Korian’s Multi-data Approach for Better Elderly Care Image | Presentations and Cases

The battle against negative brand image in elderly care


Did you know that in Belgium 62% of people aged 50 and over have a negative image of nursing homes? A Belgian study also found that 70% of older adults fear living in retirement homes due to losing independence, isolation, or mistreatment. Public negativity towards an industry can put it in the spotlight with scandals and bad press. This relentless negativity can permanently tarnish the reputation of brands within that industry.

Korian, provider of long-term services for the elderly, knows this brand image challenge all too well. The Belgian branch of the group manages 116 nursing homes as well as diverse services tailored to the elderly population all over the country. While the overarching brand is Korian, the facilities, and the homes operate under local and regional brands that do not include the name “Korian” in their branding. This approach allows Korian to integrate with local communities and retain the names of acquired retirement homes that already have established reputations. This is a detail of our story that becomes important later.

Bad buzz can arouse curiosity. For prospective clients or their families, the Internet is often the first source of information, making search engines very often the starting point of the customer journey. Thus, as no surprise, Korian is reliant on search engines to drive business. To put it down into numbers 70% of the total digital marketing budget is allocated to Search campaigns.

Trust is a critical factor in the healthcare industry, especially when dealing with vulnerable people like the elderly. Negative information, whether true or not, can significantly impact a company’s reputation. It can erode trust making it much harder to attract new residents and retain current ones. To overcome these challenges, Korian has to engage in active reputation management. On the one hand that means well-designed PR initiatives. But there is more that can be done than positive messaging. It requires actively minimizing the presence of negative content. In this industry, very prone to bad buzz, acquiring traffic, and thus investing in audiences that are relevant and associated with positive intentions becomes a very practical challenge that has a direct impact on the bottom line.

Utilizing multi-source data for reputation management


To proactively manage Korian’s online presence on search engines, it is clear we need to find a mechanism that avoids brand association with what we call “negative sentiment”. The answer is not to strategically combine phrase match keywords and negative keywords to maintain reach while controlling relevance. Semetissians are strong believers in shifting towards broad match targeting alongside smart bidding strategies. When used correctly it can significantly increase the visibility of your ad to potential customers who might use different terminology than what is explicitly targeted in more restricted match types. The vast majority of data signals that the algorithm can digest to predict whether an ad is likely to perform well is beyond what human brains can do. Even though we observe great results, sometimes AI glitches, as it is part of the learning curve. However, in Korian’s case, there is absolutely no margin for error. So, we had to build a solution that allowed us to keep a broad match-targeting strategy balanced out with automated tools and algorithms that have a nuanced understanding of elderly care and its impact on Korian’s brand image. 


Identify keywords from various data sources

 The starting point was listing all the online sources where negative content can be consumed about the elderly care industry in general, the 116 retirement homes of Korian, and the specific retirement homes of our competitors. Two concrete categories of data sources could help us identify in real-time the sentiment:

  • Google Alerts sends daily notifications whenever new content is published about the industry, one of Korian’s nursing homes, or one of our competitors ;
  • A Social Media listening tool (powered by Meltwater) that automatically scans social media platforms (X (former Twitter), Facebook, Instagram, TikTok, Reddit), forums, review websites, articles on news websites and blogs, videos (YouTube and other video material), and even radio/audio spots for which a daily report is received.


Set up a script that connects data sources together

Next, the content needs to be read and keywords need to be extracted. For this part of the process, we built a script that scrapes all content flagged by Google Alerts and the Social Listening tool. The scraping happens hourly and produces a list mentioning concretely the name of the elderly home affected (if mentioned in the piece of content) and the associated term. This list is then automatically exported into a Google Sheet, that is accessible to the full team.

Besides digital data sources, there is another important success factor. Involving a very important stakeholder; the marketing and PR teams of Korian. These teams are often already aware of an upcoming negative press release that hasn’t even been published and thus reached the news sites. They communicate and warn us immediately so we can manually add the terms alongside the name of the retirement home to the Google Sheet. This allows us to anticipate, rather than having to react.


Sentiment Analysis Model to label the dataset

With Google’s Natural Language API a sentiment analysis is performed on the text and the keyword. By utilizing this existing machine-learning algorithm a sentiment label can be attached to every “query”. We kept the scoring model simple with three categories: positive, negative, and neutral. Words related to neglect, abuse, complaints, and lawsuits for example would receive a negative sentiment label.


Connecting this script to our performance campaigns on Google Ads

By using Google Ads Scripts we automate the process of scanning terms, labeling them with a sentiment, and automatically updating the negative keywords list based on the predefined criteria. Next, using Google Ads API this list is connected directly to the search campaigns running, excluding any ad appearance on a keyword that could damage Korian’s brand image or reputation.

Let’s give a very specific example. Imagine the elderly home with the fictive name “Sunrise Senior Living” located in Zonhoven has recently negatively hit the news. Apparently, a resident went missing for a few hours, which created quite some local fuzz. The automated alarm system put in place will catch this news update immediately, probably through a Facebook post in the Facebook Group “Ge Zijt van Zonhoven als”. The data retrieved from the source would be “Sunrise Senior Living”, “missing person”, “Zonhoven” in the Google Sheet. As the Sheet is directly connected to our search campaigns, the following will happen: the specific nursing home name in combination with related terms and dynamic geo-targeting data has been immediately excluded from our search campaigns, preventing our budget from being wasted on users only looking for the gossip… Handy, right?


Increased quality of traffic and cost-effectiveness of paid search campaigns


This strategy effectively shielded our client's online reputation and had a direct business impact. Let’s put it into numbers. The automatic filtering out of non-relevant queries led to a +17% rise in CTR, resulting in a 14% higher conversion rate due to decreased irrelevant traffic on Korian's website. But above all, relieved teams.


Why should this deserve an AMMA Award?

Why is this case noteworthy? It presents a common challenge many advertisers face. Our solution, while simple and not revolutionary, is not widely considered by many in the field. If you find yourself as one of the advertisers allocating 70% of your budget to search engines and sensitive to brand image issues, this approach earns attention.

Looking ahead how can we further exploit this?

  1. We plan to refine the model to achieve a more nuanced classification. Sometimes, terms might seem negative but could be used by individuals seeking solutions to those negative situations.
  2. Subsequently, these terms can be used to create a custom intent audience targeted on YouTube. This enables us to launch brand image campaigns that effectively communicate Korian's philosophy and practices in elderly care, thereby casting a significantly more positive light on the brand for those who have been influenced by negative press.

publication auteur Diane Tremouroux
Diane Tremouroux

| LinkedinThis email address is being protected from spambots. You need JavaScript enabled to view it.



Get in touch

Semetis | Rue de l'Escaut 122, 1080 Brussels - Belgium


Connect with us

Cookie Policy

This website uses cookies that are necessary to its functioning and required to achieve the purposes illustrated in the privacy policy. By accepting this OR scrolling this page OR continuing to browse, you agree to our privacy policy.