Business

How is McDonald's doing?

Ask Twitter

By Pradeep K. Chintagunta     
September 1, 2015

From: Magazine

McDonald’s was in the news last spring for reasons other than its financial woes. The Oak Brook, Illinois-based fast-food company made a flurry of announcements that ranged from debuting a line of clothing at a fashion show in Sweden to raising employee wages at its company-owned stores in the United States. These moves have been analyzed in terms of their financial consequences, but what about the marketing associated with them, and the social media reaction to it?

McDonald’s made three marketing-specific announcements in March alone. First was the “imlovinit24” campaign—a global marketing effort that took place simultaneously in 24 cities in 24 countries on March 24. The effort included, among other things, converting a toll booth in Manila, Philippines, into a McDonald’s drive-thru and offering drivers a free toll and breakfast, and orchestrating a surprise performance in Los Angeles by the R&B artist Ne-Yo. The campaign was targeted at reaching a younger, global audience—a group that the company is fighting hard to attract and retain. On the plus side, such a move could generate buzz for the brand, globally and otherwise. It could also introduce potential new customers to McDonald’s products, and existing customers to products they may not have tried before. From an operational perspective (taking into account the extent to which these events went smoothly), McDonald’s could potentially benefit from the campaign. Of course, the return on this strategy is difficult to compute.

Chart showing volume of tweets about McDonald'sThe second marketing initiative took place a few days later, on March 30. McDonald’s announced that it was testing the possibility of offering its breakfast items, such as the Egg McMuffin, throughout the day rather than only during a fixed period in the morning. It’s a move that both consumers and business folk have been calling for, given (1) the appeal of the breakfast items, (2) increased competition from other fast-food and fast-casual restaurants for this market, and (3) the fact that it provides another reason for potential customers to walk into a McDonald’s. A key downside to this type of move, however, would be its impact on the operations within a store, i.e., whether it compromises the employees’ ability to deliver products quickly and at designed quality levels. If adding these breakfast items detracts from, say, speed of service, such a move could have deleterious consequences.

The third marketing initiative also launched at the end of March. McDonald’s announced that the company would give patrons a free small McCafé coffee during breakfast hours from March 31 through April 13. Clearly, this was an attempt both to get regular customers who were accustomed to buying coffee elsewhere to sample McDonald’s coffee, and to draw those customers into purchasing other items on the menu. In addition, McDonald’s announced that it would hold sampling events at transportation hubs and other high-traffic locations, where it would give free coffee to commuters in hopes of convincing them to buy coffee from McDonald’s rather than from competitors such as Starbucks.

Chart showing sentiment of tweets sent about McDonald'sOf course, people decide whether to eat at McDonald’s based on the quality of its food (at a time when many people are moving away from fast food in the direction of more natural and organic offerings), its value, and its convenience. Unless the company makes strides on all these fronts, it will continue to struggle in the current environment. As Dominic Toretto (played by Vin Diesel) in Furious 7 puts it, “This time it ain’t just about being fast.” McDonald’s would be well served by focusing on all the dimensions of success, not just operational efficiency.

But what does the data say?

What impact did these marketing moves have on social media, specifically Twitter? (As with most of the analyses involving Twitter that I’m writing about, my thanks go to Yogesh Kansal, senior analyst at LatentView Analytics, for making this analysis possible.)

Let’s take a look at the volume of tweets corresponding to the fast-food chain’s actions. The data in figure (A) represent a random sample of tweets about McDonald’s (representing a fixed proportion, say 1 percent, of the tweets about the company each day). We will look at the time period from March 24 (the day of the “imlovinit24” campaign) to April 2, shortly after the “breakfast all day” and “free coffee” announcements.

The chart suggests that the later announcements of the all-day breakfast and the free-coffee giveaway had a bigger impact than the 24-city campaign, at least for this time period. The next question is, what happened to the sentiment (either positive or negative) expressed in these tweets over that time period?

Word cloud showing frequency of words in tweets about McDonald'sFigure (B) shows the sentiments expressed in tweets throughout the period analyzed. Yellow represents positive sentiment, blue represents negative sentiment, and gray represents neutral sentiment. Overall, it appears that a majority of the sentiments expressed were neutral. The number of positive sentiments shared exceeded the negative ones. The time-series pattern is also interesting: the company received a bump in positive sentiment and a reduction in negative sentiment around the time of the marketing announcements. So while the “imlovinit24” campaign did not generate many more tweets, it appears to have shifted sentiment in favor of the company. The later announcements appear to have increased both the volume as well as the positive valence of the sentiments.

Next, we take a closer look at what is being said about the company in the tweets themselves. For this, we first analyze the words in the tweets, and then depict these words in increasing font size corresponding to the greater frequency of their mentions.

Figure (C) indicates that the word “breakfast” gets a large number of mentions. To see if this is driven by the announcement, however, one needs to look at a time period without such an announcement. Indeed, repeating this analysis for March 9–18 reveals the presence of the same term but with a much lower frequency
of mentions.

We also see, in the data from March 24 to April 2, the terms “imlovinit” and “neeyocompound” show up, which is reflective of the campaign on March 24. (Those terms hadn’t been mentioned between March 9 and March 18.) Finally, other initiatives also appear to have triggered some conversations, such as those about “workers,” a word that shows up in the figure.

Chart showing volume of tweets about McDonald's compared to Burger KingIt’s instructive to get a sense of what was happening with other firms during the March 24–April 2 time period. Figure (D) compares the volume of tweets about McDonald’s with the volume of tweets about Burger King. The figure shows relative movement in the volume corresponding to the McDonald’s announcements. Such effects are absent in the Burger King volume.

Next, we compare the overall sentiments for the two firms over this time period. Figure (E), on the next page, shows the word cloud corresponding to Burger King during the announcement period of interest.

This figure suggests an entirely different focus for tweets about BK. This is not entirely surprising, since the sets of tweeters for the two companies show little overlap, as the Venn diagram in Figure (F), also on the next page, illustrates.

Word cloud comparing tweets about McDonald's and Burger KingFigure (F) also shows that McDonald’s has a greater positive sentiment associated with its tweets and a lower negative sentiment as compared to BK. Again, these numbers in isolation don’t say much about the impact of the announcement, so one might want to know the sentiments associated with the two firms earlier, before the announcement. In early March, the positive sentiment level for McDonald’s was 25 percent while Burger King’s was 14 percent. So even though one may not be able to draw any statistical or scientific inferences from this comparison, at least in terms of directionality, McDonald’s appears to be better off after the announcements.

In summary, the events held March 24 do not appear to have had a big impact on tweet volume compared to the announcements made March 30 and 31. However, all three events seem to have generated positive sentiment among people posting tweets. The word analysis also indicates the firm succeeded in getting some traction for the messages it was trying to communicate. Such is the power of eggs and coffee.

Chart comparing tweets about McDonald's and Burger KingPradeep K. Chintagunta is the Joseph T. and Bernice S. Lewis Distinguished Service Professor of Marketing at Chicago Booth and blogs at the Kilts Center Faculty Blog.
 

More from the Magazine

An error has occurred

An error has occurred