Speaking for myself, and in more general terms that apply to a lot of licenses:
There's many factors at play when it comes to what a company will license, a few of which are a projection of how much the license will make,
who the company is requesting the license,
and how active the brand is.
When it comes to the projection, a licensee (the person or company who pursues and holds the license to make product based on a company's brand) must estimate their sales across the term of the license. A seasoned company will know how to create that projection, but it's not an exact science because some brands perform better with different licenses.
No matter what that projection may be, the licensor (the person or company that owns the property) may only be willing to license to them if they are a reputable company. There are so many instances where licensees have failed to pay royalties. No licensor wants to chase down what they are owed. So the licensor has to be careful who they license to.
Then for everyone involved, the licensee may or may not perform well based on the current success and popularity of the brand. And even then, as mentioned above, some products just don't sell well for certain brands, even though they may have performed well for other brands. Sometimes it's a complete mystery as to why some things take off and others do not.
There's also exclusivity at play. One company may want to do posters. But the licensor may already have an exclusive agreement with another company to do posters, or a deal that includes multiple printed items including posters. So some licenses aren't available just because of existing deals.
There's also a time VS expense factor as well for everyone involved.
As an example, let's say there's a pencil company that wants to make He-Man pencil sets. Let's say the retail price on a set is 3.00, and the wholesale price on that (the price the pencil company sells the sets to retailers for) is 1.50. Royalties are usually around the 10% mark on wholesale (this is no secret. you can Google search this kind of info). That means Mattel would make 15 cents per set sold.
This pencil company projects that they can sell 1000 sets per Quarter 1, 3 and 4, and 5,000 sets for Quarter 2 (back to school). So per year, they feel they can sell 8,000 sets.
Mattel's royalty is 8,000 * .15 = 1200.00 a year
The problem there is, 1200.00 a year for Mattel is next to nothing. Just the time involved in the Legal paperwork and review, and the time any employee takes each quarter to review the product almost eliminates the profit, or possibly eat up time MORE valuable than the royalties the license generates.
So, in this case, Mattel may not issue such a license. It may be a cool product and people may really want it, but it's not a practical business decision.
Let's do an example of a projected success turned failure.
Let's say a major t-shirt company wants to do 3 different He-Man shirts. They think they can sell 5000 of each design, per quarter. So that's a total of 60,000 shirts for the year.
The shirts retail for 9.99, with a wholesale of 3.99. At 10% royalty, that means Mattel would make .40 cents a shirt
So per year, Mattel is projected to make 60,000 * .40 = 24,000.00
That's a lot more appealing for royalties.
BUT, let's say to everyone's surprise, the shirts don't sell. Instead of selling 5000 of each design, they only sell 1000 of each design. That immediately drops their royalties to 4,800.00 for the year.
Not only would that disappoint Mattel, but it would really disappoint the licensee who may be sitting on unsold product they had to pay to produce. This would cause Mattel to be concerned about future t-shirts licenses, AND it would cause this major t-shirt company to possibly never want to license He-Man shirts again.
There's all sorts of scenarios you can come up with that outline the difficulties faced by licensing, and to explain why you don't see more of it for any given brand.
There's things like royalty advances and contractual guarantees the licensor can demand from the licensee to help protect the licensor from having their time wasted.
but if the license performs poorly despite paying an advance and/or guarantee, it's doubtful that licensee will return in the future for a license on that brand. And other licensees have some idea of how products perform in their same industry. So it may cause other companies to be weary of a particular license.
And if the company is a young upstart and/or an unstable company, despite a guarantee, they may close up shop and the licensor would get screwed in the end.
Also, a licensee can't just do what they want, as they need to follow standards and guidelines in place from the licensor to insure the product has a look and feel that is in line with the image of their brand. This is why a lot of companies have styleguides that have a set of approved images and design standards licensees must follow. That's also why you see a lot of the same images show up on different products. It also cuts down on design time for the licensee and approval time for the licensor. And no matter how good of a product a licensee may create, there's no guarantee it may sell.
Licensing is a tough business. I could go on and on about it, but hopefully that sheds some light on the topic from a guy who's both licensed products and also helped create licensor styleguides for many years.