4 Common Mistakes in Personalization and How to Avoid Them

Though consumers are aware we’re collecting data from them, they have expectations for how we use it to personalize their experiences. See the most common mistakes in personalization and how to avoid them.

Personalization in marketing isn’t new. Businesses have long used demographic information from prospects and clients to segment and personalize their messaging.

That said, there’s so much more data and different types that we can collect these days. And so many channels to get it from.

Study after study tells us that consumers want personalization in their digital experiences. So long as we personalize our content and offers the right way, they’re fine with us doing it.

But what exactly does that mean? How do you know if you’ve done enough personalization or if you’ve gone too far with it? Below, we’ll look at four mistakes that are often made when personalizing digital experiences and ways to avoid making them in your own marketing campaigns.

4 Mistakes Made When Personalizing Digital Experiences

You can personalize so many different aspects of the digital experience. For example, you can:

• Add a personal greeting to the tops of your emails.
• Display custom recommendations on your homepage.
• Let shoppers know when a product is available for purchase at their local store.
• Show targeted offers in Instagram ads to a certain segment of customers.
• Send discount codes to customers on their birthdays.

When done right, personalization can be a huge boon to business. Personalization can help you increase sales, customer retention and loyalty. When it’s done the wrong way, though, it can hurt your business and cost you customers.

To avoid this, you have to first understand what these no-no’s in personalization are. Here are the most common ones:

1. Over-personalization

Can you overdo it with personalization? You certainly can.

Think about an online experience you had where the website or app provided a recommendation or showed you content that seemed to know you a little too well. You weren’t entirely sure how they knew that tidbit about you—like that you’re in severe debt, you’re three weeks pregnant or your pet just died.

The info is valid, but you didn’t share it with them. Or perhaps you did in some indirect way, like by looking at baby products in their shop or watching relevant videos on their Instagram page.

The fact that this website or brand seemed to know you too well was incredibly off-putting and creepy. So, what was your response? Chances are good you didn’t do much more interacting with that brand. Because who knows what else they’d do with your data?

Brands going overboard with personalization is one of the biggest faux pas in marketing. Yes, your users know you’re gathering data. But just because you have the ability to do so doesn’t mean you should collect as much as possible and use it in invasive ways.

The better approach is to review the kinds of data you’ve collected from your users. Then decide which bits can help you create a great user experience. It should:

• Catch the user’s attention
• Get them interested enough to interact
• Boost their satisfaction with your brand

Let’s look at the example of Indeed. This is a website that collects tons of data on its users. That is, if they want to give it up.

On Indeed, job seekers fill out applicant profiles when they sign up. They include all kinds of info about their prior work experience. They can also set preferences that help improve their search results and recommendations.

In this screenshot, you’ll see an example of what the “Jobs for you” page looks like. Job summary cards appear on the left; full descriptions on the right.

Notice the green, yellow and gray blocks within each section. Green means it matches your prior work experience or preferences. Yellow or gray means that it doesn’t.

It might seem like Indeed is overdoing it with personalization here. After all, they’re not only showing me job recommendations based on previous searches and my work experience, they also use a plethora of data to demonstrate the compatibility or fit of the job.

But it works in this case. For starters, I actively consented to sharing all of this data with the website. Secondly, it makes it much easier to sift through job postings if I can see how my experience and preferences align with what the employer is looking for.

So, over-personalization doesn’t always mean using too much data. You have to understand who your users are and what they expect and need from you in terms of personalization.

2. Disregarding Data Privacy

Another mistake that marketers can make with personalization is ignoring their users’ privacy concerns as well as their data security rights.

The EU’s GDPR, California’s CCPA and Canada’s CPPA all set out to define the data privacy rights of consumers and how businesses need to act in accordance with them. There are other regulations that deal with specific industries and types of data, like PCI DSS in finance, SOC 2 in software and HIPAA in healthcare

To ignore these official regulations puts your business at risk. Not only will you have violated official guidance on the matter of data security, you’ve abused the trust of your users and potentially put them and their data in harm’s way.

Other ways you might be disregarding your users’ data privacy are as follows:

The first is by using consumers’ data in unsanctioned and inappropriate ways.

This one is obvious. Whether you’re collecting data that your users are unaware of or you’re doing something with the data that’s unethical, this is a huge problem.

While consumers might appreciate personalized marketing these days, they don’t look kindly on brands that abuse the access they’ve given them to their data.

To avoid getting in hot water with this, come up with a data privacy and security policy for your organization. This policy should include:

• The generation of a Privacy Policy, publication of it on your site and regular upkeep
• A framework for how you’ll secure your website and encrypt the data gathered from it
• A policy on what data you’ll collect, where you’ll store it and for how long
• A procedure that allows users to manage their data and request its deletion from your servers
• A process for users to easily opt-out of personalization

This policy should be available to everyone who works within your organization and collaborates with you on your website and marketing.

The second is not informing users about what data you collect and what it’s used for.

It’s not like you can force anyone to read your Privacy Policy. However, you can at least make it apparent to users that you have one while also providing them with more detailed information about the cookies and other data-tracking methods you use.

It’s also important to explain what you’re doing with the data. Or, at the very least, explain what you won’t do with it.

HubSpot, for instance, provides a brief explanation of what they’re going to do with this person’s email address when they download this report.

In addition to linking to the Privacy Policy page, HubSpot informs the user that they’ll receive more than just a report from the company. They’ll also be contacted about relevant products, services, as well as content. If they don’t want to receive these communications, they can unsubscribe later.

This brings us to the next point.

The third mistake brands make is not getting users’ consent to use the data they collect.

HubSpot is transparent about what it’s going to do with users’ email addresses. That’s good. There’s just one problem.

Users don’t fill out this form so that they can be marketed to or pitched new services and products. They fill it out because they found a report with data they’re interested in reading.

This report is strictly about data. Forcing researchers to consent to marketing and sales when all they want is access to data seems inappropriate in this instance. I know I’ve been on the receiving end of this. I use lots of these types of reports when writing articles. And I inevitably get that phone call or email from a salesperson wanting to pitch me on their service or software.

Whenever users give you data, you need to put it to use in an appropriate manner. Not only that, you need to make sure you have their consent to do so.

One way to avoid this problem would be to add a checkbox beneath the statement. This way, users can opt in or out of these communications. If you don’t, forcing consent might backfire on you.

Another thing you can do is to use a cookie consent banner. The only problem is that they were so pervasive after the implementation of GDPR. While these banners often gave users the ability to pick and choose which data the website could collect and use, most people developed banner blindness to them. Instead of considering what each website wanted to do with their data, they mindlessly clicked “Accept” or “Decline” to get past it.

Cookie consent forms are not a common sight online anymore. Although they weren’t being properly interacted with by users (understandably so), they helped brands communicate their data collection policies to users. And get visitor consent to use that data as outlined.

Whether or not you have a cookie consent form, you should always provide users with the ability to opt out of communications. Customers also shouldn’t be forced to share every piece of data with you. So, when setting up accounts, they should be allowed to opt out of certain kinds of notifications and have full management rights over their data.

3. Using Outdated Data

Here’s another one I’ve bet you’ve encountered from the user side of things. You open up your email and see this message:

Hi [fname],

We’re sorry to hear you were unhappy with your most recent purchase. In accordance with our 365-day refund policy, we’ve issued you a refund of $62.31 to the credit card ending in -2631.

Kindest Regards,
Your Favorite Retailer

When you see something like this, you probably start to think about a number of things. Like:

Do they really not have my first name?

If they can mess up filling in the details, what else can they mess up? Have they sent my credit card info to someone else, for instance?

If their customer service process is this shoddy, no wonder they’re making such terrible clothing.

Creating personalized content with errors isn’t going to instill any trust in your users. Whether you leave placeholders in the content or the data is missing altogether, it makes your brand look incompetent and unprepared.

Now, you could blame your personalization software on this kind of error. But is that really the issue?

If the software is to blame, then you’re using the wrong software. It should be able to handle a simple task like pulling in requested info from your existing database. Another problem could be that you’ve entrusted your entire process to automation and failed to have someone review the resulting content to verify that it looks good.

But let’s say the software isn’t to blame. Then who is?

Well, it could be the person creating the personalized content. They should not be calling on data that either doesn’t exist or is incomplete. For instance, if the [fname] field isn’t one that you require users to fill out, you shouldn’t be including it in your content. That’s how you run into these incomplete messages.

On a related note: Another thing to consider isn’t just missing data, but outdated data. This one you may have encountered as well.

Let’s use this SMS as our example:

Severe weather alert! There’s a tornado warning in effect this evening. The evacuation plan is now in place. If you’re planning to leave, please get to safety prior to 8 p.m.

This seems like a pretty straightforward message, right? A tornado is rolling through, and a weather notification service is letting this person know about evacuation orders. The only problem is that the message leaves out an important detail—that this order is for the city of Jacksonville Beach and not St. Mary’s.

The subscriber should not have received this message as they’d already unsubscribed from Jacksonville alerts. They changed their ZIP code and communication preferences last year after they moved. For some reason, though, they were left on this company’s SMS subscriber list for their old address. Now, they’re scrambling to find out what’s going on because they think a tornado is about to hit Georgia.

Outdated data can be just as harmful as missing data can be your brand and the relationships you make with customers. It makes your operation look sloppy and unaware.

So, it’s important that, before you personalize your communications, you’re assured that your data is in good shape. If you don’t have a data management policy in place, make one now. This might include:

• Having employees follow up with users in your CRM to confirm their data is accurate each year.
• Prompting users by email to review their account and profile information and keep it up to date once a year.
• Removing inactive users from your database.
• Scanning the database for missing or clearly incorrect info, then following up to fix and validate it.
• Double-checking that all personalized marketing messages pull in data only from required fields.

Another thing to think about is if missing data is hurting you. And I’m not talking about the occasional missing field about a user’s demographics.

For instance, is there certain behavioral data you’d benefit from having, but have continuously cut your personalization efforts short because you don’t? Say you have an online bookstore. You’ve been wanting to show reading recommendations on the homepage based on browsing habits. However, you’re not tracking that data.

If it’s because you haven’t updated your privacy policy or cookie consent form, what’s holding you back? Or maybe your CFO doesn’t want to spend money on personalization software, despite the evidence that users want personalized recommendations like these. If you can create better personalized experiences and recommendations with this data, don’t miss out.

4. Not Testing or Optimizing Your Personalization

Whether it’s regular ol’ static content that everyone sees or personalized content that a certain segment does, you should always be working to improve it.

Studying your analytics will help. It’ll give you an idea how certain content or campaigns were received by your users as a whole.

But it’s not just content that misses the mark you should watch for. Even personalized content that performs well can be improved upon. That’s why you should have an A/B testing and optimization strategy as part of your process.

The first thing you should test is personalized content vs. non-personalized content. This will help you determine if personalization makes sense in this context. Then you can try out further variations of your personalization.

Let’s use an email I received from LinkedIn for this example.

I’ve given LinkedIn the ability to communicate with me about my post performance every week. This is what the message looks like:

LinkedIn tells me how many impressions my posts got while also highlighting the top-performing post. It’s a helpful summary email.

But one of the things we know about email marketing is that subject lines can impact open rates. If you can’t get someone to open the email, then they’ll never see how helpful you’re trying to be.

So, the first thing I’d test is the subject line.

This is the current one LinkedIn uses: Suzanne, your posts got 114 impressions last week.

I’d also try a non-personalized one like: See how your posts did on LinkedIn last week.

LinkedIn might find that people are more likely to open the non-personalized one since it’s clearer about where it’s coming from. The personalized subject line is fine, but it also is a bit too vague and could be from anyone. I’ve gotten messages like this from agencies trying to pitch me their SEO services and it makes me reluctant to open them. So, it’s something worth testing.

Another thing to test and optimize is the personalized content in the email or the order in which it appears. For example, you could try replacing the top performing post preview with a list of links to your top posts along with numbers of impressions and clicks.

With LinkedIn, that approach might be difficult as posts don’t have headlines. However, this is something you could certainly do with a blog writing platform like Medium.

When testing and optimizing, always keep the end goal in mind. In other words, why did you personalize this piece of content in the first place? What is it you want your users to do more of? The variations you test and the changes you make (even if it means doing away with the personalization) should help you better achieve these goals.

Wrapping Up

Personalization in marketing isn’t optional these days. Consumers are more than happy to give up relevant personal data so long as they know the info is safely guarded and it’s used to create better experiences and more beneficial offers.

So, personalization requires a lot of care. Think of it like walking a tightrope. Leaning over too far in one direction is risky.

To play it safe, familiarize yourself with the most common mistakes that marketers make in personalization. Understanding what doesn’t work or doesn’t sit well with consumers will help you avoid those pitfalls. Something else you can do is adopt an ethical personalization strategy. This will help you prioritize your users’ safety and priorities before anything else.