How would you test this?

[kamakiri]

I finally fixed a problem with my GameBRAIN web site. (http://www.gamebrain.us/) It has sat collecting dust for months while the market grew. The main competitors (here (http://www.fpsbrain.com/?switchLanguage=english) and here (http://www.manapotions.com/)) have expanded their product line, proving that there is a market there. Both of them added an energy drink to their product line.

I just got that free $100 coupon that is rolling around, and I have been thinking of testing my site again. Anyone have any advice there?

What tools can I use to see what search terms are leading people to the competition?

How would you guys go about testing this muse?

[MattHavok]

have product photos on your site. give the eyes something to look at. gamers like visual.

[AntonTheKhan]

I finally fixed a problem with my GameBRAIN web site. (http://www.gamebrain.us/) It has sat collecting dust for months while the market grew. The main competitors (here (http://www.fpsbrain.com/?switchLanguage=english) and here (http://www.manapotions.com/)) have expanded their product line, proving that there is a market there. Both of them added an energy drink to their product line.

I just got that free $100 coupon that is rolling around, and I have been thinking of testing my site again. Anyone have any advice there?

What tools can I use to see what search terms are leading people to the competition?

How would you guys go about testing this muse?

Go to compete.com and quantcast.com. I think spyfu.com and there are several others that can actually see what keywords are driving traffic to websites. Manapotions seems to be getting steady traffic, whicle fpsbrain is only a few hundred visits a month, so nothing special.

[jakediddy]

No offense, but how does one website know the traffic of another site? Is there some formula they apply to the results of google ads after a batch search?

[kamakiri]

I have gone over the basics. Not too sure about SEO or how to make mine better though. Any specifics there?

@jake - You serious there? I thought you knew better.

@Anton - FPSBrain has 3 sites actually, and that English one is probably the least busy. The German and Dutch ones are busier. Many times they don't even update that English one. The company is located in the Netherlands, and I would assume only threw up an English site to gain PR.

@Matt - No.

[jakediddy]

@jake - You serious there? I thought you knew better.


I'm totally serious. I'm pretty confident there's no way for a website like Compete (or any other traffic estimating website of an external nature) to accurately collect traffic information without actually being involved in every request to the website-in-question, either Google Analytics style with a script or image, or installing a BHO/plugin inside unsuspecting internet surfers' browsers.

Turns out my suspicions are valid: http://venturebeat.com/2007/02/22/traffic-measuring-continued-why-compete-doesnt-work-and-why-quantcast-does/

These websites are not accurate at all unless you volunteer to put a tracking pixel (or JavaScript or whatever you want) on your pages. Without that, they need to fall back on formulas to estimate traffic.

Perhaps websites such as JCPenny.com or youtube.com would be accurate, since many different "classes" of computer users visit it; but websites pertaining to IT or software development (maybe even gaming) wouldn't, as visitors to those websites are generally more computer-savvy and don't install browser plugins like that very often.

Perhaps I phrased my initial post wrong, but at least you motivated me to actually look it up.

[kamakiri]

Jake - That post is from 2007. For all intents and purposes it could have just as well been written on papyrus. Saying that those services are somehow incorrect would be like saying that the Nielson Ratings are incorrect because they don't have a monitoring device in every household on every television set.

Tracking pixels? Do you really think google, you tube, or MSN use them, and if they do let anyone else see the data? Especially when given the alternative of just throwing a line of code in there to do just that?

The whole topic of that article has nothing to do with what I am talking about any way. If you need a third party tracking pixel to know your web hits when any free web counter can do it for you, you are in trouble to begin with.

The size of the market is proven with statistically viable data. Embracing tech or sticking your head in the sand is your choice. Dig a bit deeper and come back again.

[networkmemetics]

There is no way 100% accurate way to determine traffic to a competitors website.

Not a lot has changed since 2007 Kamakiri. This is what JakeDiddy was saying. He was trying to help you - I almost didn't want to help you after your "stick you head in the sand comment".

In order for a 3rd party (quantcast etc) to collect data, each individual site owner must place a meta tag / tracking pixel / javascript code on the site. You can track your own site easily. You can get an idea of the market using the sites mentioned, but it is not an accurate account of traffic.

Googles Keyword Tool, Wordtracker etc, are not accurate either...they are estimates.

You can analyze a competitors pages for keywords using Googles keyword tool. You can use SpyFu to peak into their PPC campaigns.

You can analyze the anchor text of their backlinks to determine what keywords they are targetting for their organic seo campaign.

[officer_dibble]

I'd agree with all of you on this.

It won't be 100% accurate - but the trend over time/comparisons between sites are still very useful.

For one of my own sites - Google analytics under reports by at least 18% for transactions.

[AntonTheKhan]

Jake - That post is from 2007. For all intents and purposes it could have just as well been written on papyrus. Saying that those services are somehow incorrect would be like saying that the Nielson Ratings are incorrect because they don't have a monitoring device in every household on every television set.

Tracking pixels? Do you really think google, you tube, or MSN use them, and if they do let anyone else see the data? Especially when given the alternative of just throwing a line of code in there to do just that?

The whole topic of that article has nothing to do with what I am talking about any way. If you need a third party tracking pixel to know your web hits when any free web counter can do it for you, you are in trouble to begin with.

The size of the market is proven with statistically viable data. Embracing tech or sticking your head in the sand is your choice. Dig a bit deeper and come back again.

Kamakiri,
Google is not the only advertising avenue there is. You can try with Facebook and myspace, or the smaller Yahoo and MSN. Facebook seems like the best choice. Thing is you never know what search engine or avenue will bring you in the highest conversions. With AdWords, you can also advertise on specific sites (I suppose you want to target gaming sites and such). This might be more viable than advertising in search. You have to test. Good luck.

[jakediddy]

Kam - I understand the point you are making with regards to Nielson ratings. What they are doing requires them to call a decent number of viewers and ask them various questions. They can then assume that these results more or less represent the population or certain demographic as a whole. Simple statistics.

The only problem is that when you apply this to computers, you use toolbars (such as Yahoo! or Google's toolbars, I'm sure they sell their statistics) instead of phone calls. Many tech-savvy users do not install 3rd party toolbars that come "packaged" with software such as Google Earth or Java runtime, so you miss out on data from virtually the entire tech-savyy demographic (or at least software developers who hate unnecessary software being installed on their machines).

Regardless, I had no intention of turning this thread into a discussion on Alexa and Compete, and I apologize. Just watch out for my sand hole. :cool:

[kamakiri]

I agree that this thread has gone well down hill. I guess my point is that all of the tracking tools we have err on the conservative side.

If the tech savvy users aren't calculated in, then they can just be grouped into the 80% of people I ignore, as there probably is no good way to determine their web habits to begin with (correct?).

I have only started looking at the sites mentioned here, but I am familiar with Alexa. They use tool bars and have a huge number of installations. More than enough to be statistically valid. That they don't count everyone and every group really doesn't matter much. I am not looking for the 100% mark.

@Anton - We all know google isn't the only game in the verse, but if you don't have at least marginal success there, you probably aren't going to anywhere else either. The Japanese have a saying that if you chase 2 rabbits you catch neither. I haven't ruled out those other routes, but I need to have a few success with google ad words (ie burn through $100), and see what works, needs improvement, fails...

@Dibble - Thanks. Under reports by 18% means it is right 81% of the time, which is more than what I am looking for and need. If I can target even 50% of my target market that is what I am looking for. The low hanging fruit. That last few on any media (google, FB, yahoo) will not be my target for a long time. I need to exhaust my other easier opportunities before turning back and starting from the beginning to target the more marginal prospects.

@Network - I disagree that they are inaccurate. You can't really make a broad statement like that and not cite it. Under-reporting? Probably. Inaccurate? Probably not. They have some serious sample populations that they base their data on. Would you agree that even if they were inaccurate, that margin of error doesn't matter for my testing purposes?

That is kind of my whole point here. If under reporting is the least of my problems I could have worse problems.

[kamakiri]

I wanted to drag this post up because of some awesome resources that I read about on Seth's blog:

Discover the truth about a site's online traffic (http://sethgodin.typepad.com/seths_blog/2009/12/discoverthetruth.html)



Pretty cool stuff. I am going to spend the evening digging into it.

[Monkiii]

I know compete.com is waaaaaaaaaaaay off for my highest traffic site.

[Recoveringfool]

Dated article, but interesting real world experience is cited.

http://venturebeat.com/2007/02/22/traffic-measuring-continued-why-compete-doesnt-work-and-why-quantcast-does/#more-3481