Here is BrandBucket's way:
1. Recruit sellers and see what they've got.
2. If a name is great one (a killer name, a pronounceable LLLL.com etc.) that either is hard to obtain or costly to obtain, then list them. They get $10 for listing, traffic that those names normally have, natural buyers that might come with them and 30% of sale if it happens.
Math works this way:
Cost to obtain very high quality names: average $300. Listing price for this kind of names: average $5000. Sell %: 3%. If you have 100 of this names, upfront expense of $30K, renewals 1000, logo payout $300, revenue $15K annually, so BB would make around 45% return before other expenses.
Now if the cost burden is shifted to sellers, then upfront cost $0, renewals $0, revenue 3x5000x30%=$4,500.
Obviously, $4,500 on 0$ investment is better than $13K+ on $30K, as that is "only" around $8.5K incremental on your "idle" $30K in the bank.
It is completely different dynamics on brandables that are not unique. Meaning there are literally hundreds of thousands of equivalents for them.
BB can buy 100 of those at $2000 average, sell 3 a year at $2K average, minus renewal and logo, for a bit under $5K annually
With your 100 names, though upfront is $0, they get only 3 x $2K x 30% = $1,800
Obviously, $5K on $2K upfront investment is better than $1.8K on 0$ investment. Why? Because the incremental $3.2K on $2K that you have on your bank account is still very healthy return.
So, if they do have "idle" 30K in the bank account from the example above, they are better off filling up the segment of generic brandables with their own 1500 names for around $50K annual earning than using it towards quality names and making around $8.5K with it.
So to summarize: they don't want your good names as for that segment, they'd rather have names of owners, managers, ambassadors and Matmor, whatever his capacity in that organization is. They will still take some, not to completely turn you off, but don't look for a logic here. They only want your great names.