Alexa is becoming a joke lately. Some people claim that over the previous two years it was already losing reliability, but lately it went completely nuts. Whatever they did on the last algorithm update, it messed up their rankings badly.
I know that many people already ignore Alexa, but most advertisers still consider it when evaluate ad buys on websites, so I think we need to raise the awareness about it, and encourage them to use other available tools to gauge traffic.
The problem emerges when you make comparisons of different websites, and when you bring into the picture other traffic tracking services like Compete.com or Google Trends. I have 5 interesting cases to illustrate this.
Case 1: Daily Blog Tips vs. Daily Writing Tips
This was a natural example since I own both sites and have access to the real traffic figures. Daily Writing Tips (as of today) has an Alexa rank of 91059. Daily Blog Tips has a rank of 112363. The Alexa graph of the reach of both sites tell a similar story:
So if one was to trust Alexa, he would conclude that Daily Writing Tips is getting a lot more traffic than Daily Blog tips right? Wrong!
In fact it is exactly the opposite, Daily Blog Tips generates around 15,000 daily page views, while Daily Writing Tips only 8,000. The proportion of unique visitors is similar.
Now the interesting thing is that both Compete.com and Google Trends show a very similar traffic trend for those two sites, and one that is much closer to the real numbers.
The time frame of the graphs is different, but it should not matter. The Alexa graph is displaying the trend for the past 6 months, while Compete and Google are displaying it for the past 1 year.
Case 2: PhotoshopStar vs. VentureBeat
PhotoshopStar is a photoshop blog created by a friend of mine. It has around 2,500 RSS subscribers and gets some decent traffic. As of today it has an Alexa rank of 30,422. VentureBeat, on the other hand, is one of the most popular technology and web 2.0 blogs on the web, with tens of thousands of RSS subscribers. It is Alexa rank is 31,397. Hmmm. This would mean that PhotoshopStar gets more traffic than VentureBear? Check out the Alexa graph:
Now what would you think if I tell you that PhotoshopStar gets around 300,000 monthly page views (source is my friend), while VentureBeat gets over 1,000,000 (source is Federated Media)? Alexa must be kidding right? Not only that, but again both Compete.com…
and Google Trends managed to display a more realistic situation.
Case 3: JohnChow vs. Techmeme
JohnChow.com is a popular “make money online” blog. According to John himself his blog generates around 300,000 monthly page views, and its current Alexa rank is 58,000. You probably also know Techmeme, a tech news aggregator. Its Alexa rank is 73,938. Again, one would expect that Johnchow.com is receiving a lot more traffic than Techmeme if he was to follow what Alexa says right?
Well this is probably far from the truth. Techmeme must be getting at least twice as much traffic as JohnChow.com. Both compete…
and Google Trends in fact are showing a very similar pattern, where Techmeme has by far more traffic than JohnChow.com.
Case 4: BloggingTips vs. Copyblogger
BloggingTips is a meta blogging site, with around 2,500 RSS subscribers, and its current Alexa rank is 103,872. Copyblogger, on the other hand, is the authority copywriting blog on the web, and counts more than 40,000 RSS subscribers. Oddly enough, Alexa ranks Copyblogger right aside with BloggingTips, with a rank of 102,259.
Over the last month Copyblogger shows a spike on Alexa, but still you can see that was going hand in hand with BloggingTips for a while. Do they really receive a similar traffic? I highly doubt so, and the Compete.com traffic report confirms that.
Google Trends shows a trend to both websites that is very similar to Compete.com (that is, the BloggingTips traffic is so small in comparison to Copyblogger that it is not showing in the graph).
Case 5: PSDTuts vs. DZone
Someone could think that the crazy Alexa rankings covered above were due to websites coming from different niches. Well let’s take two on the same niche then: web design.
PSDTuts is one of the most popular blogs with photoshop tutorials. It is current Alexa rank is an impressive 11,957. DZone.com, on the other hand, is a popular social bookmarking site for web designers and developers. One would expect that the latter would have a higher traffic right? Not for Alexa… which is giving a rank of 23,233 to DZone.
Compete.com is telling a whole different story though…
And so is Google Trends. Yet another time both of those traffic tracking services are agreeing with each other, and completely disagreeing with Alexa.
Conclusion
As you can see, on all those 5 cases Alexa was reporting very strange traffic trends. Sometimes it reported the opposite of the real situation. I am pretty sure it would be easy to find 10 or 100 more similar cases.
Compete.com and Google Trends, on the other hand, are behaving very similarly, and judging by the numbers that I know, they appear to be far closer to the real situation.
In my opinion the only factor that makes people still consider Alexa is its ranking system. People like rankings. They like to compare A with B and C. Compete.com and Google Trends still need to develop such a continual ranking system (right now you can only do on-spot comparisons).
Should they come up with such a ranking system, though, I am not sure how many people would still look at the Alexa numbers.
Alexa is by far the worst traffic website out there..
One should not worry more about alexa ranking. It calculates trafic from the pc’s having alexa tool bar. So it don’t give correct traffic information.
I forgot to mention that the entire time, my Quantcast rank has been relatively stable.
Alexa has just gone berzerk on me the past few weeks. My most popular website (Info-Mac) dropped from in the 400,000s to the 10 millions, then went back into the 600,000s within a matter of two weeks.
My other, much smaller website with way less traffic (System 7 Today) is somehow in the 200,000s, is supposedly getting 99.9% of traffic from South Korea, and is the 3,399th most popular website in that country.
I’m now completely done with Alexa.
I dont know how exactly alexa work,.! but my site even reached 150.000, know back to 163000. I almost update with new post everyday!
I have no idea lol…
I thinks so, cause get alexa rank high is not so hard any more.Many people cheat about it.
Dunno why, my alexa rank make me little confuse, coz they rate unpredictable – i mean, to day my rank at 4.977.425 and tomorrow the point at 12.545.235 but 3 days later the rank at 4.977.425.
Thanks for your informations
Alexa results are not reliable and bias,it is very easy to manipulate the system,example if the website modarators download the alexa toolbar on their pc and used every time on their website it will effect the results regardless even they got only few visitor a day
Well researched post Daniel, I think this will give for thought for the Alexa people
aha finally i’ve got answer while i have stay in here.thank you sir for the explain.
I read somwhere that social site traffic from digg and even stumbleupon is now being discounted by alexa. I am guessing you get a lot of social traffic here?
Well, I don’t disagree that viewing Alexa stats can be MISLEADING, but I’m not sure I agree that they are not ACCURATE.
First of all, I believe you are completing misinterpreting the Alexa REACH stats. Their REACH stat is a percent of all Internet users that the site is being viewed by. So theoretically, if Alexa was a truly accurate and unbiased, representative random sample of all Internet users worldwide, then if you take the percent that they say your site ranks and multiply times the total number of Internet users worldwide you would get a number which equals your site traffic.
So there are what, 190 million Internet users? 300 million? Who knows? It’s a moving target and no one knows the exact answer. The point is, whatever the number is, it’s huge and your monthly uniques are going to represent a very teeny tiny percentage of total users. So, allowing for some error, it is entirely possible that all of the comparisons you disagree with on Alexa are in fact accurate. Remember you are looking at a comparison of relative PERCENTS ot total Internet users, not the actual user counts that Compete shows. It is very difficult to move the needle when you are measuring a what your traffic represents as a percent of all users. All but the top 100 sites will look very small on this measure.
And there is no possible way you can compare a percentage ranking from Alexa with estimated visitor counts from Compete without doing some further math like I have described.
The second relevant point I would like to make here is no two web analytics reporting systems EVER agree. None. Nada. Zip. Zilch. So forget about getting agreement. You never will. The reason is because you do not know all the relevant details about which traffic was counted and which wasn’t. Particularly session length is the one variable which can totally throw off comparisons for unique visitors. I say 30 minutes, you say 20 minutes, somebody else says 15 minutes, another 60 minutes, and 90% of analytics package users leave the setting at whatever the default setting is for their package and the setting varies from analytics program to analytics program.
Another obvious but not so obvious factor is whether the stats throw out visits from robots or include them. And the list goes on and on.
There will never be a defacto standard that ever totally agrees with your web logs. We might get some agreement strictly for convenience sake that this analytics or that analytics is the current benchmark for comparing sites, but there will never be an unambiguous absolute truth here. You will always be making decisions under conditions of uncertainty. It’s just an estimate.
Today you are railing against Alexa. In the future, you will very likely not be happy with Compete, or Google, or Hitbox or you-name-it.
I’ve noticed almost every site shows less traffic at Alexa and I guess the only reason is that there is not so many user with Alexa toolbar.
In fact, the toolbar did not exit for Firefox and we know that this browser has grown. And about the future, will Alexa provide a tool bar for Google Chrome? No… that’s not the question. Are the users willing to install an Alexa toolbar? I don’t think so. The trends show less Alexa toolbar users.
It is not a matter of whether Alexa, Complete, or whatever page ranking site tells you. From what I see, they are still working on that and there is no surefire way that they are accurate. Because everyone have different observation and page ranking architecture.
Their methodologies are also different. Not every pageranking site is the same. Everyone have different experience they have been through and learn the years of experience. Same theory like page rank site. They are still in the works and they aim for improvement not perfection. As the saying goes “No one is perfect” same thing, same theory applies to the mathematical problem solving that involves page rank. That is why their methodologies is not an easy task to develop.
Only the owner of the website who works on the ranking architectures, methodologies, optimization and allocation knows how it works but it is not yet perfected in anyway. It might take years for it to come into fruition. So bear with them.
Since there is no scientific track record to prove that claim up so you can’t be so quick to judge them. It is a learning process so we also continue to evolve. Same goes with the website. Till then take care.
Wow,
You have made your point so clear that a blind can see. But as you rightly said, advertisers are still taking it into consideration, so, we can’t forget that name ALEXA…
The figures are pretty poor estimates of actual popularity and traffic rank. Alexa numbers should not be taken as a way of judging the true popularity of a site.
I honestly dont know what to think about this alexa. I’ve been pulling my hair out trying everything I know to get my websites rank up. I look at some comp. site and they are ranked the top 5 in google and under 100,000 here! The market I’m in isn’t even very big.
if theres anyone who could help me see something i’m missing please write!
our site is
http://www.ipodrefresh.com
Interesting things you’ve got here… I just found another same thing on
Just spoke to an Alexa rep at Ad:Tech in Chicago last week. He acknowledged some of the slow downs in their data reporting.
More interestingly, he hinted at Alexa evolving to become a different kind of company in the coming weeks. He asked a lot of questions about analytics and reporting needs.
Just thought I would share!
Something nobody appears to have mentioned is the fact that the Alexa toolbar is still not available on Vista.
Here in the UK vista is becoming more and more popular which means sites that attract UK visitors will drop in the Alexa rankings.
Other countries, where vista isn’t as popular may still have the toolbar installed which will mean their rankings are not effected.
A perfect example of this is a popular UK Business Forum. it’s Alexa rankings are in a steady decline, and yet everything else says it is as busy as ever, another thing Alexa highlights is the growing percentage in overseas visitors.
This could be caused by 2 things;
1. It is getting more overseas visitors
2. UK visitors have Vista, oversees visitors don’t
I’m more inclined to think 2 is more accurate.
I started watching this trend some time ago. My blog, which is in a similar niche as DBT has a significantly better Alexa rank sitting at 66,417. Sure, I get good traffic but I’m not in the same league as DBT.
Go figure. Good post Daniel.
Well researched post Daniel, I think this will give for thought for the Alexa people
The only time I remember Alexa being useful was when some of my friends did PayPerPost sponsored blogging. PayPerPost use Alexa as a measure and you are assigned sponsored posts to write based on your Alexa ranking. Apart from that, I can’t remember Alexa ever being useful for anything else. Its funny something as a big as Alexa is so irrelevant to web users.
While I agree that Alexa’s rankings are widely inaccurate, Alexa still produces rankings that many advertising systems pay attention to. As a result, we still need to play the game!
The Alexa numbers are completely useless for getting solid estimates and there is no justification for using Alexa data to evaluate websites.
Compete shows more realistic data than Alexa, but is also useless if you are interested in non-English or maybe even non US based sites.
I guess Google trends for websites will be the authority in the near future when it comes to traffic evaluations. With Google Analytics, Google Toolbar, Google Search and other services they have so many data sources that they should be able to draw a realistic picture.
Thanks for finally confirming what all of us felt was going on. I track a few sites and compare them to a range of competition and have been definitely noticing similar trends.
hey daniel it was the same old story for alexa…
Hi Dane.
Same shit is on my blog, my friend blog gets less then half then traffic of mine but he is like in 40k in alexa and me still crapping in the 80k. I think the new restructuring of the ranking algorithm is totally worthless.
The old one used to give better results.
I think that there is a downward spiral on most sites according to Alexa is because less people are using the Alexa toolbar.
Daniel your just now noticing this, lol. Alexa has been garbage to me for years.
If I’m not mistaken Alexa ranks are based on its tool-bar. The only traffic counted by Alexa is the traffic coming from individuals with their tool-bar.
So a simple explanation for the lack of rank compared to other sites would be the other sites get more traffic from individuals that have Alexa’s tool-bar installed.
It is a terrible ranking system, but I believe that’s the explanation for it. It definitely needs to improve.
Alexa was never accurate. Don’t count on it. You can fake Alexa ranks easily by installing their toolbar and going to your site a few times a day.
I think its about time advertisers smarten up and stop checking the Alexa rank. Or maybe its just good for people with low traffic. 😉
Even with proof like this, advertisers still tend to consider the Alexa ranking before purchasing ads on a website
We’ve all known Alexa to be worthless for some time now but evidence like this really puts a nail in the coffin.
I get the same kind of results. I think their problem will continue to get worse unless they think of a really great reason for people to download their toolbar. They probably have more people uninstalling it than installing it these days.
Alexa is definitely becoming more and more worthless, especially since they last changed the algorithm many people i know just do not take it in to consideration any more.
Now my rankings have improved since the change, however i know for sure that i do not get anywhere near as much traffic as this site does so why should my site have just over double the rank of yours.
It seems very silly and inaccurate, i am not sure how they even measure or calculate the results anymore.
To be honest they would probably get more accuracy by picking sites from a hat.
Good post. I agree that something is definitely off.
FYI, you’ve got a typo in your title—it’s “becoming,” not “becomeing.”
I think this is a temporary problem. They are obviously changing the way they are monitoring sites.
The thing is, scores are based on 3 month averages.
Whilst Bloggingtips has a score of 103 thousand just now, it’s 1 week average is about 600,000. Compare this with CopyBlogger’s 1 week average of 137,183.
Likewise, DailyBlogTips has a score of 112 thousand just now however your 1 week average is 265 thousand.
All these scores should head towards the 1 week average in time, Alexa is just slow in changing their main score because thats based on a 3 month score.
I don’t think much of Alexas ranking system but I do think it’s getting a little better.
The above doesn’t explain why bloggingtips is currently higher than dailyblogtips as I don’t believe that traffic has been higher at BT for any length of time (perhaps maybe a day at most). This suggests that previously scores were wrong but perhaps the 1 week scores we are seeing now is more accurate. Maybe they were messing with the algorythm. All of these are guesses but since Alexa hasn’t released any news releases since their announcement in April that’s all we have to go by.
Daniel, you know what! My blog has got traffic in 100s which is on an average 1000 uniques per day but my Alexa is better than yours – 63000.
This is simply messed up Alexa algorithm I would say!!
Thanks for not being the only one looking at this, my traffic and subscribers have tripled in the past 4 months and yet Alexa is showing a constant downward spiral. The funny thing is look at sites like paypal (which are owned by amazon.. aka alexa) and they are on a sharp turn up.
What is frustrating about this is so many people have put merit into this system and it is still looked at by advertisers. I wish they would either fix it or terminate it.
From the beginning, I thought there was something wrong with Alexa, ans as you so throughly showed in your post that there is something to be concerned about. Great Post, thanks.
This is very interesting Daniel. Just looking at my Alexa rank of 119,454 compared to dailyblogtips is 112,365 – doesn’t make sense since my traffic is very low in comparison to your site.
For a while we also heard that placing Alexa’s widget on your site helped your rank. It might sound funny that when I had the widget, my alexa rank was lower and when I removed it with the new blog design, the rank is higher! So Alexa is a real joke..
Wow. That’s way too erratic to be worth much.
Wouldn’t that showup as huge boost for every1?
But anyway, thats great. (if it works)
@SJL, they launched “Google Trends for Websites” a while ago, which is supposed to track website traffic also.
Alexa also gets traffic data from AdBrite now, so if you are an AdBrite publisher, you get a huge alexa boost.
Just s quick follow up question…
Doesn’t Google Trends only track Google SE traffic?
Interesting. Thanks for bringing awareness to this. I haven’t ever looked at compete.com before.
Mine as well. Alexa ranking strangely drops but i get way more traffic.
And if someone had some doubts, i guess they wont anymore after this post.
I agree. I don’t pay any attention to Alexa. My traffic doubled and the Alexa rank went down and the graph showed a drop in traffic.