CrowdTwist Chief Customer Officer Emily Rudin recently hosted a webinar with Emily Collins, senior analyst at Forrester, titled, “How to Build Loyalty through Predictive Analytics and Personalization.”
Loyalty360 talked to Rudin to find out more about this compelling topic.
What role does machine learning play in refining predictive models that can lead to stronger brand loyalty?
Rudin: Machine learning enables brands to refine predictive models by:
- Creating an elite tier of a brand’s best performing customers based on an algorithm that includes commerce metrics (ACV, frequency) AND advocacy (e.g., social sharing). Brands can then target these customers with exclusive offers to show them appreciation and acknowledge their status.
- Optimizing the shopping cart experience based on previous behaviors and preferences.
- Mapping a customer journey that predicts the behaviors of new shoppers and creates a path for repeat purchases that replicate those of a brand’s best customers.
- Curating personalized content and messages based on algorithmically generated customer segments to increase conversion rates.
- Creating look-alike audiences that resemble a brand’s best customer segments and advertise to them on various channels.
What are marketers doing well to build loyalty through predictive analytics and personalization and where do the challenges lie?
Rudin: Marketers are beginning to fully realize the value of making customer data actionable. Customers today have high expectations and they expect a company to give them relevant experiences based on their interactions with the brand. Marketers today understand that they need to go beyond segmenting and personalizing customer experience based on transactional data. They also need to capture how their customers engage with the brand.
That’s why we are seeing more and more brands seeking a holistic view of customers across all channels. This is a great starting point, but the challenge then often lies in what happens to this data after it’s been collected. What’s been done to turn this data into experiences that deepen customer loyalty? The challenge behind this is that most brand marketers have access to “oceans of data” but only have “puddles of insight.”
What are some best practices related to this theme for loyalty marketers?
Rudin: No. 1 is to invest time in understanding the nature, consistency, and completeness of the data. If companies don’t understand the data, they will likely misinterpret the results provided by applying predictive analytics. Second, challenges arise when a model fit on one set of data might not yield results that generalize to other data sets. I would recommend that brands perform validation testing and learn on their models and data sets. Third, I recommend that marketers accept and plan for errors from bad data or unexpected results.
How do you build rich first-party profiles?
Rudin: You need to connect the dots for your customers’ interactions with your brand across the entire ecosystem. This can best be achieved by having an omnichannel loyalty solution in place that tracks and rewards activities across websites, mobile and social media so that your brand is everywhere your customer is. Create, capture, and reward for every interaction with your brand, whether it’s a transaction or engagement activity such as survey fills, polls, social interactions, or content sharing. These activities help to build a 360-degree view of who your customer is.
How do you know you’re offering relevant experiences?
Rudin: We recommend that you isolate a subset of your customer base and apply a test and learn strategy to your personalization initiatives. You need to set KPIs for engagement, frequency, and customer lifetime value, and measure the impact of your efforts. If your goal is to increase engagement and your members are returning to your website, social media profiles and referring their friends, you are offering relevant and memorable experiences. If your goal is to increase sales, you know you’re offering relevant experiences if you’re seeing not only an uptick in previously purchased items, but an increase in new business, upsell, and cross-sell transactions.