The rule's own list of what counts as financing distribution is broader than most people picture. It names, without limiting itself to them, advertising, the compensation of underwriters, dealers and sales personnel, and the printing and mailing of prospectuses to people who are not already shareholders and of sales literature. So the fee is paid by current shareholders and a substantial part of what it buys is directed at people who are not yet shareholders. That is the structural feature worth understanding about it, and it is not a criticism: the argument originally made for such plans was that attracting new money grows the fund and spreads its fixed costs across a larger base.
What the rule requires is process, not restraint, and reading it is the fastest way to see that. The rule contains no percentage figure of any kind. What it requires is that payments be made under a written plan describing all material aspects of the proposed financing of distribution, with all related agreements also in writing, and then a list of conditions on that plan. The plan must be approved by a majority of the outstanding voting securities if it is adopted after a public offering. It must be approved by the board, including by the directors who are not interested persons of the fund and have no direct or indirect financial interest in the plan. It continues beyond a year only if specifically approved at least annually in that same manner. Anyone authorized to direct the money must give the board, at least quarterly, a written report of the amounts spent and the purposes they were spent on. The plan is terminable at any time by a vote of the independent directors or of a majority of the outstanding voting securities. It may not be amended to increase materially the amount spent on distribution without shareholder approval. And the directors may only approve or continue it if they conclude, in the exercise of reasonable business judgment and in light of their fiduciary duties, that there is a reasonable likelihood the plan will benefit the fund and its shareholders.
So where does the familiar cap come from? Not from this rule. The numeric limits investors have generally heard of are set by the Financial Industry Regulatory Authority's sales charge rule, which constrains what a member firm may accept, rather than by the SEC rule the fee takes its name from. That matters for a practical reason: a reader who goes looking in Rule 12b-1 for the ceiling will not find one, and might reasonably conclude they have misread the rule.
One provision is unusually specific and is worth knowing because of what it closes off. Paragraph (h) of the rule forbids a fund from compensating a broker or dealer for promoting or selling its shares by directing the fund's portfolio securities transactions to that firm, or by directing to it any commission, mark-up or mark-down received from portfolio transactions placed elsewhere. In other words, a fund may not pay for distribution out of its trading business. Whatever it does pay for distribution has to go through the written plan the rest of the rule governs. The route the paragraph shuts would have buried the same cost inside the fund's trading costs, which no fee table breaks out.
The difference from a sales load is timing, and it decides who pays what. A sales load is deducted once, from your money, when you buy or redeem. A 12b-1 fee is charged to the fund every year, so its cost to you accumulates with the length of your holding period. That is why share classes exist in the combinations they do: a class with a higher front-end load and a lower ongoing fee and a class with no load and a higher ongoing fee can suit holders with different time horizons, and neither is generically cheaper. The fee is included in the expense ratio, so it is not an extra charge on top of that figure, and the fund's fee table shows it on its own line as distribution or service fees.