New Report Suggests That iPhone Loyalty Has Fallen to Its Lowest But Something’s Wrong

Alright, kids, here’s a lesson on news. Never read the headlines alone and believe it to be true.

Although maybe you guys already know since we explored that in our work-life balance and Faceapp articles.

Oh yeah, it’s time to de-bias some surveys again.

In the headlines: iPhone loyalty falls to an all-time low

CNET first released an article with the headlines “More people are trading their iPhones in for Androids, study says”, citing a survey conducted by BankMyCell.

The survey, conducted by BankMyCell, says that iPhone retention is down 15 percent over the last year. That number is from 38,043 people trading in smartphones on their site.

If you read only the headline from BankMyCell, you would get…

“Apple’s Decline in Sales During Q1-Q2 2019 Visible as 15.2% Less Trade-In Savvy iPhones Owners Are Sticking With Apple”

Dayuum, that’s a HUGE drop eh?

Survey data sources are wonky af

But hold on, let’s visit the site and see how they conducted the surveys.

Just scrolling down a little, we can now read that “2019’s trade-in savvy iPhones owners are -15.2% less loyal than CIRP’s 2018 survey”. And further down, we see this graph:

Is that… a different source of data set samples for each year? Yes, yes it is. Their 2019 data uses their own data, but 2018 relies on a report by CIRP, 2017 by Morgan Stanley, 2016 by UBS analysts…

To give you an analogy, that’s like tourists coming in to Singapore every year specifically to eat chicken rice. Every year, they would eat at different stores. From the year 2011 to 2018, they ate chicken rice and gave it a review on Google maps, giving them a range from 4 to 4.5/5 each time.

But in 2019, they ate a store that just somehow sucked a little. Maybe it’s one of those cheaper, average or less popular stalls. Then they put on Google reviews “Chicken rice standards have dropped in Singapore”.

It doesn’t make sense. The people each survey collects data from is different. We can just compare the CIRP data in 2014, which has customer loyalty at about 80%. In BankMyCell’s report, it shows 90%.

Scrolling a little bit more, we see this graph:

In 2018, the number is 70.1%. In 2019, it’s now 72.8%, a 2.7% increase for Apple brand of new devices for iPhone trade-ins.

Hold on, doesn’t this mean that if we use only customer data from BankMyCell, brand loyalty actually increased?

So at this point, I’ve pretty much rendered the rest of the report on BankMyCell useless. A pity, considering that the report looks very professional.

Possible bias? Or just poor statistics?

Is there some possible kind of bias in this reporting?

Well well well it would appear that Apple would be releasing their earnings announcement on 30 July 2019, about 2 weeks after this report.

…And while on first look it might look like some fear-mongering thing to temporarily shoot down Apple stock so that someone can short the stock, on second thought that doesn’t quite make sense. At this point, I’m not able to connect more dots, or at the very least don’t have the information to yet.

I’ll shrug it off as poor statistical skills rather than any bias .¯\_(ツ)_/¯