TechBytes with Jordan Cardonick, Director, Analytics and Decision Support, Merkle

Jordan Cardonick Merkle
Jordan Cardonick

Jordan Cardonick
Director, Analytics and Decision Support, Merkle

Marketers have to stay focused on personalization, segmentation, optimization, and automation at various stages of the customer lifecycle. It all starts with the ‘truth’ in data and then measuring it over time. To dig deeper into how Merkle tackles data performance, and the use of AI/ML in data modeling/data mining, we spoke to their Director of Analytics and Decision Support, Jordan Cardonick.

Tell us about your role at Merkle and the team/technology you handle.

I manage data and oversee analytics to ensure the financial success of our clients’ media campaigns. I’m responsible for helping clients evolve their KPI selection and data enablement by managing multiple teams that deliver day-to-day analyses, forecasting and scenario planning, budget and financial management, and marketing insights.

How would you segment data based on the contemporary marketing and sales demands?

Given that Merkle’s focus is primarily from a media lens, we develop segmentation around the various attributes of media campaigns, which may include standard elements of audience attributes. We use this information to guide our media optimization and insights, recommending more efficient ways to manage that media moving forward. Additional common segmentations include things like brand vs. nonbrand, creative size, geo publisher, and client generated segments like product/service offering and customer type.

At Merkle, how do you verify the authenticity of data? How does it fit into data orchestration platforms?

Data is the foundation of what we do at Merkle. There is a high level of scrutiny placed on ensuring that data is properly loaded, maintained, stored, and protected. For as much access to digital data that we have, we are not able to dictate how platforms attribute/characterize data. The best example of this was in the early days of the internet: Most data generated by AOL was centralized to its main server farm despite actual customers locations. So even now, if a device type or geo may be wrong, we need to still play by the Googles and Bings of the world’s rules. Based on rules, we determine the best strategy to target and manage data when the platforms don’t conform to one another.

How would you define personalization, segmentation, optimization, and automation at various stages of the customer lifecycle? How do you manage customer experiences at each level of this journey?

The answer is in the goals and objectives of the client and what that brand hopes to accomplish at each level; and what access to data there is at each step. For instance, at the awareness stage, we may segment to find new audiences vs. look-alikes of the people who eventually buy. Building automation around look-alikes for the model and optimizing around those qualities and engagements is then more traditional and streamlined. For net new characteristics of potential buyers, i.e., selling a new product to a current market or entering a new market with current products, we may need to design a long-term test or longitudinal study to see what the downstream impact of that is. This makes automation and even personalization difficult because we don’t know what quite to expect yet.

Why should brands leverage data optimization software for marketing campaigns?

I recommend companies stay away from pre-built software that is ‘out of the box’ at first because it may not consider all of the nuances of that business or in the data itself. Out-of-the-box software packages work well once you have an initial foundation and understand your data. I do suggest software like Tableau or Python that facilitate data analysis if there is a proper team in place to do so.

How do you leverage AI/ML and data science at Merkle? Which AI companies are you particularly interested in?

While AI/ML in their current iteration are areas that we are continuing to refine and develop services around, these elements have been key to our business in recent years. We’ve built our own bidding technology that considers many characteristics, rules, and models to populate the engines with the right bid at the right time. Modeling and mining the data are crucial to helping us uncover opportunities and efficiencies for the campaigns that we are managing. Companies like Amazon, IBM, and Microsoft are so interesting to follow in terms of the potential future for AI, but there are a lot of companies that are doing cool work like Uber and the self-driving cars and all the cool behind-the-scenes work that Netflix does to promote what shows I might want to watch.

What is the one message for 2018-2022 that you want to share with other analysts and researchers in the marketing and sales technology industry?

Data is going to continue to explode and come from all sorts of places. Don’t get overwhelmed by all of it — just make sure you take a step back and think about the “so what” before diving in.

Thanks for chatting with us, Jordan.
Stay tuned for more insights on marketing technologies. To participate in our Tech Bytes program, email us at news@martechseries.com

MTS
Previous ArticleNext Article

Leave a Reply

Your email address will not be published. Required fields are marked *