interventions on the .com catalog site will impact revenue from distributors due to the lag in revenue impact,” Earley added. “It is possible to identify signals that predict an increase in revenue, but those signals may be subtle without clear correlation or causation. Of course, correlation does not indicate causation, but we don’t always need to prove that.”

Oftentimes, many processes are not well understood and measuring baselines are not well instrumented. Brands don’t have an easy way of understanding, Earley said, the levers that have an impact or a way of separating performance metrics when there are multiple factors that are not easy to tease apart.

Related Article: 8 Analytics Trends to Watch in 2020

No End to End Understanding of the Value Chain

A big part of leveraging analytics is to build feedback loops to leverage data in ways that impact the user experience, Earley said. Because there are multiple processes supported by multiple, sometimes redundant or duplicate tools or different groups own different parts of a process, it is very difficult to capture the right metrics and apply them to the experience.

“Ultimately analytics are part of an activation process which depends on a full understanding of the customer lifecycle and corresponding supporting internal processes,” he said. “Not understanding all of the components and processes will impede the application of analytics from those components and processes. Analysts need a thorough understanding of the value chain in order to convert analytics to insights about the customer and their experience.”

Converting Insights to Actions

Connecting analysis to insight require the detailed understanding of the various levers that impact user behavior. But even when the insights are clear, Earley said, converting those to actions requires:

  • The ability to communicate insights to the people who can act on those insights.
  • An understanding of which levers to pull to leverage the insight.
  • The digital machinery to automatically present information of interest to the target based on analytics of digital body language.

“Lack of the digital machinery that enables feedback between the customer’s digital body language and the user experience means that insights are acted upon through ‘acts of heroics,’ which do not scale,” Earley said. “Feedback loops need to leverage information from a customer data platform (CDP) that consolidates, integrates and normalizes data from each touchpoint (click throughs, purchases, social media engagement, responses to digital campaigns, etc.). This requires that content models, customer attribute models, knowledge and decision-making frameworks (knowledge architecture), content models and product information models be aligned and orchestrated along the customer journey.”

Lack of Skill Development

Do marketers need to get better at analytics themselves? Gartner researchers also found that skill development ranked lowest among all marketing analytics activities. It was even lower (23% vs. 39%) this year versus two years ago when Gartner researched marketing data and analytics.

Marketers spend most of their time on data management, data integration, data formatting, ad hoc queries and requests, exploring data for new insights, generating reports and dashboards and advanced modeling more so than skill development.

Gartner’s Foo Kune said marketers plan to invest in areas of automation, such as data automation and decisioning automation, meaning activities around data management and ad hoc queries stand to be automated.

“If we're going to be automating so much of this," Foo Kune said, "yet we're not investing in developing the skill sets and developing our talent, what are those folks going to be doing another year or two?”