Using Social Graphs to Prioritize Marketing Spend

It may not surprise you that data science and artificial intelligence are the new trends changing the marketing profession. This is a response to having access to more data than ever before and consumers increasingly become impervious to marketing efforts. Data and AI can help re-establish the relationship with the customer and make marketing more effective. More and more, data insights are driving how we make marketing decisions.

All marketers have to answer the same questions:

  • Who is my consumer?
  • Where do they seek information about products?
  • Who are my top competitors?

The answers provide clues to a more practical question — how should I spend my marketing dollars? This important question is historically answered by using a combination of past experience and intuition. This is not exactly a science. So what happens when you are expanding to a market where you have no precedent, or looking to connect with a new consumer? At KUNGFU.AI we worked with a consumer product brand to connect with a new customer using Social Graph AnalysisⓇ — a technique that can bring value to any B2C marketer. It harnesses social media data on target consumers and can provide tremendous intelligence on how to spend marketing dollars effectively.

A CPG Case Study

Karbach Brewing was tasked with a common challenge. They are expanding to a new region and targeting a non-traditional consumer. There was no precedent, and therefore their intuition was limited. Purchasing the data on this consumer was expensive and there was not a lot of useful data to be found. The CMO needed to create a marketing strategy and budget, placing bets on where his dollars will result in the best returns. He had several key questions around event marketing, a primary strategy for the company:

  • Am I spending my event marketing dollars wisely?
  • Which events should I spend my event marketing dollars on?
  • What are my consumer and competitor behavior and preferences?

Historically, event marketing was proving to be very effective, but finding the right customer at the right event for this new initiative would be like throwing darts blindfolded. Data and social graphs turned out to be the key.

Unstructured social media data is a wealth of intelligence for all consumer brands. Social media is where we express ourselves, connect with family and influencers, follow interests, and hop on new trends. Brands learn a lot from their followers, but you can learn a lot from those who don’t. And how do you anticipate who would be a follower? A social graph draws an edge between you and the people, places, and things you interact with online.

 

Social Graph

 

A graph helps you understand how people are connected to their attitudes, behaviors, and one another. For example a graph may inform you that your followers also like motorcycles, Converse, Stage Coach, fine dining, and follow Taylor Swift. If you wanted to find new customers, you could use the social graph of your followers to their connections and correlate others who have similar characteristics.

For our client, we developed a graph database built on a proprietary data set of similar companies, local competitors, events, and social media accounts to understand consumer and competitor behavior and preferences.

The proprietary customer social data set we created included:

Local Competitors

  • 761 local competitor handles, pulling larger national brands down to local brands in the seven state region they were targeting

Events

  • 835 Twitter handles from all events in the seven states

Users & Relationships

  • 6 million Twitter users that are followers of the events and/or competitors
  • 11 million ‘following’ relationships from these 6 million Twitter users

We analyzed some of the data using a technique called Jaccard Similarity. It’s a measure of similarity for the two sets of data, with a range from 0% to 100%. The higher the percentage, the more similar the two populations.

 

Jaccard Similarity

The data sets and analysis allowed us to understand which new regional events were most like the ones favored by current consumers in other states. These events would be the best place to spend their event marketing dollars. Lastly, our team built an interactive Consumer Social Graph Insights tool that Karbach can interrogate for quantitative answers to a myriad of marketing related questions.

Social data provided a path forward to reaching a new consumer and market. Data was the first step. By understanding how social data is closely connected to other people and behaviors, insights start to emerge. A social graph and similarly analysis provided concrete answers to an important marketing question; where do I spend my money to maximize return?