Data-driven marketing has become incredibly powerful for businesses large and small. That's undeniable. Whether you're automatically sending marketing emails based on previous clicks, or you're using dynamic remarketing in PPC to follow your website visitors around the web, personalizing the online experience based on "big data" is arguably the most advanced marketing strategy available today.

But what if your precise targeting and personalization go too far?

When used unwisely, big data-driven marketing tactics can turn customers off (or worse, creep them out), and might hurt your bottom line instead of helping it. The worst part is that you may not even know when it's happening.

The Problem: Marketers Are Blatantly Ignoring Consumers' Expectations of Privacy

People generally like to assume a certain level of privacy on the web. Even if we don't make a conscious decision to remain anonymous, we typically like to keep our online activity to ourselves, at least until we voluntarily choose to reveal our identities.

The reality is that online privacy has eroded over the years, and we've known this long before NSA revelations unfolded. We already knew that websites could place cookies on our PCs and track our online activity. We already knew that programs like Google Analytics were being used to track our behavior on individual websites. And we already knew that nearly everything we posted online could be accessed and collected by unknown parties.

Because of this conditioning, it's easy for marketers to inadvertently disregard society's time-tested desire to maintain privacy online. The reality may be that online privacy has been diminished, but at the same time, reality hasn't convinced us that privacy isn't important.

This is where marketers muck up. One particular dynamic remarketing service follows your site visitors around the web and targets them with display ads based on their online behavior. Then, after gathering enough data on your site visitors, the platform uses that data to identify and advertise to people who haven't visited your site, but fit the same profile as people who have.

Sounds great, right? This is a fantastic example of just how powerful data about customers can become. The problem is that this marketing tactic is risky. Imagine the following scenario.

You're an athlete. You occasionally shop for athletic products online and you've seen your preferred store's ads follow you around the web. No big deal. You understand how retargeting works and it doesn't bother you. But one day, you notice you're being followed around the web by an athletic company you've never heard of. That's weird. How does this company know who you are and what you like to buy? Do you want to do business with them? Or would you rather stay away?

My bet is that many of your potential customers would answer, "No, I don't want to do business with them."

And that's bad news for the advertiser who hasn't realized that customers are getting turned off by these ad practices. If you want to learn more about the risks involved with remarketing, check out this great list, "The Hidden Dangers of Retargeting in Search & Display Campaigns".

The Solution: Dial Down Precise Targeting and Stop Over-Personalizing

Remarketing, Personalization, and of course, Big Data are big buzz words in 2014, and I don't want to disregard their value. In some cases, consumers are willing to trade their privacy for a price -- freebies, coupons, and discounts. But only to an extent. Personalized email marketing and creepy retargeting ads following customers around the web can turn them off.

So, I encourage the marketing industry to take it down a notch. If you're using data-driven marketing tactics yourself, consider these tips to avoid creeping out your customers.

Retargeting Tips:

• Never remarket sensitive products or services.
• Make sure retargeting ads aren't shown too frequently. Keep it casual.
• Limit the length of time you retarget any one site visitor. Leave people alone after a reasonable amount of time.
• Don't retarget people who have already purchased.

Other Data-Driven Marketing Tactics:

• Dynamic ads and landing pages shouldn't reflect a user's search activity to the point that they can tell. Choose dynamic keywords carefully and embed them in a context that will always make sense, no matter which keyword is dynamically inserted.
• If you discover a unique insight about a customer, such as the time of day an email subscriber gets up in the morning and first checks his email, use that to your advantage, but keep your knowledge a secret. Don't send an email wishing him a good morning right when he logs on at 7 a.m.

Think about how people actually perceive your data-driven marketing campaigns. Are your personalized emails really helping you, or are they turning off loyal customers? Your data might not be telling you the whole story.