| 2:51 pm on Sep 28, 2003 (gmt 0)|
Depends on the type of affiliate, not all use it in the same way.
From an affiliates point of view I would not limit it to 5, as you may not have the best performing 5. A managed approach to who it goes out may stop the area becoming spam. I can't ever see it being as bad as email marketing.
| 3:00 pm on Sep 28, 2003 (gmt 0)|
Why? Maybe the sixth affiliate will turn out to be more successful than all the others put together.
Anyway, affiliates don't necessarily use the datafeed to create thousands and thousands of webpages.
They can serve the data up as search results, use the datafeed to identify a subset of relevant products, use it to identify links which are no longer valid, use it to keep existing links up-to-date (e.g. price), use it to speed up or automate the process of creating individual links, etc. etc.
| 7:05 pm on Sep 28, 2003 (gmt 0)|
I think Mayor's concern is very real.
Numerous affiliates are using datafeeds to create megasites that are little more than clones of assorted merchant catalogues. I've seen affiliates claiming to have several hundred thousand pages indexed by Google, and I consider their claims plausible. "Vomiting into Google", one consultant calls it.
Some merchants are starting to see datafeed clone sites as a detriment, not something to encourage, because they often push the merchant's real pages down in the SERPs, and cause the merchant to pay commissions for sales to traffic that they were formerly reaching without benefit of affiliates.
| 10:04 pm on Sep 28, 2003 (gmt 0)|
Yeah, but a merchant shouldn't have an affiliate program if they begrudge paying that money. The commission structure should ensure that the transaction is always profitable, and therefore the merchant should be HAPPY to pay the comissions in exchange for higher sales.
Lots of affiliates can boast more traffic, higher SE rankings, or simply a better interface than the merchant's own site, and IMO they're doing the merchant a huge favor. Those sales wouldn't have gone to the merchant without the affiliate's help; in most cases they would have gone to a competitor.
| 10:41 pm on Sep 28, 2003 (gmt 0)|
One thing I wonder is if we're looking at a snowball rolling downhill. Let's say I put up an affiliate site with 10 pages for a merchant. The next person puts up 20 and beats me out on the numbers. So I put up 100 pages. My competitor hops on a datafeed and puts up 1000. I do likewise and put up 5000. They one-up me with ten clone sites and 50,000 pages.
Along the way, some good affiliates get buried and give up, or seek out the merchant's competitors. The merchant's competitor starts spewing out data feeds to anyone who will take it in order to regain a presence in the serps. Did I hear buckworks say several hundred thousand pages?
On top of all this, the merchant gets a shady reputation because they lose control of affiliates putting up all those clones that surfers assume are the merchant.
I like datafeeds, they are a great marketing tool. But I'm wondering if some of them are getting out of control. I hope not, but I'm just wondering where these things are heading.
| 11:30 pm on Sep 28, 2003 (gmt 0)|
|a merchant shouldn't have an affiliate program if they begrudge paying that money. |
Merchants don't mind paying commissions to affiliates who actually bring NEW business. But some -- not many, but some -- are starting to take a closer look at how some affiliate activities mesh with what the merchant is already doing. Sometimes things that work well if one or two affiliates do them don't work so well if 200 are doing them.
One merchant (Tiger Direct) recently decided to stop allowing affiliates to use datafeed info, and one of the biggest reasons cited was the need to maintain corporate control of their trademark/branding, not have it diluted by copycat affiliate sites.
| 1:27 pm on Oct 3, 2003 (gmt 0)|
I think datafeeds are a great way to promote products if you are an affiliate.
If you look at some special products you can see more than 100 Amazon associates and their datafeed-generated "shops".
I think a fee for datafeeds would help a little. But the only solution would be to limit the number of subscribers. For example the following solution would reduce the spam a lot:
after 3 month the 3 worst-performing subscribers will be dropped and replaced by new ones (have to proove their interest and knowledge of implenting the feeds)
Inactive subscribers (less than 3% of total traffic) can be dropped within 2 weeks.
If it's a big product portfolio you can also make different groups and adapt the rotating-subscribers idea per area and also have 10 top subscribers with access to all data.
| 2:51 pm on Oct 3, 2003 (gmt 0)|
Limiting access to a fixed # of affiliates doesn't make any sense. Like I said in my first response, there are plenty of excellent reasons for using a datafeed other than simply dumping every record into its own HTML page.
But if the merchant is concerned about this, it would be sensible to vet sites individually before offering them access to the datafeed.
| 3:21 pm on Oct 3, 2003 (gmt 0)|
There's one merchant I work with who refused to do datafeeds but has an easily crawlable site so I wrote a program to scrape all of their data into a table. I told them I was going to do it first and they didn't object.
The month before I made my own make-shift datafeed, I did $250 in sales for them. The two months after that, I did in excess of $8,500 in sales. They're STILL not interested in sending me a datafeed of their products.
| 3:52 pm on Oct 3, 2003 (gmt 0)|
I don't think limiting the databases is the right answer in most cases. If affiliates are simply creating hundreds of thousands of pages, many times education is the best bet. Affilaites who use these databases to complement reviews, create niche sites or even set out the data in a way the merchant has not for some reason are my main reason for promoting databases for affiliates. An affiliate with a god grasp on marketing can often take our database and make a better product for their users than exists anywhere else- including the merchant site.
| 5:03 pm on Oct 3, 2003 (gmt 0)|
My feeling is that for the average mid tier merchant datafeeds are typically not a big problem. The technical skill required to build a good datafeed site, or the fear of trying something new, somewhat screens out a larger percentage of affiliates. One program I manage has over 3000 affiliates, the datafeed is free for all, yet I only really have about 5-8 partners that have generated good, high ranking sites. Many of the datafeed affiliates I know DO build a better site than some of my merchants!
I think with very large programs (like Tiger) that have thousands and thousands of active affiliates datafeeds could get out of hand and evidently did for them.
| 5:52 pm on Oct 3, 2003 (gmt 0)|
Yeah, I'm afraid that the Tiger mess may be the proverbial handwriting on the wall re the future of datafeeds. Regardless of the merchant's intentions, the end result will be the flooding of the serps with repetitious catalogs and directories generated from the same source.
| 8:12 pm on Oct 3, 2003 (gmt 0)|
> They're STILL not interested in sending me a datafeed of their products.
I'd stop asking and encourage them not to persue it
| 4:01 am on Oct 9, 2003 (gmt 0)|
I just wanted to say "thanks" for this discussion. We've just recently started offering datafeeds to our affiliates (we have been offering them to our drop ship customers previously), and the points mentioned here have certainly given me food for thought.
| 6:17 am on Oct 9, 2003 (gmt 0)|
>> flooding of the serps with repetitious catalogs and directories generated from the same source.
Good point, and it's already happening for some keywords/phrases.
| 6:35 am on Oct 9, 2003 (gmt 0)|
>> "Vomiting into Google"<<
Is it time now for SE's and consumers to seriously reconsider whether database size (no. of pages indexed) is important to a good search experience?
It may well be that just (intelligently) indexing an upper limit for any one domain may not miss much unique or substantive but certainly miss a lot of duplicate or template driven pages? Why does a SE need to deliver someone direct to the actual page? Is it sometimes good to deliver them to the home or category page where they can see a link to what they want in the context of the site? Not stirring anything up, just questioning whether SE's are just getting TOO big for their own good!