Thanks for your input, enigma1, explorador, and MrFewkes
It shouldn't show (for long) as duplicate content because the link will is no longer generated from within your domain, so google should clear it up.
And that could be the case. I really need to do some analytics digging into SE referrals for specific product pages which have gone out of stock, and see how long it takes SE referrals to drop (i.e. for them to recognize the content of /widgets/round-widgets/blue-widgets is now actually the content from /widgets/round-widgets/ and to adjust their rankings accordingly).
However, the SEO concern I have is less about lost rankings for specific products than with the report I'm seeing in Webmaster Tools, which says I have 40,000 duplicate title tags and description tags.
Does that make sense? I guess to say it another way: if we are filtered by Google for duplicate content for [blue round widgets], that is an extremely long tail term for us and so the loss of traffic won't have much revenue impact.
But from a higher level point of view, I am concerned if Google looks at our site as a whole and says, "Wow, they've got 40k duplicate pages. The overall quality and care of the site is less than competitor X, so we will reward the 'cleaner' site with better rankings for [widgets]."
I have to imagine that sort of high-level qualitative comparison is among the signals used in the algorithms, right? Otherwise, it doesn't make sense for Google to include it in Webmaster Tools.
Also depends how often the catalog is updated. If you update often (eg: 2nd hand items, in/out of stock different products) is not a good idea to let the out of stock pages flooding the SEs index.
We do update the catalog frequently with both new and used items. Can you say more about the issues of leaving a lot of "out of stock" pages open for indexing?
It sort of feels like a "damned if you do, damned if you don't" scenario. If we leave the behavior as is, we're leaving it up to Google to clean up our presence in the SERPs, which means periods of duplicate content. If we change the behavior, we risk flooding the SERPs with out-of-stock pages, which may frustrate users who are ready to buy, only to find an out of stock/suggestion page instead.
If we do a 301, my concern is the SEs not re-indexing the pages since we told them we had permanently moved them to a different URL.
I am also a proponent of the "sign up" alerts. I think newegg and zappos (to use two prominent ecommerce sites as examples) both have sign up/notification alerts for products or product combinations that are out of stock, and they certainly must have data to support its efficacy.
But there is huge concern for the UX of landing on a page that then disappoints. While I can completely understand those concerns, I still think there is value to a user in an out of stock/suggestion page, as you mentioned in your response.
But I am fighting an uphill battle in trying to shift to an "out of stock" model.
I can totally relate to your statement that customers are highly driven by their own bottom line, but I've been on the receiving end of the "In-Stock -> Out of Stock" checkout process, and that's a nightmare in and of itself.