On January 28, 2021, Apple increased the number of city labels shown on its maps. This change arrives nine days after Apple had decreased the density of its city labels.

The increase is most pronounced in rural areas. A few examples:

The added cities in the rural areas above are a welcome change—and make the map much more useful.

That said, I wish Apple could find a way to increase the density of highway shields at this zoom. Looking again at our West Texas map, it’s surprising to see no shields in this view—especially for Interstate 10 and Interstate 20:

Within urban areas, the density of city labels is unchanged. But there’s a noticeable increase in the surrounding areas:

I like the overall density Apple has arrived at in the zoom above. Given what’s shown (and not shown), Apple appears to have developed an algorithm that isn’t entirely population-based—and it’d be interesting to know what else is driving it.

That said, I’m surprised by a number of locally-important cities that Apple ended up dropping from this zoom, including Oroville, Paradise, Lodi, Livermore, and Auburn. It’s equally surprising to see Monterey missing (likely collided out by Salinas’s unfortunate bottom-left placement).

Chicago (below) is another example of a major city that saw no new labels added directly inside of its urbanized area—but with plenty of labels added on its outskirts:

While the density above is generally good, it’s surprising to see such a large gap in labels in Chicago’s northwest suburbs—specifically, Lake and McHenry Counties. To that end, I was surprised to see Apple ultimately drop Crystal Lake from this zoom.

I was also surprised to see Apple ultimately drop Belvidere, Peru, Sterling, and Streator. On the other hand, it was impressive to see Apple add the tiny—but heavily visited—cities of Wisconsin Dells and Lake Geneva. Perhaps Apple’s labeling algorithm is incorporating iPhone probe data?

Meanwhile in the New York area, it was surprising to see Apple add Bronx as a separate city alongside New York City:

I was also surprised to see Apple ultimately drop Kingston, Peekskill, Torrington, and Newark (Delaware) from this zoom. But, in general, I really like the density Apple arrived at here.

Going down the East Coast toward the federal district, it was bizarre to see Apple add Chevy Chase Section Three—a legit incorportated village, despite its doofy-sounding name:

The labeling of Chevy Chase Section Three over, say, Silver Spring (population 800 vs. population 82,000) at this zoom is something I don’t quite understand—and it’s why it’d be interesting to know what other signals are feeding into Apple’s city labeling algorithm.

Apple’s changes also appear to have been extended to Canada, where there’s a noticeable city label increase in Southern Ontario:

However, if you zoom in and look closely at the areas along the U.S.-Canadian border, you’ll see that Apple is apparently applying different city display rules for Canada than it is for the U.S.:

Up until that last screenshot, all of the screenshots I’ve been sharing are from Apple Maps’s eighth zoom-level (z8). But if you zoom in to z9, you’ll see that Apple has increased the density of city labels even more dramatically.

For instance, here’s Central California at z9:

On the old map, only cities with populations greater than 2,500 appeared at z9. But now Apple appears to be showing cities of any population at this zoom.

The z9 changes are even more noticeable in the Midwestern U.S.:

And the further you get away from Chicago, the more dramatic the changes are—such as here in Indiana:

And here in Illinois:

Part of the reason why these changes look so dramatic in the Midwest, as opposed to California, is because there’s generally a much greater and more even distribution of individual cities in the Midwest:

That said, I wish Apple would add a bit more hierarchy to its city label styles at z9. In the view below, for instance, some cities have populations of just 100 or so, while others have populations approaching 50,000. And yet they’re all shown in the same text size and weight:

While traveling through an unfamiliar region, it’s often useful to get a sense of which cities are more likely to have clusters of businesses, like restaurants and gas stations (without explicitly searching for them). And you just don’t get a good sense of this by glancing at the map.

This is something that old-school, paper road atlases excelled at. For instance, here’s that same area in Illinois in the U.S.’s bestselling road atlas. Notice how many levels of hierarchy are being used in the map’s city labels:

And now look again at the area on Apple’s map—everything looks the same:

While I’m not saying that Apple should copy the paper road atlas (it’s far too dense and busy), it’d be nice to see an additional level of hierarchy among Apple’s city labels at z9.¹

Put another way: with the city label increase, the map is now much better at showing us “what’s here?” in a given area. But with so many cities added, it’s no longer quite as good at showing us “what’s important here?”—and an extra level of hierarchy would improve this.

Hierarchy aside, it’s remarkable that Apple is able to show twice as many cities as Google in this area without making the map look cluttered:²

Now if only we could get a few more shields...

__

1  I think it would be ideal if Apple took its label style for the smallest cities (those with populations between 0 and 50,000) and broke it into two groups: one style for cities between 0 and 5,000, and another style for cities between 5,000 and 50,000. ↩︎

2  It’s fascinating to see how many cities Google has removed from its map of this area over the years:

I’d rather have the 2013 density over the handful of cities that Google offers today. ↩︎