Long before we lived in a world powered by data and connected devices, our definition of “personalization” was pretty limited.
A personalized gift might mean getting a pair of pajamas with your initials monogrammed on the pockets.
Personalizing your car was all about ordering a license plate with a name or phrase that had some kind of meaning to you.
When you went out to a store or a restaurant, personalization happened when a store associate or server recognized you by name, and remembered what you ordered last time.
As organizational success is increasingly predicated on the quality of the customer experience a brand delivers, personalization has shifted from a nice-to-have to a critical element of business strategy. The problem is that most businesses don’t realize the full scope of opportunity to personalize experiences that lie before them using data-enhanced design thinking.
In its most recent Next In Personalization report, for example, consulting firm McKinsey found that 71% of consumers expect companies to deliver personalized interactions. Even more (76%) said they get frustrated when personalization doesn’t happen.
Too often, however, personalization in this context is focused almost exclusively on digital experiences and reduced to the basics, such as including a customer’s name in an e-mail marketing campaign.
There is obvious and genuine value in personalizing product recommendations on an e-commerce site, or suggesting movies to watch on a streaming TV service. At the same time, though, organizations can’t risk overlooking how the growth of data science and connected devices is expanding what personalization could offer customers today.
Market research firm IoT Analytics has forecast that between now and 2025, the number of connected devices will rise from 14.4 billion this year to 27 billion. At the same time, the promise of harnessing data science has never been stronger: research published by Gartner Inc. earlier this year found almost half of Chief Data Officers are focused on customer experience improvements, and 63% see data and analytics as a way to bolster decision-making capabilities.
If other businesses are failing to prioritize the use of data science and connected devices to personalize customer experiences, it’s usually due to a lack of understanding of the potential use cases and how to get started.
A few examples of real-world applications include:
All these innovations are possible through algorithms, which are used to describe a process that can be accomplished by a machine and replacing the decision making that traditionally required human intervention.
Many organizations may have plenty of data – including both customer and operational data – that could fuel these kinds of personalization opportunities.
Many organizations may have plenty of data – including both customer and operational data – that could fuel these kinds of personalization opportunities. Working with the right partner, they can develop an algorithm to make use of the data. This usually involves the following steps:
Even if you’ve never developed an algorithm before and lack the necessary skills, you probably have the subject matter expertise that will inform your use of data and connected devices. A mortgage company, for example, will have developed extensive knowledge about what drives home-buying decisions and the financial considerations that customers need to navigate.
Similarly, creating an automated irrigation system would start by envisioning the ideal scenario – what a well-watered garden looks like – as well as the decisions a gardener would take to irrigate a garden manually.
When we make decisions as human beings, our mental process usually includes taking in information and then taking action. This is how algorithms work too.
Inputs might include data that the user enters in an app, data from third-party sources, APIs that cull specific data such as weather or data from sensors that measure everything from temperature to whether machine parts are wearing down.
Next, you need to think through the outputs that ensure the algorithm is personalizing the experience properly and consistently. These need to be expressed in a way that connected devices and applications can use.
Some algorithms are more complex than others, but it’s always best to start simple – a spreadsheet can work as a starting point -- and add more functionality as you go.
In a first prototype you might use basic operations like additions, multiplications and conditions, and simulate how it works with common and less common real-world scenarios.
Once your prototype has proven viable, it’s time to simulate the solution with actual customer data. It’s also important, wherever and whenever possible, to involve customers in the process and have them experience personalized experience you’re using an algorithm to deliver.
Whether you use a formal survey or conduct one-on-one user interviews, make sure you gather as much actionable feedback as possible. Assess whether your users truly understand the value proposition behind the personalized experience, how their data is used, your privacy settings and more.
Reinforce the message that you will continue to refine the solution until you’re truly read for your first release.
Managing an algorithm for connected devices may require a significant shift in how you store and manage data. Information that was formally siloed across disparate systems, for instance, may need to be consolidated or integrated in a new hosting solution.
These interconnections should be mapped out and summarized in a data architecture diagram that can serve as a blueprint for ongoing back-end development and maintenance purposes.
While the potential for personalization may be limited only to the imagination, your use of algorithms can always start small. It is also important to use data analytics to keep learning after launching the first version to ensure you’re achieving the right results.
As with any business initiative, everything you do with data science and connected devices should be aligned with tangible objectives. A good example is retention: will customers keep coming back to the experience you’ve developed – and if so, does that lead to business metrics such as increased revenue and market share?
Analytics should also help you gain visibility into how customers use your products and services at scale. This helps you steer development and allocate the right resources to fix recurring issues and improve the algorithm to keep up with (and even anticipate) their needs and expectations.
In fact, personalization should not be treated as a once-and-done activity but a continuous work in progress. That’s why your plan should look beyond the initial launch to what kind of features and functionality you can build into Version 2, Version 3 and beyond.
When you work with a trusted advisor or partner, meanwhile, create a long-term plan that includes a process for handing off intellectual property and other resources to any software developers, data architects or data scientists you employ internally.
Finally, use whatever personalization efforts you undertake to develop a holistic data strategy for your business, where everyone can understand how data can support your business goals and inform your product and service map accordingly.
No matter what kind of organization you’re running, you’re ultimately serving people. They deserve the most personal approach to an experience you can give them.
Oct 22, 2022
TOPICS
SHARE
Sign up for our newsletter to stay in the loop on all of the cool things we’re doing!
.png)
Data
