Again, let me say explicitly I am reading between the lines, but it is the closest sensible concept I know compared to what was said.
People in general often poorly understand the degree of capital expenditures made by companies, and I suspect those of us in the software vision may have an even more skewed vision since we see billion-dollar companies started with a laptop and ~50 square feet of office space, but when you have real capital expenditures required to make your product (even ignoring advertising, etc.), those little single-digit-percent wiggles right around the break-even point can be huge.
> ... American Airlines president Scott Kirby gave a peek into the lopsided economics of his company, saying half of its revenue last year came from the 87% of its customers, who only flew the airline once.
The flip side is 13% of the regulars providing the other half of the revenue. But those passengers are already familiar with you and using you regularly. The "light buyers" comment is showing that the long tail matters too, and that they aren't that familiar with your business/brand, and also are unlikely to hang out/be reachable in the same way those 13% are.
The regulars are a known base - it is the irregulars who do need focussing on.
Having not read the book referenced, I haven't seen the data. But this seems to go against the Pareto Principle AKA 80/20 Rule which suggests 80% of your sales come from 20% of your (heaviest) customers, and against my own anecdotal experience as a marketer.