Communicating on social networks is no easy task, even in 2019. While 2 out of 3 French people say they are irritated by online advertising (source: IFOP), however, spending on digital advertising is constantly increasing (source: E-Marketing).
So how do you optimize your speaking on social platforms, in a context of audience saturation?
Our response at The Metrics Factory: to better know your audience and identify the people who say bad things about you (the so-called “haters”) in order to be able to exclude them from your communications.
But who are these haters?
Haters are users who will speak out with hate speech against your brand. They will do what is called “trolling”, which means they will try to harm your brand and create controversy. As they say “Haters Gonna Hate” so why keep talking to an unreceptive audience that's going to attack your brand?
Why spend money advertising to an audience that hates you?
To illustrate The Metrics Factory's know-how in audience analysis and the optimization of media buying, we took the example of a brand in the banking sector since it is a sector particularly affected by this phenomenon.
1st step: Audience analysis using a listening tool
The first step is to use a social media listening tool to listen to and monitor conversations around your brand.
Brandwatch, Talkwalker, Linkfluence, Linkfluence, Digimind, Synthesio... There are now a multitude of monitoring tools! These tools are generally used to monitor a brand's online reputation because of their ability to track down all the times it is mentioned on social networks, media sites, online news, online news, blogs and forums, etc.
If we take the example of Crédit Mutuel, we can see that the brand has been mentioned more than 52,000 times online, with the majority of mentions (70%) coming from Twitter. Indeed, Twitter being THE platform for instantaneity, it is widely used by users for after-sales requests or complaints:
In addition to the volume of mentions, we are going to focus in particular on a feature that listening tools offer: tone analysis (or feeling analysis).
This feature will allow you to associate a tone of expression with each mention: does someone speak well of your brand (positive mention), poorly (negative mention), or is it a mention without positioning (neutral mention)?
With regard to Crédit Mutuel, we note that the mentions are mostly neutral, i.e. they bring visibility to the brand but will neither promote or denigrate it.
There are also 10% positive mentions and 12% negative mentions, which we will focus on in more detail.
2nd step: audience segmentation
The second step consists in segmenting the audience in order to identify specific segments of users to exclude from your campaigns or to address through ultra-precise targeting.
To be able to arrive at a precise segmentation of haters, we will only keep the people who say bad things about Crédit Mutuel on Twitter (since it is the platform that delivers the most data and on which haters express themselves widely).
Once we have this user base/haters, we use a segmentation tool that allows us to share the audience into different communities using differentiating criteria (socio-demographics, interests, biographies, biographies, biographies, biographies, hashtags used).
We then obtain 5 segments within Crédit Mutuel haters, which we qualify according to their characteristics:
This graph of co-occurrences represents the distribution of the various segments of haters within the Crédit Mutuel audience. The more circles there are, the more the segment is present in the whole sample. As for the area of the circles, the larger it is, the more interconnected the user profiles are.
Segmentation then allows us to draw up fine portraits of these haters based on socio-demographic criteria, their behaviors (the media and influencers they will follow, the hashtags they use), their biographies... And to obtain strong insights on these different communities.
If we take the example of “Fintech experts”, this is a segment composed of 80% men, aged between 35-44 and living mostly in Paris.
These users are interested in finance, the economy, and new technologies. They will therefore consume media related to these areas and follow pages such as BFM Eco, BPIFrance, La Tribune or even French Tech.
They will therefore be more sensitive to these issues and will be more likely to express their dissatisfaction with technological problems (failure of the website, application, etc.).
They also define themselves as belonging to the world of finance, entrepreneurship and new technologies.
3rd step: translate this segmentation into media targeting and capitalize on a cross-platform strategy
After drawing up a detailed portrait of each audience segment, all we need to do is import this list into the Twitter Ads console to create a tailored audience.
Custom audience is a very powerful form of targeting that allows you to target or exclude a specific audience, using a list of users. A process that is still little used by advertisers. However, this method allows real optimization of advertising investment by limiting the negative repercussions by excluding people who have already said bad things about your brand and who therefore present a risk.
To go even further in the strategy of optimizing targeting, we at The Metrics Factory suggest that you capitalize on this Twitter audience segmentation through other advertising consoles.
For example, it is possible to translate this segmentation into Facebook Ads targeting, based on the insights (behaviors, interests, etc.) that we have been able to observe.
It's important to keep in mind that users don't operate in silos on social media. A Twitter user can also be present on Facebook, Instagram or even LinkedIn (although the uses of the platforms may differ).
So this is a good way to optimize your media investment and improve the quality of your audience on all social platforms, by excluding all of your haters from your speeches.
Audience analysis can also meet other needs.
Here are 2 other possible ways of optimization:
- The analysis of the promoters: This time, it is a question of analyzing and segmenting users who say good things about your brand in order to offer them specific content (a commercial offer for example). As the audience is more receptive, it will be easier to activate.
- Cannibalization of competing audiences : analyze the audience of your competitors to activate them in media and gain market shares.
