Insight/ Strategic Planning:
3 Steps For Using Big Data To Drive Faster Marketing Growth
CMO.COM | Digital Marketing Insights for CMOs
by John Conley, Vice President, Data Warehousing Dotomi
Most marketers agree that marketing programs empowered with big data could radically improve business results. But understanding big data’s potential business value and realizing it are very different. While many brands have invested millions of dollars to collect marketing information, few CMOs can say they are fully leveraging its benefits.
by John Conley
Vice President, Data Warehousing
The Big Opportunity:
The biggest opportunity to leverage big data is in turning your big target audience into millions of small ones. Using big customer data and available technologies, marketers can actually treat every prospect and customer as an individual. Best of all, that can be done using anonymized profiles that protect consumer privacy.
Insights from big marketing data can be leveraged to create compelling experiences tailored to individual behaviors and preferences. By accurately processing and "actioning" big data, marketers can customize individual messages at remarkable scale.
By taking the time to leverage your company's existing big data assets and following three simple steps, you can deliver personalized marketing that can grow sales this year:
1. Unite offline and online customer data and associate it with individual profiles:
The first step is to unite a brand’s offline and online customer data, associating all of an individual’s actions with a single, anonymized user identity. Bringing together all of this information results in a more complete picture of users and provides an opportunity to learn how to drive more purchases.
Indeed, the ability to combine the most granular SKU-level offline and online purchase information, rewards program records, CRM interaction data, e-store activity, and more will help you discover what a consumer needs and how best to motivate his purchases. Arriving at those conclusions is a complex process and not always easy to accomplish by yourself.
2. Create personalized messages on a one-to-one level and then deliver them across devices at scale:
Once profiles are created, the next step is to make them actionable. Using big data, it is possible to make personalized ads that reflect a person’s interests, behaviors, and lifestyle.
A personalized ad is customized to each unique consumer across many levels. It might reflect the imagery and colors that a shopper is most known to respond to, the product that she is most likely to want, her lifestyle, marital status, and, of course, her brand preferences. The message could also reflect whether that person typically buys in a retail store or online. Personalized creative development can produce thousands, or even millions, of dynamic ad variations and provide individual relevance at an even greater scale.
Personalized ads are also very different from retargeting ads, which feature the last product an individual viewed on a particular site. With personalized ads, two consumers interested in the same watch could see two completely different ads. For the status-oriented customer, the ad could feature the watch in context with edgy, luxury-driven imagery. For someone more interested in quality, the ad might highlight the watch’s time-tested reputation and customer ratings.
This sort of personalization is possible due to recent technology advances that enable rapid analysis of massive amounts of user-level information. Massive parallel data processing is necessary in order to ingest numerous data sources and channels in real time.
Big data can also help optimize when and how a communication reaches the consumer. A brand can message one user across the different devices used throughout the day–and do so in the ad format that is most effective for that particular user. For example, a consumer best persuaded by video can be reached as he browses on his phone at breakfast, his PC during the workday, and on a tablet at night.
3. Create a "best next" personalized communications stream:
We've all heard of "next best." But with personalization, marketers should be more concerned with "best next." Best next means delivering personalized communication streams that are dynamically optimized based on both recent and historic consumer actions, both online and off.
For example, imagine that a customer goes to a store and buys an overcoat. When that retailer prepares to message that person next, it’s likely that they won’t need to buy another new coat. Instead, an ad featuring a scarf and gloves that match the new coat would be more likely to drive an incremental purchase. And not just any scarf and gloves, but the set that best reflects the buyer’s lifestyle and brand preferences.
Our own internal client research has shown that the near-real-time component is critical because the retailer is more likely to sell the matching scarf if the message is sent right after the coat is purchased. But best next isn’t only about related items. Profile analysis reveals exactly the right set of goods and stimuli to motivate the customer’s next purchase, whether or not the item is related to what was just purchased.
Personalization Works To Drive Profits
By associating big data with individuals, marketers can create robust, dynamically updated profiles that drive extraordinary results. Statistically, validated testing across more than 170 leading retailers has found that personalized media programs yield significant incremental sales results. In fact, our research has shown that personalized media can drive 3 to 6 percent of total brand sales.
Also, measurement warrants some specific attention here. When assessing the business impact of personalized media–or any other marketing tactic–marketers need to understand attribution and the total incremental sales delivered. Each of these can be measured only by evaluating both online and offline channels. Incrementality is measured by separating the sales that were caused by the program from those that would have occurred anyway. We recommend using a continuous A/B testing methodology that encompasses online, in-store, and other channels in order to get an accurate read. This methodology uses statistically validated techniques to determine total incremental sales from all channels.
Big data and personalization could be growing your business in a way that maintains the highest standards of consumer privacy. So don’t wait another year to get your data house in order. The opportunity you’re missing might be enormous. The time to act is now.
About John Conley
John Conley is vice president, data warehousing, for Dotomi, a unit of ValueClick Inc.