tightrope walker

Marketers are feeling the pressure to personalize. Seventy-four percent of online consumers reported feeling frustrated when content doesn’t adapt to or recognize their needs according to one report.

But personalization can go too far.

Nearly two thirds of customers are uneasy about the amount and type of information that companies have on them — a feeling that can make them less likely to give up any identifying information and erodes brand trust. When deciding how to personalize customer experiences, brands need to ensure their strategy doesn’t make their customers uncomfortable. 

Here are some do’s and don’ts for avoiding the “creepy” factor:

Do: Understand the Difference Between Implicit and Explicit Data

Much of the danger of being perceived as “creepy” comes from brands misunderstanding the difference between implicit and explicit personalization. 

Implicit personalization makes use of data that you infer from a user’s behavior, such as demographics based on IP addresses. Explicit personalization draws on information that users have willingly provided by filling out forms or otherwise volunteering data on themselves and their preferences. 

While brands need to use both kinds of data to personalize, it’s much easier to cross the line and make users uncomfortable if you use too much implicit data and over-assume their qualities and wants. Users will likely be more comfortable seeing personalized touches based on information that they know they’ve given you, rather than data they didn’t realize they were giving away.

Don’t: Be Greedy

Your users trust you with their personal information, and it’s your responsibility to use it in an effective way that adds value to their experiences. Don’t ask for data that you can’t use: you’ll erode their trust, and they will be less likely to share anything with you.

Likewise, if you prove that you can use their information to deliver a more valuable experience and relevant content, you’ll increase their engagement, trust and willingness to offer up more data. And it should go without saying, but if you’re collecting potentially sensitive information, be sure to have the proper privacy protocols in place.

Do: Plan

While this may seem obvious, it’s often overlooked and the root cause for providing over-personalized experiences. Planning personalization — identifying who, what and why — is often more important than the how of implementation. Don’t execute random scenarios. Understand the user you’re targeting and what segment they belong to, what you’re going to personalize against and why you’re personalizing.

This planning process will also help you tailor the levels of personalization to different users to avoid alienating a particular group. Using analytics tracking, engagement and lead scoring, you can segment users into groups based on how engaged they are. In general, users who are more engaged trust your brand and will be more comfortable with higher levels of personalization. 

Engaged users are also more likely to have provided explicit data and interacted with more touchpoints on your site, giving you higher quantities of more reliable information. Less engaged users require more subtle personalization: they trust your brand less, and you risk alienating them if you make too many assumptions about them.

Part of your planning process should be deciding what kinds of data you want to personalize against, and what kind of data is appropriate for what you’re hoping to accomplish. Names, emails and preferences for certain products or services are usually fair game; more identifiable information around demographics, for example, can sometimes be inappropriate to use. Don’t personalize just for the sake of personalizing; always be sure to have an end goal in mind.

Don’t: Over-Personalize

In general, personalization needs to be subtle. You shouldn’t be creating completely different scenarios from user to user. Rather, you should be creating a more relevant experience so that your users move more quickly down the path you want them to. Most of the time, it should be more about nudging the user than pushing them headlong into a fully personalized experience. 

Start slowly with your approach to personalization and gradually add more scenarios as you measure your progress. You’ll likely notice there is a fine line that can be crossed where over-personalizing is not adding any additional value, or having an adverse effect on performance. You’ll win the trust of your customers over time with each value-add, personalized content scenario that you deliver to them. 

Do: Measure, Measure, Measure

Did I mention measure? All the planning in the world still doesn’t guarantee an effective model. Mitigate the risk of having your model go wrong by having a strong measurement model in place that allows you to keep track of the effectiveness of the live scenarios you’ve implemented. Be sure to choose metrics that allow you to measure how your scenarios are providing value against others, those that you have in play and the original, unpersonalized scenario. Performing regular check-ins against those metrics will allow you to react quickly if your scenario isn’t working as well as it should.

Creative Commons Creative Commons Attribution-Share Alike 2.0 Generic License Title image by  Wiros