Back

How to Scale 1:1 Personalization in Less Than 60 Days With Machine Learning

July 22, 2020

How to Scale 1:1 Personalization in Less Than 60 Days With Machine Learning

Today’s marketers understand that customer segmentation is becoming a tool of the past, and that transactional offers don’t deliver the relevance that customers now expect. But they may be unsure how to replace these tactics with a personalization strategy. If this is a problem your marketing team is currently facing, the answer is in your customer data—but extracting it is easier said than done.

Your brand is most likely already gathering customer data from multiple sources: transaction data from your loyalty program, online shopping habits from your ecommerce platform, behavioral data from your customer data platform.

But gleaning actionable insights from all of this siloed data is difficult. Your company may have a data science team to help, but they typically aggregate data and use it to create propensity and spend models. These models identify trends, but can’t help you create relevance for customers at the individual level. Developing more personalized messages, offers and rewards will require another look at your MarTech stack.

Optimizing personalization with machine learning

Adding Formation’s machine learning (ML) solution to your Martech stack will make your data more actionable and help you better use it to implement 1:1 personalization. Our machine learning engine analyzes the data from all of your data sources, but instead of just identifying trends, it uses that data to uncover insights about each individual customer, then develops individual offers based on what it learns. Our solution also connects insights to your marketing channels, enabling you to send offers everywhere you connect with your customer.

Formation’s solution is unique in that it combines ML and AI to make offers more personalized over time. Our algorithms are continually running tests, analyzing results and learning from new data in a process we call multi-step behavioral offers. As you may have guessed, these offers require multiple steps to complete—and as each step is completed, the system gains new behavioral data on the customer, adapting future steps, offers and rewards based on what it learns.

How machine learning and AI create multi-step behavioral offers

Multi-step behavioral offers are designed to benefit the customer and the brand in equal measure, because they align the goals of both parties. For example, if a customer prefers to buy coffee on weekday mornings, and a coffee shop brand wants to increase incremental revenue, the ML and AI engine may design an offer that rewards the customer every time they purchase an add-on item with coffee on a weekday morning.

As the ML engine learns the customer’s preferences, subsequent offers are optimized accordingly. For example, the engine may learn that our hypothetical customer has tried breakfast sandwich add-on items on 5 occasions, but a dessert item on only one. The AI would then update the offer to reward the purchase of breakfast sandwiches, while eliminating the rewards for dessert items.

For customers, this provides a high level of personal relevance, because the offer speaks to their specific motivations. It also provides customers value by leveraging their personal preferences, keeping them in the customer loyalty loop so they keep returning to the brand. For brands, this reinforces the customer behaviors that they find valuable.

Over time, the cycle of learning and optimization enables our ML and AI engine to make offers increasingly more personal, relevant and effective. Machine learning allows Formation to prompt more of the customer behaviors that brands value, but also to change customer behavior over time so that it becomes more valuable.

Scale 1:1 personalization rapidly

Formation’s combination of ML and AI allows for fast offer scaling, because it can individualize offers for every customer regardless of the size of your customer base. That means you can quickly reach millions of offer combinations that are completely tailored to individual customers, fully replacing your segmentation strategy with 1:1 personalization.


To deliver greater relevance to customers, you must act on the data from those customers. Adding Formation to your MarTech stack enables you to do this in an automated way, giving you greater ROI from your stack and your strategy. Learn more about how the next evolution of marketing calls for the right technology in our white paper “The Future of Marketing: Relevance + Value.”