Discriminated by targeting: why you need to be mindful of users’ feelings when conducting digital campaigns

We used to think that each advertiser dreams of targeting their audience more accurately (meaning: more narrowly) and, at the same time, of saving on advertising. The previous decade in media can be generally described by the phrase “segment or die out”. Digital media provided such strong targeting capabilities that some brands really took off. The recent popularisation of programmatic technology in Russia, the hype around Big Data and the growing volumes of customer data give a feel that you are able to conduct a meaningful dialogue with almost every individual. In fact, this is not quite true.

Rather, it is possible, but is not (yet?) necessary. And here are three reasons why:

1. An interest in a certain topic or an intent to purchase expressed in a search do not always attribute to a prospective customer. Often, the target audience does not meet expectations whatsoever and the budget “saved” is just wasted.

2. By using audience data to create personal offers to customers, you can end up discriminating against certain groups.

3. By presenting narrow, personally targeted desktop advertising, you risk becoming a source of embarrassment for the user. Just imagine: an adviser wants to show a certain government minister a site on their laptop; the page loads and there is an ad for foot fungus cream and a personal discount. To win over a person, you need to rescue them from uncomfortable situations, and not put them into one.

When the scope is too narrow

It is like the Facebook news feed: you can spend the whole day discussing the blue & black/white & gold dress or Kim Kardashian’s really nice trousers, but you won’t learn anything different, which is sometimes more important. Having defined the segment “women aged 25–55, owing kittens up to 12 months old”, you can miss those who just bought/rented an apartment, changed a job or married/moved in together. But these are typical situations where a family gets a kitten. And, don’t forget about all the crazy cat guys over there!

Not guessing the target audience is not the worst thing that can happen. Since the TV age, the audience have got accustomed to the stream of unconsciousness: during a 3–5 minute ad break, you would learn a lot of new and interesting, but completely-off-the-mark stuff. The price of failure is a “wasted budget”, and that’s all. Worse things might happen, but more on that later.

How can you avoid this? It sounds corny, but before starting your advertising campaign take time to research the audience (as Kotler teaches). But please, use analytical sites and audience data, rather opinion polls. Install a pixel for a programmatic platform you trust on your site. In 2–3 weeks, it will give you an insight into who those people are, who already visit your site and are loyal to your brand: social demographics, geography (technically, you can also find it in GA), their interests and intentions — have a look whether it is the way you pictured the target audience to yourself in early planning? And then, by using Look-alike, you can target users with similar characteristics.

Digital Discrimination

Companies with a large volume of data about audience may voluntarily or not abuse this knowledge to their detriment.

The US Federal Trade Commission (FTC) has recently published a report, which described the future of Big Data and the risks associated with the development of this approach in marketing. Specifically, the regulator cited examples where business people, using information available to them, took actions that infringed upon the interests of the disabled and disadvantaged.

Any technology is a binary code. “Yes” or “No”. “A kitten” or “a puppy”. “To eat” or “to drink”. However, a machine cannot always flawlessly tell one thing from another based only on the available data. In other words, you can put a prospective customer into a group of too poor, too sick, too purple, etc. based on insufficient data or incorrect interpretation. You deprive yourself of earnings and your potential customer — of the needed product or service; if your creative approach/message breaches moral and ethical standards, your reputation will suffer as well.

How can you avoid this? If you have already started playing with data-driven marketing, make sure that you or your data supplier have enough data and that it is exhaustive. If you do not have enough, buy more. Make assumptions when segmenting your audience, but when formulating a personalised proposal, make sure that the person understands the way you thinks and your logic.

Media

Finally, an over-precise targeting (even if correct) used via a non-personal media — e. g. on the desktop — is sometimes more annoying than anything else. And do not let the example of the advisor and the minister confuse you — almost everyone has encountered an inappropriate retargeting. It is like checking personal social network messages on your TV in the living room. Nobody does this.

How can you avoid this? When making someone a personal offer in an ad, imagine that you are whispering in their ear: no one else should hear you. Choose inherently personal media, such as mobile, e-mail/instant messengers (if there is a user base of those who wish to hear from you, of course).

What is the bottom line?

The answer is simple: when limiting the audience for your communications, do not limit your perception of reality.

What do you think on the matter of targeting?