Purchase Intent, a measure of the probability that a consumer will purchase a product, is being used by marketers to identify the most meaningful actions that their customers take online en-route to a purchase. A recent Cardinal Path webinar outlined the opportunity for organizations to use Predictive Analytics and Purchase Intent findings to hone in on the specific actions that are most likely to lead to a conversion, either online or offline.
Dave Booth, Cardinal Path’s Co-Founder and Senior Partner, was joined by Katie Birmingham, U.S. Cellular’s Digital Analytics Manager, to discuss three recent, real-world use cases of Purchase Intent research including Katie’s own findings for U.S. Cellular. You can view the on-demand webinar here.
We received so many great questions from attendees that we ran out of time to answer them all! We had some Cardinal Path experts take the time to respond and thought it would be worth sharing the Q&A, below.
Q: What tools in addition to basic web analytics are needed or useful for predictive analytics?
A: There are two sides of a technology stack that you want to look at when you start to build out models. The first set of tools is for data collection. Figure out what kind of data you’re getting, and from where. For example, if you’re going to be using customer data you’ll probably want to have a CRM in place; if you will be using data from your website then a solid web analytics implementation is necessary. The second set of tools will be used to model; you will need tools like SAS or R. So it’s possible to get started with a number of free tools; you can use the free version of Google Analytics, there are free versions of CRMs, R is a free tool, and you may have a lot of data in systems that you already have paid for.
Q: How do you define the long-term value of your users?
A: Every organization will decide what’s best for their unique needs, but ultimately a customer’s lifetime value is typically defined as an estimate or prediction of the net profit an individual will contribute over the length of the customer relationship. Being able to predict the kinds of products and services that are bought, the frequencies they’re purchased at, and the probability of a customer churning all go into customer lifetime value.
Q: How do you evaluate time with brand (also known as “dwell time” in other circles) when you’re evaluating the multi-channel customer journey if a customer is exposed to different types of micro-conversion triggers?
A: Engagement scoring is a great way to incorporate signals that represent users consuming your content and engaging with your brand and messaging. This can be done across multiple touchpoints and devices by assigning anonymous UserID’s as known, logged in, or otherwise identifiable users interact with different digital assets across different platforms and devices. The engagement scoring algorithm(s) or calculation(s) will be unique to each organization, and can serve as inputs to models.
Q: For data collection with google analytics: is it necessary to implement the user ID feature in order to get analyzable atomic data?
A: It’s certainly helpful ;-). The GA UserID feature of Universal Analytics allows you to assign an anonymous but consistent identifier to known visitors that allows you to track what users are doing (anonymously) as they cross sessions, browsers, and devices. You can even see interactions with mobile apps and web properties, so it’s extremely valuable in understanding the various touchpoints users are exposed to and interact with. That said, the UID feature is not the only way to accomplish this. Using custom variables set to (still anonymous) customer ID’s can then be extracted in tools like BigQuery (with GA Premium) and then merged back with customer data sets to enrich customer profiles and ultimately serve them better.
Q: Can you please elaborate on Mobile data analytics?
A: Generally, web analytics tracks what’s happening on websites. What brought visitors there, what they did while they were there, if they converted on goals, the paths they navigated, if they bought anything through e-commerce…those kinds of things. Websites are accessed on a variety of devices, including desktops / laptops, tablets and mobile phones. When we talk about mobile app tracking, this is referring to mobile applications that are downloaded and run on a mobile device (tablet or mobile phone). So when you go to the iOS or Google Play Store and actually download an application to run, it’s not tracked the same way a website is. For example, instead of pages, you have screens. Tools like Google Analytics and a host of dedicated mobile tracking platforms can be configured and used to track interactions on these applications.
Do you have a question you’d like to pose to the Cardinal Path team? Feel free to contact us.
View the on-demand webinar Purchase Intent: Understanding Customer Behavior to Drive Sales.
