Everyone knows that distinctiveness can be inherent or acquired, and that some kinds of trademarks — product-configuration trade dress, notably — can “never” be inherently distinctive.
But not everyone knows that the same principle applies, though not (to my knowledge) utilizing the nomenclature of “acquired,” to trademark strength as well, as John Welch notes:
TTABlog 101: In determining the strength of a mark, the Board considers both its inherent strength, based on the nature of the mark itself, and its commercial strength, based on marketplace recognition of the mark. See In re Chippendales, 96 USPQ2d 1681, 1686 (Fed. Cir. 2010)
— TTABlog (@TTABlog) December 15, 2018
(See here for blogging here about the Chippendales case.) This fine but potentially important point recently came up in a case we litigated to a verdict involving a trademark that was very descriptive. (We even argued that it was generic, but the judge wasn’t having any and the jury, well… it’s complicated. And sub judice, so…. ) This took place within the Eleventh Circuit, where the two most important LIKELIHOOD OF CONFUSION factors are (1) trademark strength and (2) actual confusion. There was zero actual confusion despite years of market coexistence (and there was no survey), so the plaintiff hit the “strength” button really hard.
Throughout the trial and during closing argument, the plaintiff kept coming back to its mark’s supposed strength. “We did advertising! We gave away free samples! We sold a bunch of stuff! We sold even more of it! We did all the things!”
And our response, consistently, was that all this was nice, and it’s certainly what you do, but still the mark was very descriptive as a matter of law — in other words, inherently. Because there was no evidence in the record of whether all these activities had actually accomplished anything, i.e., resulted in greater recognition in the marketplace, the fact that these things were done did not support a finding that mark had, as we might say, “acquired” strength.

To non-specialists who are asking what I am about to answer, here is the answer to what you are wondering: The difference between distinctiveness and strength is the difference between whether the trademark exists and how good (“strong”) a trademark is. “Distinctiveness” is term of art that is synonymous with secondary meaning. If a purported trademark is not distinctive, either by virtue of its inherent ability to function as a trademark or because of public recognition (or both), its user earns no trademark rights at all.
In contrast, or, perhaps, further along the trademark life cycle or timeline or what have you, trademark “strength” is a way of expressing whether the trademark has a lot of, well, that distinctiveness. Yes, strongness (or, if you must, strength) is distinctiveness, in terms of what we’re measuring: Trademarkness. But once we have established that a mark is “distinctive” and is in fact a trademark, we stop using that term for purposes of expressing degree. “Distinctive” is a toggle: Yes or no? Yes = trademark. No = not a trademark, or at least not yet.
The same qualities, however, are certainly being measured when we ask, is that a lotta mark? Or merely a wisp of a mark? This is an important consideration not merely on the beach but in the scope of rights available to the trademark holder. When a strong mark steps up, junior users part way. They may not even come close to a strong mark in choosing their own.
If you have a big strong trademark, therefore, such as Coca-Cola or Sony, other LIKELIHOOD OF CONFUSION factors such as competitive distance, similarity of goods and channels of trade are much less important. When people see your trademark on something, they think of you and your product with the strong trademark on it, almost no matter what. (Trademark dilution claims for “famous” marks are essentially a prepackaged version of this concept.)
In contrast, if you have a weak trademark, you’re not allowed all that much exclusivity beyond your blanket. If it’s inherently weak, meaning highly descriptive, you’re lucky we don’t treat you as generic — not a mark at all. So the only way we will let you stop a competitor from using a similar mark is if it steps right your blanket, i.e., offers more or less the same product to the same folks the same way using pretty much exactly your trademark. In other words, you have to show that a lot more of the possible components of the LIKELIHOOD OF CONFUSION analysis are present in order to establish infringement.
Unfortunately, the Eleventh Circuit doesn’t seem to distinguish, as the Chippendale’s case does, between inherently strong marks and marks made strong by public recognition, and, in the latter case, what kind of evidence prove it. But perhaps, maybe soon, it will.