The ROI of personalized experiences: Audience measurements

This is the first of a three part series on the ROI of personalization. The second and third parts will look at content and process measurements.

Recent statistics support the need for brands to create more personalized customer experiences.

• 80% of consumers are more likely to purchase from brands that provide tailored experiences.
• 70% of consumers say their loyalty is impacted by how well a brand understands their individual needs. 
• 71% of customers get frustrated when they don’t have personalized experiences. 

While all of the above may be true, it can be hard to determine the return on the investments needed to create truly one-to-one experiences. The path to doing so can take people many months, plus millions of dollars, to get right. 

In this three-part series, I will explore how marketers can measure performance and returns when creating personalized customer experiences. I will also cover several questions that any organization should ask before embarking on what could potentially be a large-scale initiative.  

Let’s start our discussion of measuring personalized customer experiences by focusing on the important aspect — the customer. Measuring personalization by audience performance means that we are measuring both individuals and groups of individuals and how they react to more tailored content, offers and journeys. 

In this article, I will go over different ways of looking at audiences and personalization and address some of the skepticism around the returns on one-to-one personalization’s efficacy. 

Understanding the audience is key to personalization

We’ve all gotten emails or text messages that say the equivalent of, “Hey [insert your name here], would you like 50% off on [insert product or service here]?” 

Some may call that “personalization” because your name was inserted instead of simply saying, “Hey, random person.” But this is not the type of personalized experience I want to discuss here. 

Let’s call the approach I just described “substitution” rather than “personalization” and focus on more robust examples and ideas instead. 

To do anything beyond simple substitution, however, we need to understand much more about our customers or potential customers than just their first and last names, email addresses or phone numbers. 

Enter the first-party data strategy and the reason you see so many brands investing in tools like customer data platforms (CDPs) and even second-party platforms that pool customer data among trusted parties. This strategy is crucial now that third-party cookies and mobile device ID tracking are being deprecated by major technology companies. 

So here’s a question, if personalization depends on understanding your customers well, how do you know how well you understand them?

The answer lies in creating a first-party data strategy and infrastructure that allows you to build customer profiles and evolve their information over time. 

Building a first-party data strategy requires: 

• The right infrastructure (e.g., CDPs and CRMs). 
• The trust of customers who provide their data to you in the first place.
• A way to continually enrich and serve customers with personalized experiences based on that data.

Dig deeper: What is personalized marketing and how is it used today?

Relative measurements are critical to understanding personalization lift

This one is for the true personalization skeptics out there. How can you understand if your personalization is working unless you set up a true experiment and measure the difference between using personalization and not? Not very well, indeed!

This is why a relative measurement can provide many insights about how your personalization is working (or not). Think of it in terms of the following (which you’ve probably seen before):

• Show a customer the exact product they bought in imagery on the website or an email.
• Show a customer the product they customized on their last visit.
• Customize imagery the customer is shown based on their geography or other demographics.

Then, compare the results of doing that by showing them a generic product or other one-size-fits-all text or imagery. You may see a lift in some areas and not others, but this is part of the value of using relative measures.

Because creating personalized content, offers and experiences takes more resources than a one-size-fits-all approach, it’s important to have a better understanding of the levers that generate the most value. 

Eventually, you may be able to personalize everything for everyone on every channel. But in the meantime, knowing what aspects have the biggest impact can help you make meaningful improvements without the resource drain. 

Dig deeper: How to humanize the digital experience with first-party data

CLV is the ultimate measure

While there are many helpful measurements to determine the effectiveness of personalized experiences, the most beneficial can also be the most challenging to use. Customer lifetime value (CLV) requires both a wealth of information about an individual’s full set of actions and, perhaps, the most precious commodity of all — time. 

Measuring CLV enables us to truly see the effects a comprehensive, personalized customer experience can have on the buying and product or service usage of an individual. It factors in the cost to acquire a customer, which can often be an investment to convert them and then demonstrates how a single customer can drive value over time.

Of course, the time-based component does make this the most challenging. For instance, if the average lifespan of a customer is over five years:

• How do you get a useful CLV model in a relatively short amount of time?
• How can you tell what role personalization plays in that model?

I’ve seen many different models for calculating customer lifetime value, but they will use historical data to create averages for spend over lifespan, churn and more. You can use these as your baseline measurements.

In addition, you can use relative measurements to see the effect on customer lifetime value for those customers that received less personalized experiences versus newer customers that might have benefited from more personalized ones.

I’ll talk more about multi-touch attribution in a later article in this series, but being able to attribute value and conversions to specific interactions and channels or touchpoints can also help when you are asked a question about determining the value of personalization in the overall CLV.

How much personalization is enough?

Beyond the statistics telling marketers personalization has a positive impact on buying behavior, as consumers ourselves, we appreciate it when our experiences are tailored by brands. 

But that personalization comes at both a cost to the brand (which might be passed on to us as customers) and potentially to the amount of data we give away (which can affect our data privacy). So the question remains, how much personalization is enough? And is there such a thing as too much personalization?

There are a few ways to look at this. 

• From an internal resource perspective, too much personalization too quickly can drain resources if the right systems and processes aren’t set up to handle the increased needs. 
• True one-to-one personalization relies on artificial intelligence and machine learning (AIML) models and predictive analytics that can work very effectively but needs time and training to do so. 

So perhaps instead of asking how much personalization is enough, it would be better to ask: 

• How much personalization is enough to create improvements now?
• What should we be building for the future?

Taking this approach means that your customers can benefit from a more tailored experience while your internal teams and infrastructure adapt to the changes needed to continue making these personalized experiences more effective.

When you pay attention to these important aspects of your audiences, measuring the ROI on personalized content, offers and experiences becomes greatly valuable.