SEO for Firefox Updated to Include Compete.com Data

Hi Aaron,

do you have an idea how many visitors a site should roughly receive so that the data from compete on unique visitors is sort of accurate (in the right ballpark)?

For German sites they often say not enough data available (makes sense, I guess), but for other (relatively small sites) they do give data that sometimes leaves me wondering "Is this data of any use or should I forget about it?".

Does it make sense to look at the traffic volumes of sites with say only a few thousand visitors/month or will the data usually be highly inaccurate for such small sites?

thanks!

For smaller US targeted sites (say 2,000 to 5,000 uniques a month) I sometimes see unique visitor volumes that might be off by a factor of 2...showing half or double the expected traffic.

With such sites they are still good at giving you their most important keywords and showing general traffic trends.

Compete estimates my traffic within 5% every month. Pretty impressive.

They also removed the supplemental results. It got to where every page I had on one site (22,100) was showing to be in the sup index.

That is why the supplemental results part was dropped Mack. Google killed the hack I was using to find supplementals and made those using it get bad data. I could have tried switching to another hack, but Google would likely just disable that too, and I would have been passing Google's intent to mislead webmasters on to the general SEO community. I didn't want to do that.

thx for the reply, Aaron. I didn't think they were nearly as accurate for small sites to be honest, but good to hear that.

This is a bit off-topic, but do you (or anyone else) have an idea how accurate the g traffic estimator volumes for the paid listings are? Are they similarly accurate as compete's data or can they be way off for frequently searched terms, too? thanks

Hi Patrick
They can't predict upcoming spikes due to news events, but they should be somewhat decent at predicting the general trends. Their numbers are on a different scale than compete though...because Google is estimating ad clicks, not searches. Thus in that regard the numbers will be skewed low. Plus if your bid prices are low, or you have a low CTR, or the current ads have a low CTR...any of those further skew the data lower.

Also remember that they have broad phrase and exact settings, so if you check for broad match it might show lots of related volume that is just not there for the core term you want to rank for.

Yeah, I have noticed trying different strings doesn't really seem to work well anymore. I had worked out one lengthy combo at one time to find sup pages, but I checked it a few days ago to find out it had been busted.

Guess that is what I get for sharing it. All the same I am glad to see it removed as it was very misleading most times.

Keep in mind that I never intended to mislead people with the supplemental stuff...it was just a datapoint that was made available by Google only to be quickly taken away by Google.

thx again. I understand that google's traffic estimator estimates clicks on paid ads (and that I have to use diff. match types), so I always sort of multipled those numbers by a factor of 4 or 5 (expecting the organic listings to get clicked about that much more). I know that it's far from an accurate prediction obviously (as kw tools in general), but I only found out about google's traffic estimator tool a while ago (until then I had only used the adwords suggestion tool and looked at the colored bars), thats the main reason I was wondering if those numbers really meant anything or if I should better ignore them.

Another reason was this example that more than surprised me:

for "speak french" (phrase match) the estimated clicks are around 30 in the traffic estimator tool. Thus I was thinking around 100-150 (or so) should go to the no. 1 organic listings, or 3000-4500 (or so). However the site www speakfrench co uk shows very little traffic on compete.com though it ranks at the top for a variety of these keywords (#1 for the most generic "speak french"). The same thing for other sites that have a page rank for "speak french" (and permutations) like speakfrenchsite com or learnhowtospeakfrench

Now, I assume the reason for this is that these sites are searched for by people who know English all over the world (not focussed on the US) and consist of only 1-3 pages that don't do anything to attract related keywords or long-tail phrases...?

Hi Patrick
You can always buy the keyword on Google AdWords and test it.

yes youre right, I'll do that. thx for your answers as always!

Hi Aaron,

The delicious count seems to be off. I tried with and without the www, and I always get the without the www result, I only have 2 on the without and like 180 with.

I'm on Mac OSx if that makes a difference, though maybe it's just a delicious count problem. I was getting the correct count before the update.

You really did a nice job with the update!

The last SEo for Firefox was starting to kick out errors, but I have not seen 1 since updating!

Thanks again and great tool! Maybe my all time favorite!