Yesterday I wrote an article about a fact that has been widely overlooked: Apple's FY2017 Q1 was 14 weeks long rather than the usual 13 weeks. Some people have noted this fact, but there has been an effort to explain it away as a "slow week".
I'm perfectly willing to admit that it was a slow week. Let's just assume for the sake of argument that it was a slow week. The question is, how slow is a slow week for Apple?
Total revenue for FY2016 Q1 was $75.9 billion, while total revenue for FY2017 Q1 was $78.4 billion, an increase of $2.5 billion, or 3%. Total revenue for FY2016 Q4 was $46.9 billion, so an average week in FY2016 Q4 was $3.6 billion. That's well above the $2.5 billion difference. However, perhaps that average number is too high to use. Over the past 2 years, Apple's slowest quarter was FY2016 Q3. This seems like a good place to look for a slow week. Total revenue for FY2016 Q3 was $42.4 billion, which is substantially lower than either FY2017 Q1 or FY2016 Q4. Thus, I think it's fair to take Q3 as roughly representative of a slow week in Q1. The average weekly revenue for FY2016 Q3 was $3.3 billion. But that's still higher than the $2.5 billion difference between Q1 in 2016 and 2017! Of course, some weeks in FY2016 Q3 had to be higher than others, but it seems reasonable to think that even a slow week in Q3 brought in at least $2.5 billion.
The inescapable conclusion is that even if the 14th week in FY2017 Q1 was one of the slowest weeks of the past two years, that's still enough to account for the difference with FY2016 Q1. Ergo, FY2017 Q1 did not in fact represent a "return to growth", as so many media outlets have incorrectly reported.
I'll leave it as an exercise for the reader to perform the same kind of calculations for individual product revenues and unit sales. I suspect you'll find that most of the so-called growth disappears.
To end with a positive, it's worth noting something I've never considered before and perhaps you've never considered before either: the staggering amount of money that Apple makes in only one week. I estimated that the slowest of the slow weeks brought in at least $2.5 billion. One can only marvel at that number. If I could bring in that much in one week, I'd retire after one week. Really, I'm not that greedy, so I'd retire after one day.
One could argue that FY2016 Q1 had an advantage over FY2017 Q1 because of the iPhone release dates. FY2016 Q1 started on September 27, 2015, and FY2017 Q1 started on September 25, 2016. iPhone 6s was available on September 25, 2015, and iPhone 7 was available on September 16, 2016. Therefore, FY2016 Q1 may have captured more of the initial iPhone sales than FY2017 Q1, because the first week of iPhone 7 sales occurred during FY2016 Q4 instead of FY2017 Q1. Nonetheless, this does not change the fact that FY2017 Q1 was not a growth quarter. It doesn't change the numbers, doesn't change the weekly averages. We're only talking here about how to explain the numbers, and that's an entirely different argument. In any case, the iPhone release dates have nothing to do with Mac sales or iPad sales, so they can't possibly explain away the 14th week advantage for those.