"Consumer Sentiment Insights" Towards Halloween Products 2019

Trick-or-Treat? Consumer sentiment towards Halloween products from big brands: McDonald’s, Burger King, Oreo, Jaffa Cakes.

Happy Halloween!

Every year, big brands roll out seasonal special products. It’s an opportunity to build fun social media, TV commercials and in-store campaigns around them. Spooky shapes and flavours colour the isles of supermarkets and menus of fast-food chains. 

Rivals like McDonald’s and Burger King take this occasion to create playful campaigns against each other. For example, in  2017 Burger King’s Halloween commercial used a scary “IT” look-a-like version of Ronald McDonald for their Halloween campaign. 

Although these campaigns may look fun from the outset, what is the real feeling of consumers about them? Are they really liking the creativity of the campaign and the taste of their products?

In this blog, we are going to explore consumer behaviour and sentiment towards big brands’ Halloween products. For some Halloween fun, we wanted to try out some free and open source tool from the web to find interesting consumer behaviour insights. This tool tries to capture the consumer sentiment analysis that is an estimate of the emotion contained in a tweet’s text.

Data Visualisation Explained

For each product analysed in this article, the keyword analysis will be shown in three ways:

  1. Tag clouds: common words are put into a cross graph and divided into four different sections, or “emotional regions”; Upset, Happy, Relaxed, and Unhappy (clockwise order). Words that are more frequent are larger. 
  2. Sentiment Graph: each tweet represented by a circle and positioned by sentiment. Unpleasant tweets are drawn as blue circles and pleasant tweets are in blue.  
  3. Screenshot of the Tweets that the tool is able to extract from Twitter.


McDonald’s have a history of Halloween campaign releases, from spooky Happy Meals to horror burgers. This year, McDonald’s launched a competition for their consumers to get a chance to win a visit from McDonald’s McNuggets VANpire. In addition, they focused on creative designs for packaging and on introducing Halloween themed toys such as a Hello Kitty glow-in-the-dark Skeleton.

What are people saying on Twitter about McDonald’s Halloween campaigns?

Most frequently tweeted words: “Halloween”, “McDonald’s”, “Hello Kitty”, they are all positioned in the “relaxed” emotional region, with very close frequency to the “unhappy” emotional region. Unfortunately, the brand has been criticised for controversial Halloween decorations displayed in a few stores resulting in people sharing negative opinions and affecting the results (see below tag clouds after the criticism on the controversial decorations).

Overall, the sentiment towards Halloween at McDonald’s is concentrated in the “relaxed” section. The words used in the tweets indicate a calm and relaxed feeling with words such as “enjoying” and “simple” being used, as can be seen in the below top tweets.

The customers sentiment is positive. However, consumers have been tweeting suggestions for new Halloween products and… The fun (not so much for McDonald’s) is that most of the Tweets are about dressing up Burger King stores like McDonald’s for Halloween. Will McDonald’s ever respond to the joke? Will McDonald’s versus Burger King, be the next Marketing epic battle, just like Coca Cola and Pepsi? 

Burger King

This year, Burger King launched a new Halloween version of the famous Whopper; a Ghost Whopper and it’s not the look of the burger that changed, but the taste. In fact, the buns are flavoured with white cheddar cheese.

From the Tag Cloud analysis, the most frequently tweeted words: “Halloween”, “Whopper”, “Ghost”, “Burger” are allocated in the “happy” and “relaxed” sections, hence associated with positive and pleasant tweets with just a few negative in the emotional regions. A sign of a really good Halloween seasonal campaign!

As expected, the tweets sentiment analysis concentrates in the “relaxed” emotional region too. With most of the tweets sharing the launch of the new product rather than feedback on it, as seen here:

The sentiment is therefore neutral to the product itself, although by the virality of the posts about the launch of the product, it is safe to guess that people talked and liked the idea and shared it among their audiences.

Oreo Halloween Cookies

Love Oreo cookies? How do you feel about a Halloween version of the famous Oreo vanilla cream cookies? 

A biscuit base cookie, stamped with spooky pumpkin faces and a centre of orange cream, replacing the traditional vanilla flavour cream. 

Here is the Tag Cloud results for this product:

The most used words: “Halloween”, “Oreo”, “Chocolate”, “Orange” are located in the bottom right “Relaxed” emotional region although a few are in both the above “Happy” emotional region and the bottom left “unhappy” emotional region.

However, the sentiment map shows just a few, albeit positive, tweets towards the relaxed/calm sentiment.

Compared to others products explored, Oreo’s consumers are in fact sharing their love for the product itself with tweets like:

Jaffa Cakes Lemon and Slime Cake Bars

Due to the limited stretch of data for the tool, Jaffa Cakes sentiment analysis showed only a few results with the Tag Clouds concentrating mainly in the “relaxed” emotional region and the “unhappy” emotional region. The latter contains mostly unrelated keywords.

However, sentiment analysis results to be positive! Jaffa Halloween Cakes seem to be related to pleasant consumer sentiment, as shown in this tweet:

It was a fun exercise to play around with all of this consumer data, however, the SentimentViz tool clearly isn’t build for Halloween themed analysis. As most of the sentiment was scored by ‘relaxed’ and ‘calm’ values that clearly don’t align to scariness factors. Therefore, using a custom service like Skims, we could develop a new type of Sentiment Value around scariness. Watch out for our 2020 Halloween blog next year.

Why you need better Consumer Behaviour Insights

The above is just a small and limited search made with a free online tool. Imagine how much more is possible when you have a bespoke tool built for you to discover your own and your competitors consumer behaviour insights. 

By integrating external unstructured data to your database you can:

1- Track your own brand or a competitor’s and get a wider view into what customers like and dislike across the brand, a product range or the entire company. 

2- Track your or your competitor’s consumer sentiment after a product launch. You will be able to link common features across competitor products and gain really useful insights to help you create a smarter product roadmap. Being customer-centric is key in today’s market.

3- Win a new client pitch! Find products, services or features gaps from your competitors and make sure that you have the extras it takes to win your clients.

4- Track complaints. Know what your customers say about you, good or bad, all over the web.

How to get better insights?

We like to keep it simple and easy for our clients. We have spent the last 5 years building a powerful tool that helps market researchers and brands find insights within data.

We use Unstructured Web Data and AI to help our clients see the invisible. It’s a three-step process:

  1. We combine data with the help of our Skim Engine, from millions of different sources that can be both internal (email, support tickets, chat) and external (forums. blogs, product pages, social media) to create a live dataset for analysis. 
  2. We extract features from the text and conduct Sentiment Analysis to identify the main product features that are liked or disliked across a category. 
  3. The magic is done! At this stage, we will be able to deliver your insights. This intelligent sentiment data is modelled and presented via our analytics dashboard, on-demand.

Blogs that might interest you: