The Unique Challenge of Audience Targeting
Marketers in the consumer-packaged goods (CPG) industry face a unique challenge compared to other B2C/B2B marketers — they must target everyone who eats, drinks or uses their products — while non-CPG marketers can use tight customer segmentation to target specific groups of people. During the golden age of television (1950-1980) when limited channels meant massive reach of uniform audiences with undifferentiated ads, targeting everyone was feasible, but in today’s world of countless platforms, streaming services, and digital media options coupled with consumer ease of self-service and unsubscribe, audiences are highly decentralized. Heavy-handed “spray and pray” marketing is expensive and ineffective. So how do CPG marketing teams cut through the noise? Audience-based marketing.
Audience-based marketing allows you to follow customers where they go. It’s a nimbler, customer-centric approach based on demographic/lifestyle data — the type of data used for target audiences matters a lot according to Nielsen — to formulate impactful, high ROI campaigns.
Why Identifying the Right Audience for CPG is Hard
But audience based marketing in CPG is not easy today. It requires multiple tools to merge internal and external data and customer IDs across channels, devices, and more — tools for data prep, exploratory data analysis, dashboarding/reporting, modeling in Python notebooks — and once merged, it’s easy to make cross-channel reach and frequency mistakes (think: unknowingly sending the same ad to someone via web and mobile — which gets annoying) because cross channel coordination is difficult. Pulling off a traditional Recency, Frequency, Monetary segmentation is time consuming. Doing it as one person seems impossible, especially if you don’t want to wait for a data scientist to help you or a report developer to create the view you need.
In this article, you’ll see an easier way to accomplish audience targeting in a single analytics platform — Tellius.
* * *
I started with this marketing campaign dataset from Kaggle consisting of sample customer demographic and purchase behavior or wine, fruit, etc. Loading the .xlsx was straightforward.
Connect flat files, cloud objects, SQL and NoSQL, business apps, and more
Once loaded, data exploration is easy. Preview the dataset to get one-click summary stats.
Next, I used Tellius’s “Search Guide” for analysis inspiration. I could look for clusters and segments, top/bottom analysis, and much more.
Contextually aware guided search suggestions
I was curious what product was most popular. I searched for average wine, fruit, and fish via a Google-like search experience.
It knew “wine” meant MntWine. I got back visual, explorable answers.
I noticed a column in the data called AcceptedCmp1, AcceptedCmp2, etc. I decided to create a sum of these 5 columns. Doing data prep was easy via a point and click interface that automatically creates a data pipeline of changes.
Point and click data prep with visual transformation pipeline view example
Next, I used Tellius Guided Insights — a powerful tool that automates artificial intelligence and advanced statistical methods to answer complex business questions in a matter of minutes — to identify the top segments in my dataset by setting the target variable as “Response”.
Tellius Guided Insights is an AI-powered way to quickly go from “what” to “why”
Tellius provided me with drivers of response on the left — Total_Spend, Monthly Web Visits, Meat purchased, etc. — and provided me 5 segments on the right — with a drill down into Segment 1 in the middle which appear to be people whose Total_Spend > $1,918 and purchase more than 4 web purchases who are 4.6 times more likely to purchase than others. As expected, there is a segment of customers who spend a lot and purchase from the web fairly frequently.
Next, I explored Segment 4, who are 3.5x more likely to respond. They turn out to have a very different profile than Segment 1 — frequent web visitors with lower total spend who tend to purchase wine and meat.
Exploring another segment
To look under the hood and understand how the Tellius insight was generated, I clicked into “more info” and saw what model was used (in this case, a decision tree classifier), Evaluation, and Feature Importance.
Under the hood of a machine learning model in Tellius
Often, this is where analysis typically stops due to technical barriers. How can we flag this segment with 4.6x in our data and provide an actionable list of customers to target with offers?
With Tellius, you can automatically apply a flag to your dataset without having to know SQL by using Explore Segment → Smart Insights.
From here, I create a detailed list of target customers. Viola!
Detailed list of people who fit the highest performing segment to target
Audience Targeting Made Easier
Audience targeting has historically been viewed as too technical and too hard for any one person. Legacy tools, multiple teams and communication barriers cause audience targeting to be a slow process with output that is out of date by the time it is ready. With Tellius, I have taken on this challenge by myself and — at the end of the day — by targeting contacts my profit margin should see a lift.
To learn more about how CPG firms can mash up third-party, marketing, and customer data across multiple sources and get diagnostic and proactive intelligence to fuel decisions, check out this 30 minute webinar or give Tellius a spin with a free trial. You won’t regret it.