This week Canadian Transport Minister Omar Alghabra announced new measures to hold air carriers more accountable for flight delays. CTV ran a story detailing the differences between delayed flights in the U.S. and in Canada and the crux of it is that Canadians endure about 38.5% more flight delays when comparing major air carriers (we looked at 57% on-time in Canada vs 79% in the USA detailed in the news story).
We wanted to see if Polly reflects this in her data. The challenge is that we are comparing different datasets: government data with known flight delays vs social media posts of flight delay complaints. Also, there could be cultural differences in the level of complaining between Canadians and Americans. Still, we were curious to see what we could find. The entire study cost $10. It took 2 minutes to produce each report and then about 1.5 hours of further analysis.
We did a 3-month study for Canada and America and the question for both was simply “Is your fight delayed?” You can click on the links under each country to see the results for yourself. Note that, because America is 10x more populous than Canada, you need to multiply all Canadian numbers by ten.
First, we look at the Engagement number in the report to see how many people are discussing the issue. For Canada, the engagement on this topic was 74,468 (744,680 with multiplier):
 |  |
If the same proportion of Americans were upset, we would expect to see the comparable engagement level be just over 744,000. But it wasn’t. Americans complaining about flight delays was 492,698. The percentage difference between these two numbers is 50.25% more complaints in Canada than in the United States, and this is higher than the 38.5% actual difference in flight delays. So Canadians are complaining significantly more than the flight delays on their own can explain.
If you are a consultant on contract with or wanting to win a contract with one of these Canadian airlines, you would want to understand why Canadians have what appears to be outsized anger. The logical first step is to analyze the emotions and see if we see any differences here. We access the emotion diagram using a drop-down at the bottom of the report.
By clicking on “anger” for each sample (Canada and U.S.) we can isolate the number of people reporting they feel angry in each population. Again, we need to multiply the Canadian engagement by 10 to make it comparable with the United States. Doing this, we see that Canadians are almost twice as angry as Americans, with 12,924 (129,240 with the multiplier) expressing anger vs 64,476 Americans.
When we isolate the anger emotion, Polly also highlights only those conversations where anger is expressed. We do the same for the American conversations.
What becomes clear in the conversations is that the Canadian conversations expressing anger often speak about compensation (and the frustration trying to access compensation) whereas the Americans don’t mention a right to compensation at all. They are not being compensated, and they are less angry!
So if I were a consultant who has a contract with a Canadian airline, there are a number of further investigations I would want to conduct for them:
Isolate two lookalike samples of travellers and really probe the U.S. sample on compensation. What amends, if any, do American carriers make?
Test any above compensations (vouchers for restaurants, discounts off future flights) and use Polly to model out if these might cost less than the number of people making further claims for compensation.
Message testing to the best way to communicate with passengers when delays are expected.
Doing it with A.I. is going to be infinitely more cost-effective. Polly is here to help consultants working in small-medium enterprises succeed. You can get started here for only $100.
To satisfy yourself that Polly is credible, check out our testimonials page.