In the last days, there have been a lot of news about this company “Cambridge Analytica” (CA) “stealing” data from Facebook and manipulating the US citizens in order to get Donald Trump into the White House. CA used big data statistics to determine the pain points of each user, in order to send them the right ads and messages.
Some people now asked us, if this is similar to what Placedise does?
And if not, does Placedise target specific user groups?
What really happened
First, let’s get clear on what really happened with this presidential election. Since, especially in Social Media, people talk a lot without knowing anything, I want to point out the case in a little more detail.
The Trump team, like never before, wanted to make strong use of Social Media and all marketing possibilities nearby. That is nothing special. They selected the best fitting measure.
They realized, that it is possible to put people into groups, by making use of statistics and applying it to their user profiles. In fact, that is something even Facebook offers right out-of-the-box when you use Facebook Ads. Also in fact, this is something research institutes used even long before Facebook, when doing surveys and predicting the results of an election. In this process, they did not only collect a simple yes/no answer, but enough data to be able to say something like “people, age 33-45, living there, no income, with a dog and two children, would vote for X, because he/she supports Y.”
CA collected this data via Facebook by using different clever tactics (like fake research surveys and more). This is a process that many larger websites use to classify their visitors. I know this at least from 10 larger news websites. You ask simple questions for a fun fact survey like “which country fits your personality”. It is like a game, but in the background, they add the answers to the user’s digital profile, in order to target for advertisement. For example, they now can send special ads to all users, who love Italy and sometimes feel lonely. Imagine an ad that promotes a situation where you chill alone in front of the TV, with some good Italian wine. Buy that wine!
Only problem with the previous step: CA did not stop with those users, but also gathered connected information, that belong to other users. This was only possible, because the relevant users did not protect their profiles in their settings. That is, why the media calls it stealing, but I am not sure, if this is the right word. It has been a feature of Facebook. Another problem: CA did not collect this data directly, but bought it from Aleksandr Kogan, who originally did those surveys. Something, not allowed by the Facebook terms.
Next, CA identified different user groups and clustered them by their pain points. They then played the right messages towards them. For example, “Let’s build a wall” for all people with low income, low education, struggling in life, white, conservative, maybe hating Mexicans. And then find similar groups.
Is this manipulation?
It depends on the definition of “manipulation” in this context. If strengthen specific, already existing feelings and thoughts in people’s minds, yes, it is. If we define it as magically make people do something, they would never do without the measure, then it is a clear no.
As mentioned many times before (see for example: Is product placement ethically questionable?), it is not possible to make someone do something against their will with regular communication measures.
This means that CA did not create those voters. They just used smart statistics to tell the right people the right story. In the end, this is nothing different to all other campaign measures, except the fact, that it was (because of the targeting) way more efficient (not necessarily cheaper). In the end, they did the same, Amazon or WalMart do with their online ads. Or McDonalds or Ford or you name it. Let’s rephrase this: CA did not manipulate people to vote for Donald Trump! They just helped the campaign team to play their ads more efficiently.
However, the consequences (not directly because of CA, but more the general fear of modern tech): Looking at the EU, this kind of easy targeting will be no longer possible, because of new laws by this and next year. With the current events in mind, I would not be surprised, if this will be adopted by other countries too.
Does Placedise target people?
It depends. Placedise, of course, considers different target audiences and one of the main features is to consider way more parameters per measure than any other tool. However, we do not offer ad targeting. We are the step before the actual contact and with Placedise, targeting happens the other way around. This is 100% safe regarding any laws and the best you can get in the process where Placedise comes into play.
The process with Placedise
While in the above process, we look at a targeted channel, where you can play your ad, Placedise uses a different approach. This is mainly because of the fact that we mainly deal with measures like product placement or traditional TV ads that are hard to target when they air.
At Placedise, you define your target audience up front (similar to Facebook Ads). You say who you want to reach and what you want to achieve (for example more attention or better reputation).
Placedise then analyzes all the big data in the background as follows: The system checks your target audiences and finds it or similar groups in the data pool. It then determines how the ideal ad for this specific target audience with a given goal and product/brand would look like.
However, this is the point where Placedise stops. How you play the final campaign is up to you. This is mainly due to two reasons. First, Placedise is not a marketing channel itself, it is an analytics tool and digital assistant. It tells you what to do but does not do it for you (yet). In addition, with Placedise you look at a lot of different marketing measures. From product placement to radio ads. Facebook, on the other hand, has one measure: Facebook. Second, from our point of view, it does not matter. We help you to optimize your ad in a way that the defined target audience gets effected the way you want. All other people are irrelevant at this point of time. Imagine a product placement in a movie. You have 100 people in the cinema. You want to address people, 25 to 30 years old and want them to love you. We help you to carve your message in a way that you achieve this goal best*. At this moment, it does not matter what the other people think about the ad (despite negative reactions – we got this covered, too). Maybe now, you are thinking “but what if the other people, for example parents, are also relevant for my goal or could interfere with it?”. In this case, you would need to include them into your target audience. But usually there definitely are irrelevant groups. If you sell ladies razors, you should not care what 70 years old grandpa thinks.
In the end, you can still use Placedise in the planning and then another tool to target your measure on the channel. For example: Use Placedise to do the ideal branded entertainment movie and then use Google/YouTube targeting tools to get the movie only to those people that you want to reach.
To answer the opening question: That is the difference.
The Facebook “scandal” is nothing special if you look at the details. It is simple targeting and statistics. Nothing big illegal. Anyway, it might change something in the industry.
Placedise uses similar techniques but operates on the “safe side”. It helps you to get your message in the best manner with the biggest impact to those people that you want to reach.
*btw: That is something you cannot do in any way with Facebook ads, since those processes happen on a subconscious level. There is no way to use Facebook Analytics to see if your ad works this way or not. You would need complex market research or Placedise as evaluation method up front.