KNIME Spring Summit speaker interviews: introducing Benjamin Spiegel of MMI Agency

November 19, 2015

Registration for the KNIME Spring Summit 2016 has just opened. To give you a taste of what’s to come, today we are publishing the first of a series of interviews with the speakers invited to present at the Spring Summit. We wanted to know just what drives them in their area of work or research and what their thoughts are on topics like data analytics, predictive analytics, the big data landscape and the internet of things.

We start today with an interview with Benjamin Spiegel of MMI Agency.

Benjamin Spiegel is the CEO of MMI Agency – a Houston-based brand activation agency that has been serving Fortune 500 clients since 1986. An industry veteran with extensive experience in advertising, media, data and technology, Spiegel has developed highly successful marketing campaigns for numerous global brands. Spiegel will be speaking at the KNIME Spring Summit this February, and he recently sat down with us to discuss his passion for modern-day data analytics.



KNIME: How did you get involved with data analytics?

Spiegel: To be honest, my involvement was born out of necessity. Math has never been my strong suit and it was really the changing landscape of the marketing realm that led me to data analytics. When marketing started to become more digital, clients had more and more questions such as: “What is the right strategy?” “How do we measure success?” etc. The majority of these questions had to be answered with data. As I started working with larger brands, the complexity of data increased and so did the questions being asked. This led us to look past the number of clicks and focus on the individuals behind them. Today, this is a standard process; marketing, data and analytics are now inseparable.

KNIME: How did you encounter KNIME?

Spiegel: One of our favorite CPG clients introduced us to KNIME. The global digital realm was evolving quickly and it was getting difficult for them to gather and analyze all of these digital touchpoints. In addition, media had become so fragmented that marketers now had to normalize metrics such as Pins, clicks, video plays and impressions. Standard, out-of-the-box tools were failing to accommodate this vast array of metrics, so we started to look for a scalable solution that would allow us to consolidate global metrics, KPI, etc., and allow easy access to data for the various departments.

KNIME: Why did you decide to adopt KNIME for all your data analytics projects?

Spiegel: KNIME was great for us because rather than having a one-dimensional tool, it provided us with a platform for data exploration and innovation. In our current work, the fluid and flexible nature of KNIME allows us to accomplish multiple tasks in a number of different ways. Because it is open source, it also has introduced us to a diverse community of developers who are constantly adding useful plugins and nodes. We feel that KNIME is unique because users can focus on what they want to do rather than how to do it.

KNIME: What is the most promising data analytics application for marketing?

Spiegel: For us, one of the most exciting applications is forecasting and prediction. Consumer behavior has shifted: long gone are the days where marketers can broadcast the same message on a single platform and reach every customer. People today are far less receptive to advertising. This is why data analytics has become so vital. It allows us to better understand our consumers and target them with sequential messaging along their journey that will ultimately enable a conversion. Now, we are working on using KNIME to help us predict future movements of the consumer journey at an individual level. This will hopefully allow us to eventually reach the audience with relevant content before they even know to begin looking for it.

KNIME: What is the most enlightening result you have ever found from a data analytics investigation?

Spiegel: The one that comes to mind relates to one of our entertainment clients. In their previous movie launches, they were only able to measure the point of diminishing results after spending money on media. So despite the company continually spending money, they had already fully reached their potential in audience awareness and ticket sales. They challenged us with predicting the exact point where awareness was achieved so that they could limit their budget and basically spend money more efficiently. Our team immediately started looking at leading indicators such as trailer views, IMDB activity, social engagement and more until, suddenly, we had hundreds of indicators. KNIME was vitally important in this project because it was able gather data from a variety of different touchpoints and identify the exact point where returns began to diminish. Based on this, we created a model that could inform media buying and improve efficiency for all of the company’s future releases.

KNIME: What is the most challenging analytical problem to solve?

Spiegel: One of the most complex issues we face is knowing how to retrieve data from different platforms, in different formats. This means trying to combine search, social, display, video, etc. with different metrics to tailor a strategic approach that will suit the client’s needs. Consumers are always looking for new platforms to use and it is up to us to find ways to normalize, manipulate and integrate this new information with the existing data we have on that consumer. The difficult part is taking all the nuances of these platforms and synthesizing the data into one story that can show us which factors are driving conversion.

KNIME: What is the most complex KNIME workflow you have developed so far?

Spiegel: With the rise of social media, more and more brands are trying to understand how their consumers are talking about their category, what motivates them to buy, act or engage. While social listening is an attractive idea, the reality behind it is much more difficult. 90% of social conversations are just noise. One of the more complicated workflows we built in KNIME was a project involving 70 million social conversations in the beauty category. We had to build advanced linguistic models and intelligent filters to mine millions of lines of conversations to better understand the true consumer passion points, which could in turn, inform the client’s marketing and product strategy.

KNIME: What are the most frequently requested use cases for predictive analytics in marketing?

Spiegel: In our experience, the most frequently requested use cases involve digital marketing’s impact on physical world activity. As an industry, we are still searching for a perfect answer, but agencies and technology companies are making huge strides in proper attribution. By better understanding the customer’s path to purchase and analyzing the data from each step, we can better understand how online impacts offline. 

KNIME: Thank you very much Benjamin, for this insight into your field of work. We look forward to hearing more from you in Berlin at the Spring Summit 2016.

What are you looking for?