Site steadily losing traffic, not because of panda

[beerSEO]

Hey all,

I have an eCommerce site that has been successful for years. There's no real text content except for completely unique blog posts, having to do with current events happening at the wholesaler (you will only find this content on our site, no where else).

No blackhat tactics and all organic linkbuilding.

As of March 24 I have been seeing a dramatic decrease in traffic and revenue. At first I thought it was seasonality, but that can't be as the product we sell is used during the spring and summer primarily. A check with google insights confirms this, a random list of keywords checked showed all search trends were up.

In our robots.txt we have:
Disallow: /*?
Allow: /*?p=

This is to allow page query strings but block all others. Can this ("/*?") somehow be interpreted as telling the crawlers to not index anything?

Any ideas?

Thanks!

[caribguy]

Quick checks:

- can your /?p=1234 pages still be found in the index?
- do you use canonical urls to avoid serving duplicate pages
(e.g. /?p=1234&q=aaa != /?p=1234&q=zzz )

Consider rewriting the query string urls to something more intuitive, from /?p=1234&q=aaa to /category/topic

Once you have eliminated these types of technical problems, you'll be able to do a better analysis of your traffic and visitor behavior.

[beerSEO]

We do have canoncial tags to take care of query strings.

The URL structure doesn't have any query strings anyways except for the ?p=, which are just the next product pages (which are allowed by robots).

Is there anything else I should be looking at?

I was looking at top competitors and none seem to be actively doing [shady] SEO... so I'm at a loss here.

[TheMadScientist]

I'd recommend visiting the Apache Server [webmaterworld.com] Forum and finding out how to strip/redirect all query_strings except the ones you need ... IOW: I wouldn't try to use robots.txt to try and solve the issue, I'd go with a 'less open to bot translation' answer.