Hiya, Andrew. The difference in the yDNA haplogroup designations isn't an FTDNA marketing strategy, it's the result of what is actually being tested.
The 23andMe test you took employs the same type of microscope-slide-looking thingy that all other inexpensive direct-to-consumer DNA testing companies currently use; a microarray genotyping chip. It's "programmed" so that when a prepared (read: enzymatically fragmented) sample of DNA is washed over it, certain pieces of the DNA attach to certain points on the microarray.
Those pieces are SNPs (single nucleotide polymorphisms) and they live, based on the particular version of the human genome map being used, at very specific places on specific chromosomes.
All current microarray chips test for a (varied) number of places on both the Y-chromosome (if present) and the mitochondrial DNA molecule (mtDNA). Some don't report it (i.e., AncestryDNA) but they all test for it: it's integral to the chip. The marketing hype, if any, is there: none of the major DTC DNA testing companies actually do multiple tests for your $69 fee. It's one, single test on one, single microarray chip (which itself is reusable for a few different tests, BTW).
All of these chips, in addition to the standard, basic set of yDNA and mtDNA SNPs that come included, are customizable with the ability to program them for, roughly speaking, about 50,000 other SNPs that the testing company can define as they see fit.
For example, 23andMe chose to program their Illumina GSA chip to look at--again being specific to the version of the human genome map in use--position 12,108,819 on the Y-chromosome. That point, among the 58 million on the chromosome, is CTS2243, and its value would be positive for C, or cytosine, for that haplogroup.
Haplogroups, for yDNA and mtDNA, are defined by testing either positive or negative at a specific SNP. R-CTS2243 is, under the dated ISOGG lexicon, R1a1a1b1a3a2a1. With haplogroups, these things are ordered hierarchically, with older variants being "parents" of newer. R-CTS2243's parent is S223/Z287. Z287 is believed to have bifurcated (split off from its ancestral R-Z284)...about 4,300 years ago.
Immensely fascinating to geeks like me, but of only limited value to genealogy.
You can get deep dives into yDNA SNPs at FTDNA, from individual SNP tests and SNP panels (based on high-level haplogroup) all the way up to the flagship BigY-700. The latter, just an aside, has allowed over two dozen test-takers with different surnames (my Williams line among them) to understand that we all shared a common male ancestor somewhere on the eastern edge of the Irish Sea--be it Wales or northwest England or Scotland--most probably sometime between 800AD and 1200AD. Interesting, but of only limited value to genealogy.
Haplogroups and haplotypes are very different things. Haplogroups are useful in genealogy to disprove an hypothesized relationship, but they cannot be used as positive evidence of a relationship. "Haplotype" refers to elements of your very specific genome.
What FTDNA offers in their basic Y-37 through Y-111 tests is an examination of Short Tandem Repeats (STRs). A $69 test on a microarray chip can't do this. An STR is what's called a microsatellite, essentially positions of nucleotides that are repeated up to dozens of times in a row. Think of an office copy-machine...sometimes the machine gets stuck and it stops copying too soon, or it keeps going and generates more copies than you expected.
STRs exist in autosomal DNA as well. In fact, that's what forensic and parentage tests examine. The reason is that STRs are more freely able to mutate. Extra copies at those loci don't upset the applecart. You start messing around with important protein-encoding genes, well...the prognosis for species survival ain't great.
STR mutation can give us a far more proximal idea of relationship than can haplogroup SNPs (again, barring an ultra-deep dive with FTDNA's BigY-700). FTDNA does this STR testing with Sanger Sequencing, different than the wash-and-rinse microarray chip tests. At 111 STR markers, FTDNA predicts that an exact match has a 95% chance of connecting you to a common ancestor five generations ago.
Way better than only a haplogroup 4,300 years ago.