Loading...

A crawl, walk, run approach to real-time decisioning

This article is an excerpt from Experian Marketing Services’ 2016 Digital Marketer Report. Download the full report to discover more insights and trends for the upcoming year!

When marketers think of real-time decisioning (RTD), they tend to think of cutting-edge programs like predictive analytics and advanced modeling. While these advanced concepts are certainly making headlines, most marketers have a long way to go before they can realistically implement these complex strategies.

The reality is that a lot of marketers aren’t adequately addressing the basics when it comes to using their real-time data. In your current marketing programs, do you collect your data in real time? How deep is the data that you’re collecting? And are you able to connect this data to your other first- and third-party data sets?

Real-time data and decisioning requires a crawl-walk-run approach. The key is to be realistic about the resources you have and your starting level of sophistication. Remember, real-time decisioning isn’t just about being quick. It’s about focusing the right response at the right speed based on the customer’s needs.

  1. Crawl: Collect and connect data in real time: Without real-time data, there is no real-time decisioning. Email data – not only clicks and opens but location, time of day and device type – is often the easiest type of data to collect. But stopping at email only gives you a sliver of an understanding about your customers’ relationship with your brand. That’s why it’s imperative to marry your real-time email disposition data with other data sets, like web activity data, social data, offline data (including in-store, point of sale, kiosks and call-center data) and mobile behavior.

    Collecting and linking real-time data requires an infrastructure to house the data and help you link it to your initiatives. This is the beginning of real-time decisioning. I’m always amazed at how few people either don’t collect data or don’t house it in a place where it’s actionable “at the edge” to the Marketing team.

  2. Walk: Design and test programs using that data: The second phase of real-time decisioning is simply using the data you collect to make decisions. It’s a bit like a ping-pong game. When a customer serves you the ball, you need to come back immediately with the most appropriate response and be ready for their next move. To do that, you have to be prepared for all of their possible interactions and have the right responses ready to go. Understanding your customers allows you to map a response plan to any of their actions at any given time.There are two ways in which you can use real-time data in this stage. The first is decisioning. What do you do when you see real-time data? What is your immediate response? The second is a bit more subtle, and it has to do with how real-time data can inform the communication itself. For example, if you see that a customer just responded to a specific product category in social media, how can you highlight that category in the content of your next promotional or operational communication? By using real-time data, marketers can not only ensure that the timing or cadence of the communication is optimal but also provide the most relevant content to that customer.
  3. Run: Automate and predict using advanced data science: The next evolution of real-time decisioning is about predictive analytics, automation and data science; essentially making instant decisions based on many different variables. This strategy can be used to inform offer optimization, cadence or channel delivery, or to personalize inbound channels. Additionally, advanced RTD is only improved when more advanced data sources are included in your decisioning engine. All channel response data, purchase data, external data sets such as weather or Internet of Things (IoT) data streams all help paint a more accurate picture of a customer. It is recommended that marketers approach advanced RTD with caution. Jumping directly to these niche, advanced toolsets without first mastering the crawl and walk steps may backfire. You might be able to power a really unique, single-offer campaign with an advanced tool, but if the systems, architecture and partnerships aren’t in place to form a sustainable strategy, you’re winning the battle rather than winning the war.

Don’t be intimidated by the overwhelming opportunities in real-time decisioning. You don’t necessarily need to collect all possible data sources to get your feet wet. Take a look at the data you already collect and think about how you may be able to utilize it to create more relevant communication frameworks. If you already collect email data, bring in one or two more sources and use those data points to make decisions about how to react. No matter which data you use first, the process is the same: Find the data source then design and test programs around that data source. As you do, find ways to automate and conduct more advanced programs around them. Then wash, rinse and repeat.

Blog CTA banner_DM16 report