dhanson865 wrote:In an ideal world three or more of each reviewed product would be purchased from three or more retailers. No undue influence from a manufacturer, distributor, or retailer would be possible under that model.
Take for example the situation where a 700 to 1000 watt PSU is sent in for review when we are more interested in seeing the review of the 400 to 600 watt models. If the source of the review material chooses what gets reviewed it already limits the outcome of the review process even if SPCR is as impartial as they can possibly be.
Doug, you've touched on a couple of key points:
1) It's true that sample submissions (donations, mostly) by mfgs, distributors and retailers generally set the limits or range of the products we test. As long as our revenue stream remains limited, there's no easy way around this, even though it is not ideal.
2) Reader funding of product reviews could be workable and has a grass-roots kind of charm. This is probably worth further thought/discussion/development. It happens occasionally now anyway -- readers send products bought w/ their own $ for us to test -- but not often.
3) Your mention of reviewing up to 3 samples of the same product brings up a previously untouched issue:
The most "costly" part of SPCR production is labor. Early on in this discussion, people brought up bandwidth/hosting costs. They are not nothing at >$500/mo but a pittance compared to the cost of the time reviewers spend on the editorial/testing work. Our review process is demanding & complex & time consuming; it is what gives us the data to write the kind of reviews we do. Let me give you some examples of the time required for:
--PSU review -- on average, 2-2.5 8-hr days, depending on complexity. If we checked 2 more samples against the 1st, then another day would have to be added.
--motherboard -- 2 days
--heatsink -- 1-1.5 days
--PC -- 2 days
--case - 2.5 days
Often, we end up spending considerably more than the above, because we explore some unique aspect of the product that seems essential to us, something that falls outside our usual testing parameters.
Even if we pay ourselves just
minimum wage, the labor cost of almost any review is ~$200. How many reviews do we put up? Around 85 full articles in 2009, ~7 articles a month. Considering that we need to do other work to supplement our income, time/energy is our single most valuable commodity. The only way to change this is to increase site revenue w/o increasing our workload.
So getting back to the idea of 3 samples of each product to test -- only if the extra samples were used in spot checks to confirm the first.
PS: Aris -- your last statement is so utterly, totally wrong in every way.
Stay out of this discussion if that's the kind of comment you're going to continue making. There will be no other warnings.