It happens all the time. Your favorite grocery store rearranges items so now your morning cereal is right there where you can reach.You get a promotion email. Amazon recommends three items you could potentially buy.
These are all examples of companies utilizing next best actions, offers and products strategies. These strategies have an enormous influence on both customer retention and completed sales, and they tap technologies such as artificial intelligence (AI) and machine learning (ML). The future, however, is for them to evolve and allow next best experience.
Next best action
Next best action involves the question, “What next step should we take for the customer?”
The answer might involve a representative reaching out by phone, sending out a coupon or adding items for sale. It usually attempts to identify and eliminate the biggest or most immediate hurdles the customer has so that they convert.
Related: How to Nail Every Type of Outreach
Next best offer
Next best offer asks, “Which is the offer that is going to help and appeal to the customer the most right now?”
If you have a coffee shop and a customer always orders their “usual” drink, next best offer might mean offering them a coupon to upsize their next order. If a different customer buys something different every time, next best offer could mean a coupon to try a new menu item for free with the purchase of something else.
Next best product
Next best product identifies the items from your company’s catalog that would be the most logical for the customer to buy based on their interests and purchase or browsing history.
Someone who just bought a bike might also want to explore helmets or other gear like a basket or handlebar smartphone attachment, while someone who just bought pasta might also want sauce. Amazon and other companies use recommendation engines within this strategy to show you additional items at the bottom of the pages you view.
Using predictive analytics to get personal
Companies have used these three approaches to get customers to buy and stick around since virtually the beginning of time. The difference today is that businesses can have thousands of customers instead of just a few dozen or hundreds. It’s impossible to monitor everything customers do and interact with everyone manually to learn about them in the digital environment.
In some instances, the solution is to find a broad tactic you can apply by default at scale. Probably the best instance of this is McDonald’s famous line, “Do you want fries with that?”
This line got many people to spend more when they ordered something to eat or drink. The problem with the McDonald’s line is that it failed to recognize that customers are unique, thereby missing the opportunity to offer people other things they were more likely to want and buy.
This is where AI and ML come in. Today, it’s almost possible to collect such a large amount of data on customers, such as age, how much they usually spend, whether they have kids or even what their IP address is.
If you place all of this data into one central location (a “data lake”), you then can use AI and ML to segment your customers and make predictions about what they’ll want or need. The system can alert you that it’s time to take some sort of action and even recommend what might be best to do next based upon parameters you set — such as the customer not buying anything for a certain amount of months. Everything can be highly personalized in real-time so that you literally and figuratively can meet customers where they are.
Personalization matters first because it reduces the odds that people will feel bombarded or mistargeted. 80% of customers are more likely to buy if the business provides a personalized experience, and 42% are frustrated by non-personalized content. But good business is not just about a single campaign or purchase; it’s about the larger customer lifetime value (CLV).
Many personalized activities collectively build trust over the long haul and prove to the customer that you can consistently meet their needs and desires. When the customer feels secure in that, they become loyal. This means that you can stop spending so much time trying to find new buyers.
With this in mind, the ultimate goal for you as a business is to bring all three approaches — action, offer and product — together to routinely provide the next best experience.
CLV is the primary metric that measures this. Doing this requires shifting your execution mindset from inside-out to outside-in (a shift from thinking about what benefits the company, to what benefits the customer), and it requires addressing the entire spiderweb of your operations in a more long-term way.
From a technological standpoint, this requires a higher level of customer analytics maturity as well. Whatever system you use has to be able to evaluate larger amounts of information from more fine-tuned angles and with greater consideration of potential ramifications. It also has to be flexible enough to accommodate the rapid shifts that appear within the market that might make specific data suddenly more or less relevant.
Let’s take a look at what implementing next best experience might look like in a practical, everyday scenario.
- Gather your structured and unstructured data into one location (i.e., create a data lake or “single source of truth”).
- Create technological tools such as widgets that allow you to work with your data lake information easily.
- Gather insights from your data using your technological tools.
- Engineer rules that will guide your sequence of decisions and actions. This is essentially gathering and analyzing the data, coming up with good if-then standards to apply to the information and then moving forward based on those standards. You might send a text to a customer if their email open rate drops below a certain percentage, for instance.
- Take action according to your established rules. Design those actions based on data. For instance, if your data tells you that younger customers prefer digital tools, then whatever your next option for or interaction with the customer is, you’d probably do best to implement it on a digital platform.
- Verify whether that action was effective. If it wasn’t effective, tweak what you’re doing and develop a new standard to apply. There can be so many different elements to consider in a customer journey; reinforcement learning, which is a specific type of machine learning, can run experiments that help you determine what rules might be best based on conditions that are met.
Through this process, remember that you’re always optimizing for customer lifetime value. Although one option or set of options might provide good short-term, immediate benefits, you’ll show preference for the alternative option or set of options that offers a greater long-term advantage.
Take a more sophisticated analytics approach to deliver outstanding long-term satisfaction
Right now, predictive analytics can assist companies with all three strategies — action, offer and product. These strategies center on providing the right item to the right customer at the right time, and each can improve the relationship that customers have with your business in the short term.
However, the big prize is to use customer analytics to always provide the next best experience. As margins of differentiation become increasingly slim, it’s that overall experience that customers are going to use to decide what to do. Examine how AI and ML supported analytics can move you into this next stage so that, when customers think about being satisfied, your brand stands out.