Re: [wg-c] registry contracts

[Eric Brunner <brunner@world.std.com>]

Caroline,


One issue with the for-profit model is the probability of consolidation.

If operators have no title interest in registry data, and the mechanism
for renewal is competitive bid on price with nominal non-price conditions
which resolve to capitalization, then at each rebid, each operator will
have only their initial capitalization and accumulations, direct as well
as indirect, from registry operation.

If the for-profit registry business model is successful then higher
capitalized entities will out-bid existing operators at each rebid
event.

As a corollary, marginally capitalized initial TLD delegatees can expect
only one period of capital accumulation.

As a corollary, for-profit registries will seek to diversify to mitigate
risk of loss of delegation, and as a precondition for access to investment
capital.

This leads to for-profit registries operating as business units of larger
entities, participating in shared risk and shared access to capital.

Additionally, if no mechanism is allowed to exist to support a policy of
non-consolidation, then higher capitalized entities will accumulate registry
delegations.

Clearly, if operators have a title claim to registry data, then the cost
of displacement of an operator-in-place is dramatically inceassed, and
the liklihood that marginally capitalized initial TLD delegatees can expect
to sucessfully compete for delegation renewal is increassed.

If title to data is not assumed, then I expect the for-profit segment of
the DNS registry industry to rapidly consolidate to a stable oligarchy
market in which registry operations will be non-strategic, or a market
access tool.

If title to data is assumed, then I expect the same result by slower means,
as buy-outs rather than the lower cost bid-outs will be required to meet
the revenue stream or market access goals of the consolidators.

One thing we can expect is that the venture capital attraction of registries
will not scale. If registries offer indistinguishable products, names for
sale, the interest in and ability to capitalize registry operators is not
infinite. Whatever the capitalizations are of the initial delegatees, and
the candidates known to WG-C have unknown capitalizations, participation
by VCs in new operators after some initial period is not likely to be large.

The prospect for the for-profit registry segment is bimodal capitalization
with a highly capitalized oligarchic dominant group not at risk from revenue
fluctuations from periodic accumulation or loss of registry delegations,
and a marginally capitalized group servicing marginal numbers of customers
and highly at risk to revenue fluctuations.

---

Another issue with the for-profit model is the probability of labor cost 
containment.

I suppose someone needs to point out that labor-intensive registry operations
will move to low-cost labor pools, with only necessary bits of customer
support comming out of wage-protected labor pools. NSI wouldn't have scored
points with the USG if they'd been a shell with most wage-earners off-shore.

Assuming that registry operators are cost-sensitive, the race to the bottom
on comodity back-end labor cost will preclude registry operators from more
than a marketing presence in most markets.

Were ICANN to make the policy choice of preferential selection of late
accessing regions for comodity back-end registry labor, with some policy
guidance on the issue of labor protection, I'd be quite satisfied. Leaving
it to "natural forces" -- well I'm sure that position has its advocates.

Registry operations as a cottage industry has a lot of attraction to those
who hope a benevolent free market will extend to them some priviledged
stay of execution. However, piece-work in the DNS industry is unlikely to
follow any path other than that of piece-work in the garment trade, with
status marks and non-status labor. Tribal governments are always looking
at the labor cost and availability vs portability of economic activity, we
can do call centers (labor cost < 2 x minimum_wage, english, and office
skills), but remove english from the requirements and if portable, the type
of activity won't last in the Americas north of Mexico.

Off hand I don't know why the trademark holders care one way or another,
either about operator consolidation or labor selection. I suspect that you
may prefer the weakest operator model to be selected, as it will be the
registries that realize your remedies. If however you posit registries
generally cooperative on the issue of remedy application, and ICANN the
mechanism address the exceptional registries, then that model may not be
today's cup of tea. I don't expect that your consideration is limited to
narrow Constituency self-interest, mine isn't, we're not going to rush
to be the first consolidatee.


Cheers,
Eric

	In theory there is no difference between theory and practice,
	but in practice there is.

(signature credit Patrik Faltstrom)