Don't let personalized ads scare users: avoid overuse of data

With the rapid popularization of the network, advertising at any time to show is not a problem. We often talk about accurate marketing, to accurately locate users. But do you know that overuse of user data makes your ads look scary?

As an example of a "horror" ad, one day I opened my mailbox and found my name impressively appearing in the side-bar ads. "How did he know it was me?" muttered his heart. Where did he get my message? Did they take away my privacy data? ”

In fact, for most users, "invasion of privacy" may be just a noun, a feeling. As for the specific information and data, no user with a little technical background is confused. So there is a ambivalence: both the convenience of personalized service and the fear of privacy violations. So there are a lot of people who do not understand blindly against the user data extraction behavior. In fact, if you use the user data just right, consumers will appreciate the service you provide them.

So the key to using user data is to put what we're doing on the table, just as many restaurants now replace the kitchen walls with glass. This is the first piece of advice from InMobi's article "How to personalize without maximally creepy" (how best to use user data to personalize ads without the fear of users): Frankly acknowledging the source of the data and not hiding anything. Advertisers, who are just starting to touch user data, tend to be obsessed with making ads look like magic. In fact, this does not achieve the desired effect. Personalized advertising can be done like a service, not as if you have mastered a lot of "privacy." You should publicly tell the user why you want the ad to be personalized and where the data is being obtained.

InMobi This article also puts forward several other suggestions for personalized advertising:

Personalized ads are added under customized conditions. (such as what users can see after they log in.) At this point, the user will not feel violated, because he is in the customized environment, and advertising will have better results. For example, LinkedIn offers job referrals, Facebook sponsored stories, Groupon's personalized discounts, and more. Consumers understand and expect this kind of advertising to unfold.

Ensure that users are entitled to "exit the game" at any time. Hunch's chief scientist, Hugo Liu, says that personalized services do not give users the right to quit, like building a tower without a fire escape (there are no fire ladders in Chinese houses ...). Khan ... But you can imagine a 50-storey building with only elevators and no ladders. Giving users the freedom to choose whether or not to "personalize the experience" is likely to retain more users. Because trust is based on mutual respect.

Use only the latest data. Avoid using historical data, not only to get the best results, but also to prevent scaring users. Surprisingly, the most valuable data for personalized services is often those that expose data. Getting a history of buying, evaluating, and so on may seem pretty good, but the data is outdated and often not very relevant to the user's current situation. Information age data quickly become stale, so it is rare to get valuable information with historical data.

The most valuable data for personalized advertising is the following:

1 device type, operating system and software development kit. Providing personalized experience for different devices is the key to making ads attractive to users. Sometimes we call this "targeting" (targeting), which is actually a personalized way. If you recognize that a user is using an Android 2.3.6 system, don't animate it, and if you recognize that the user is using the iOS 5.1 system, you can add a button to buy itunes.

2 Position information. Where is the user? What is happening around him/her? Is she shopping or listening to a concert? Is she in the hospital or in the museum? What's in the neighborhood? Location information can make the effect of advertising very significant. GPS data are available in HTML5. Advertisers should dare to ask users to share the data.

3 the user will. What is the user searching for? 4S shop or auto parts? Coffee or tea? Did they give up the shopping cart? It's important to listen to subtle clues, especially when it comes to "repositioning" (re-targeting) users. Data management platforms can address these issues.

To sum up, to allow personalized advertising to achieve the best results, please do not show off the technology, do not let people feel that you have too much information, to dare to advise users to actively share those less private information. Instead of chasing and obsessing over the hard work of your users, take advantage of the information you can easily get at hand.

This article source: Lei Feng Net