Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
The Internet Businesses

Hits or Misses: Who is Your Website's Audience? 146

Posted by Hemos from the interesting-informationo dept.
securitas writes "The Christian Science Monitor's Gregory M. Lamb wrote a story interesting to anyone who runs a website: How do you accurately and reliably measure the audience for your website? From the article: 'Most websites have no idea how many people view their content. This inherent fuzziness is causing problems for commercial websites, especially online publications desperate to make money from Internet advertising... How can you charge for ads when it's nearly impossible to tell advertisers how many people will see them?' The article discusses the flaws and problems with Nielsen/NetRatings and comScore Media Metrix - they grossly undersample workplace users - and the rise in the number of sites requiring user registration."
This discussion has been archived. No new comments can be posted.

Hits or Misses: Who is Your Website's Audience? More

Hits or Misses: Who is Your Website's Audience?

Comments Filter:

  • As Ty Webb would say... (Score:5, Funny)

    by Sqwubbsy ( 723014 ) on Monday June 21, 2004 @08:49AM (#9483654) Homepage Journal
    By height.

  • Not too hard (Score:1, Interesting)

    by Anonymous Coward
    1. Cookies
    2. IP addresses
    3. Mandatory registration

    • Re:Not too hard (Score:5, Insightful)

      by onion2k ( 203094 ) on Monday June 21, 2004 @09:01AM (#9483801) Homepage
      Mandatory reg. puts people off using the site in the first place (Why register if you can see the content.. If you can't see the content who knows if its worth registering for?).

      IP addresses is half the problem (everyone behind one company firewall looks like 1 user).

      Cookies are ok so long as your users are ok with you "tracking their browsing habits".

      Its a tricky puzzle...
      • Cookies are ok so long as your users are ok with you "tracking their browsing habits".

        Or unless you're dealing with the goverment who won't even let you set a "session" cookie. (No matter how many times you try to explain that it goes away when they close the browser.)
    • 1. I don't like being tracked, and trust AdAware and misc tools to rid me of those evil-doers. Plus I clean out my Cookies and temp files occasionally.

      2. Dynamic IP's, Proxies, and Gateways.

      3. Not that again; Everyone hates signing up somewhere before getting in, and being dissapointed at the content. Repeat this over and over. You just walk away when you have to register after a while, cause you simply don't want all your info on webpages you've abandomed. (or don't trust your information, you know those
    • It was my task to develop log anaylsis software to count visitors for large web sites. I was not only surprised to find out how inaccurate the art was, but also had difficulty in convincing other web-experienced collegues on how impossible it was. "All the other web analysis programs display number of visitors" they said. Well they all make guesses is the answer. The current best practice is to count unique login names, but most sites don't use authenicated logins and even then you can have many hotmail

  • use cookies? (Score:4, Informative)

    by Anonymous Coward on Monday June 21, 2004 @08:53AM (#9483699)
    I always just set a cookie with a tracking ID, and then use that to keep track of the anon user. counting the number of tracking cookies given out each day, and the time they were used for seems to work sufficiently for me... or is there some problem with that I don't know about?

    • Re:use cookies? (Score:4, Informative)

      by PhxBlue ( 562201 ) on Monday June 21, 2004 @08:58AM (#9483765) Homepage Journal

      Only when you consider browsers that let you reject cookies, such as Firefox. But then, that's more the web developer's problem than mine, since I'd just as soon remain anonymous.

      • Re:use cookies? (Score:2, Informative)

        by Anonymous Coward
        Don't forget, a lot of people who use spyware tools like adaware and spybot delete tracking cookies on a regular basis, further skewing results.

    • Re:use cookies? (Score:5, Informative)

      by Stephen ( 20676 ) on Monday June 21, 2004 @09:13AM (#9483919) Homepage
      Read the article. They are complaining that one user may read the content from work and from home, and so count as two users. One might also point out that sometimes two people may use the same computer, and only count as one person.

      • Re:use cookies? (Score:1, Informative)

        by Anonymous Coward
        I'm glad I didn't waste my time reading such a stupid artical then.

        That's like saying a billboard company wants to know the exact number of people that look at their billboard while driving by. Some things just can't be done, what a waste of time.

      • Oh come on! (Score:5, Insightful)

        by peterdaly ( 123554 ) <petedaly@ix[ ]tcom.com ['.ne' in gap]> on Monday June 21, 2004 @09:33AM (#9484126)
        Read the article. They are complaining that one user may read the content from work and from home, and so count as two users. One might also point out that sometimes two people may use the same computer, and only count as one person.

        My wife and I both read the same article/section in the newspaper we got yesterday, even though we only got a single paper. (We "logged" 1 impression even though 2 were made.)

        I understand that is the opposite of what you suggest, so...

        Not only that, but we had some sections delivered to us that we (gasp!) threw out without even reading even though we may have been part of the target demographic. (We "logged" 1 impression even though 0 were made.)

        And the web is different how?

        -Pete
    • I always just block your tracking ID cookie and go about my merry way. How do I factor into your statistics?

  • Cookies? (Score:4, Insightful)

    by Mz6 ( 741941 ) * on Monday June 21, 2004 @08:53AM (#9483702) Journal
    Call me oblivious, but wasn't this one of the reasons why cookies were created?

    • Re:Cookies? (Score:5, Insightful)

      by howlatthemoon ( 718490 ) on Monday June 21, 2004 @08:56AM (#9483749)
      What cookies don't tell you is who the person is, are visitors in the target demographic, are you missing an audience, etc. Of course, that said, I don't want to give that information out to most advertisers.

      • Re:Cookies? (Score:2, Insightful)

        by grahamm ( 8844 )
        Why should a web site have "target demographics" anyway? In considering advertising, should the advertising not be made appropriate for the people who do visit rather than the text be made appropriate for the target audience of the advertising?

        • Re:Cookies? (Score:3, Insightful)

          by geoffspear ( 692508 ) *
          That would be nice, but the sad fact is that ad-supported corporate media exists to sell more ads, not to provide worthwhile entertainment or information. For example, do you think TV networks care if their shows have any dramatic or artistic merit? Maybe they once did, but the fact that well-written scripted shows are being replaced so heavily by "reality tv" shows they're pretty much just putting on anything they can to reach coveted demographics.

          On the other hand, if you're running a website for more a

    • Re:Cookies? (Score:4, Informative)

      by RetroGeek ( 206522 ) on Monday June 21, 2004 @08:58AM (#9483775) Homepage
      Call me oblivious, but wasn't this one of the reasons why cookies were created?

      Using the Mozilla cookie control, I regularily go through my cookies. Anything that looks like it is coming from an ad site I delete and block.

      Any site which I do not recognize gets the same treatment.

      I have not had any problems from any site because of this.
      • I've never really understood the paranoia about cookies. It is not very hard to do server-side, cookie-less profiling: they have your IP address and their access log. Any decent data mining solution would give them a site traversal graph for your visits. Plotting multiple site traversal graphs against time would give a pretty good idea whether one or more users were browsing their site from the same IP address. Cross-site profiling is possible by simply sharing access logs.

        I don't particularly like it, but

        • I've never really understood the paranoia about cookies. It is not very hard to do server-side, cookie-less profiling: they have your IP address and their access log. Any decent data mining solution would give them a site traversal graph for your visits. Plotting multiple site traversal graphs against time would give a pretty good idea whether one or more users were browsing their site from the same IP address. Cross-site profiling is possible by simply sharing access logs.

          Which would work how well for

      • Am I the only person here who uses privoxy [privoxy.org] ?
        here is my setup,
        Behind a NAT box, with no ports opened.
        Use firefox as browser and privoxy as ad filtering proxy server. and zone alarm as FW
        I have ad-aware, spybot and grisoft free antivirus, but in last 2 years I haven't had a single trojan/virus/spyware hit me.

        Besides using privoxy will save you the trouble of going through your cookies, as it filters almost all of them.
        Forget pop-up ads, it even filters in-line ads

      • Why? Do you want to see ads that are of little interest to you, or do you want to see ads that are targetted to you?

        Ads are a fact of life, get over it. Besides, targetted ads are actually good enough to click sometimes.


    • Maybe someone knowledgeable can explain what index.dat is and why it is so difficult to delete. Yes, I've Googled for it but most pages seem to be selling software to remove it (who knows what else this software might do?)

  • Isn't it obvious (Score:3, Insightful)

    by DecimalThree ( 524862 ) on Monday June 21, 2004 @08:53AM (#9483709)
    that is why most online advertising consists of fees based on the 'per click' methodology?

    "How can you charge for ads when it's nearly impossible to tell advertisers how many people will see them" --- These people use access logs??
    • These people use access logs??

      Glad someone mentioned that. I check mine quite regularly. I see every click and can track every system that connects to my server. I see referring links (although tabbed browers tend to hide the referrer), IP addresses, what they were going after (or failed to get), and when it all happened. There's just too much good information in an access log to ignore it.

      Personally I'm not making any attempts at revenue, but if I were ever to advertise goods and/or services via someo

    • Re:Isn't it obvious (Score:4, Informative)

      by cshark ( 673578 ) on Monday June 21, 2004 @09:31AM (#9484095)
      No.

      That's why the standard is per impression CPM (cost per thousand). One user even from home could generate hundreds of impressions if the content is interesting enough, and the pages are chocked full of useful ads!

      Per click is another methodology, but until Google came along, it really wasn't the standard on the ad sales end. Still isn't outside of Google and the search engines.

      That said, most web sites do know exactly what demographics are visiting their web sites and when. If it's important enough to buy software to do it, and most do, there are several useful software packages that come to mind. Web Trends is the first one I think of. That program in particular actually catches many of the problems described in the article, and it's not unusual. Many such programs have similar functionality.

      Honestly, it would have been nice to see them do their home work before writing yet another scare piece.

      • That's why the standard is per impression

        Um, sorry, forgive me for point this out, but aren't you talking complete rubbish? Most ad affiliate sites (e.g. commission junction) started moving to per-click YEARS ago, and in fact years ago most started dropping even pay-per-click, the "standard" now is pay-per-SALE. I think you're about four years behind here - this all happened just after the dot-com crash. I swear, reading your post was like a flashback to "slashdot '99".

        • Oh wait, you talking about advertiser rates?

        • You might make your commission on a click basis, but commission junction and most other advertisers everywhere sell ads on a cost per impression basis. The idea is that they get a thousand impressions for every click they pay out. That's called... a business model.

          Read the article, it has nothing to do with affiliate ad sites. It talks about people who area selling ads. Could be anyone, and it's not specific to affiliates.

          Not complete rubbish last I checked. Seriously, look into advertising rates anywhere

  • mmm... (Score:5, Funny)

    by manavendra ( 688020 ) on Monday June 21, 2004 @08:55AM (#9483736) Homepage Journal
    I dont know what the real strategy of most online newspaper websites is, but they seem to follow this pattern:

    1. Make content available online, free of cost
    2. Wait for people to start using and monitor the growth in number of hits
    3. Reduce the website response to a crawl with mind numbing popups, flash ads, quick time ads, and generally anything that would make sure the user "spends" more a few minutes on the homepage
    4. Wait for most users to go away to some other website.
    5. The few braves who remain - force them to register and read all the content, since you want to chart your users by demography.
    6. Finally, now make most of the content premium - based upon the data collected in step 5, however inaccurate it is. Flood the site with more ads, if possible
    7. Moan and bitch that there is no revenue generated.

    8. Repeat cycle

    • Re:mmm... (Score:3, Insightful)

      by Mz6 ( 741941 ) *
      I never really understood the concept behind forcing registration on users. Namely the NYTimes. Granted, I am sure it's for ad revenue, but atleast make it like CNN, MSNBC, etc...

      They know that the majority of these registations are bogus. There are even websites dedicated to fooling it. If the people at NYTimes didn't know any better, there are a TON of Mickey Mouses reading their paper.

    • Reminds me of how to make a small fortune on the Internet. First, start with a large fortune...

  • Unique identifiers (Score:1, Redundant)

    by Scoria ( 264473 )
    From the article: 'Most websites have no idea how many people view their content.

    They don't employ a unique ID stored within a lightly encrypted cookie, then? Of course, those merely provide a statistic related to the amount of individual computers viewing the Web site, not the amount of people. They obviously fail to account for computers with multiple users, such as household machines and public terminals.

  • Editors at Slashdot... (Score:5, Funny)

    by Anonymous Coward on Monday June 21, 2004 @08:58AM (#9483770)
    may think their audience is a bunch of nerds, but in reality its a bunch of suave playboys that get to have sex with many hot women. I suggest they make the appropriate content changes.

    • Re:Editors at Slashdot... (Score:1, Funny)

      by Anonymous Coward
      "if they took all the porn off the internet, there'd only be one site left, and it'd be called 'bringbacktheporn.com'"

  • Uh, No... (Score:5, Insightful)

    by EvilJohn ( 17821 ) on Monday June 21, 2004 @08:59AM (#9483779) Homepage
    This is completely backwards. Infact, it's exactly the opposite. It's quite simple to tell how many people view your webpage, and hell of alot easier (and more accurate) than radio or TV.

    This is the source of the problem with web advertising, your numbers fairly accurate and based on actual events, not some satistically questionable sampling method. There's little room for fudging.

    Demographics on the other hand are a little more complicated. There, you actually have to ask.

    ---

    • It's quite simple to tell how many people view your webpage

      Please enlighten us.

    • Re:Uh, No... (Score:5, Insightful)

      by thogard ( 43403 ) on Monday June 21, 2004 @09:40AM (#9484177) Homepage
      It's quite simple to tell how many people view your webpage, and hell of alot easier (and more accurate) than radio or TV

      No kidding. For the 1st time in the history of advertising since the invention of the shop sign, someone has a direct way to count how many people see the ads and how many of them respond in some positve way. The resutls aren't even close to the typical guesses used in the advertising game to sell ads so they simply say the web stats are wrong and go back to their old ways that say more comercials are good. Too bad the real stats show that consumers are overstaturated and ignore most ads. The problem is that consumers don't buy ads, its the large comapines that buy the ads and they don't know if its going to work or not so the compaines trust the ad providers to provide useful stats and then trust them even it it disagrees with market data. If you think some of the professional software is broken, take a look at real world ads. Some of them run away customers for years. For example Oral-B had anannoying warning sound on an ad for their toothbrushes and I hate it so much I'll never buy one of their products again and that ad ran a decade ago.

      I was in a meeting room with a bunch of ad idiots that had just charged the company I worked for about a million dollars to put the www.$COMPANY_NAME.com on the tail end of some well recieved comercials that were about "building brand". They said it would increase our hits a thousand times. I had the logs and said it had increased the sites hits to about 20 times what my personal site got. The idiot then asked me how much I spent on advertising my site. One of my coworkers made some comment about it being millions less than what they charged and that the web hits had only doubled. The team of idiots then told us we must be getting our numbers from an unrelaiable ad auditing source and couldn't deal with the concept that our numbers were from the apache logs.

      • If all they want is number of hits... (Score:1)

        by Anonymous Coward
        1. Promise to quadruple the traffic to a company's site within 24 hours, in exchange for $$

        2. Post story regarding stupidity of company's advertising model to slashdot, company's server is slashdotted

        3. Profit!!!
      • You know what really makes me wonder? Why do they spend so much energy into justifying idiotic metrics, instead of in actually making people interested in the product?

        E.g., yeah, Fake UI ads generate clicks. But do they actually make people buy your product? I don't think so.

        The one time I was tricked into clicking on one of those, I closed the window so quickly, that I don't even remember what product or company was on that page. I wish I could even say that I hate them and won't buy their products ever

    • Re:Uh, No... (Score:2, Insightful)

      by grahamm ( 8844 )
      How do you detect if someone is accessing via a proxy (maybe not even knowingly, as some ISPs have 'transparent' proxies) which does not honour the no-cache and similar directives? If such a proxy retrieves the page is there any way of telling if it is then served to 1 or 1000 people?
      • No, which is one of the main problems with counting your sheeple.
        Here's some info [goldmark.org] on that subject.

        The fun really starts when you try to deal with large accelerator-cache farms that AOL and I guess most other large ISPs are using.
        As I've learned just recently a visitor coming via AOL can actually change her IP address *in the middle of a session* because any individual request may be forwarded by any of their n proxy servers to your site.

        So the trace an invididual visitor can leave in your logs may be:
        - 0
    • yeah - next they will be complaining that they don't know if two people are looking over my shoulder at my monitor.

      Web stat tracking, while not perfect is a hell of a lot more effective than Neilson ratings. At the very least Tivo and ReplayTV and the host of Cable companies providing On Demand should have helped to improve tracking.

      Exactly what is Neilson ratings good for at this point ?

    • Re:Uh, No... (Score:3, Informative)

      by Pionar ( 620916 )
      Ok, there's a difference between counting hits and counting visitors. You're wasting money if the same guy sees your ad over and over again, because he'll get desensitized to it and will ignore it after a while. But, you can't tell if those 47 hits from the same ISP is one guy or 47 guys, as he gets a different IP each time he dials in. Is that a unique visitor, or a refresh? Why is this person viewing this one page 12 times?

      In short, sure, you can always count quantity using logs, but it's impossible
      • Oh I forgot the whole issue of cookies, since I know someone will bring them up. You can't rely on cookies for stat tracking because people either have them disabled (stupidly - cookies aren't evil, they can't hurt you) or they clear them out frequently (I do it about once a week during my weekly disk-cleaning routine).
      • All ads in every medium ever have been hit or miss. Why should web ads be any different?

        Until the advertising companies can connect retinal scanners to a database of everything you've ever bought, a la Minority Report, it's going to continue to be hit or miss.

  • Lets show Mr Lamb and the CSM (Score:4, Funny)

    by WormholeFiend ( 674934 ) on Monday June 21, 2004 @08:59AM (#9483783)
    I suppose the CSM is about to discover how many slashdotters view the content of this website...

  • Not So Sure (Score:1, Redundant)

    by somethinghollow ( 530478 )
    Most hosting services come with tracking tools. My host has tools that will even break down IPs to general locations, I believe. It has so many options that it gets difficult to use. So, if you have a good host, you should be able to find out who uses your site w/o any additional work.

    If not, as most have said, set a cookie with a tracking ID. Basically, if you make a website without a decent hit counter (when you need one), you're not much of a web designer / developer. I usually log IPs, user agen

  • Page Counter (Score:5, Funny)

    by Omega1045 ( 584264 ) on Monday June 21, 2004 @09:07AM (#9483863)
    I would put a CGI page counter at the bottom of every page. I think the one with flame numbers works the best for this, but the digital looking on also works well.
    • I like the one with the little monkey turning the crank to make the numbers turn forward.
      • Alternatively, you could throw in a pop-up box that tricks the user into loading Bonzi Buddy, then count how many angry emails you get from users with Bonzi infest^W installations.

        This is a very permanent solution, as after this you no longer have to worry about monitoring traffic to your website.

  • i just dont care (Score:4, Funny)

    by Anonymous Coward on Monday June 21, 2004 @09:09AM (#9483877)
    i dont care who looks at my site as long as my statistics page reports more than just me.

  • You can't: live with it (Score:5, Insightful)

    by Stephen ( 20676 ) on Monday June 21, 2004 @09:10AM (#9483888) Homepage
    You can't measure the exact number of human visitors to a website, any more than you can measure the exact number of people who read a magazine. With a magazine, two people may read one magazine. With a website, one person may come from two computers, or two people from one computer. The problem is only that people, especially advertisers, seem to expect that exact numbers are somehow possible. But they really need to match their metrics to the medium, and not try and force the medium to fit print-media analogies.

    Anyway, the exact numbers don't really tell you anything. You really need to know the differences between two sub-populations (are visitors from pay-per-click ads or visitors from standard search results more likely to buy?). A program which makes this sort of comparison easy will give you far more insight than one which tries to get the total number of visitors closer to some mythical "true" number.

    (I am the author of analog [analog.cx] and CTO of ClickTracks [clicktracks.com], but I'm writing in a personal capacity).

    • I find the value of web logs is more the relative growth of traffic, or from section to section. Since one can assume relatively the same degree of error each month (i.e. 2 users on the same computer, 1 user on 2 computers, etc.) you can gain a lot of information just by comparing logs over time. The same goes for section by section. If your web site has 5 distinct sections you can compare within them and then over time. Advertisers like to know absolute numbers, but if you can tell them that they'll ge

  • Holds true for me (Score:5, Interesting)

    by tuomasr ( 721846 ) on Monday June 21, 2004 @09:15AM (#9483938)

    I found this article to be rather insightful. I personally run a small IT/science-news site (in Finnish) and I'm really having a hard time figuring out visitors of the site. Of course I can get some data from the log analyzing software (awstats and webalizer are being used for the site) but it really doesn't tell me what I want. It seems that the website logs don't always tell the truth. For example I'm getting about 20-30 hits a day with a referrer pointing to a site that's a search engine for blogs (${god} knows why the site has been tagged as a blog) but browsing through the actual logs reveal the hits to belong to a indexing-robot of the site that's a little too enthusiastic.

    The most reliable way to find out about the visitors on a given site would be a user survey, although not complete as not everyone would fill it out, but it would give an idea about the habits of your most frequent visitors. I, if I were an advertiser, would be interested in more than just number of hits and visits and most advertisers would be baffled by stuff like "we got XXXYYYZZZ HTTP requests last month". Personally I would prefer to advertise on sites with a well-built sense of community and an active userbase that's keen to interact with the website, when I browse a site for the first time or a site that I visit infrequently, I rarely click on banners or ads. I'm more prone to clicking ads on sites which I visit daily or so, it gives me a feeling of supporting the site I like and I just might buy something from the advertiser if they are offering something that I need, therefore focused advertising is the key, hence again you need to know your users.

    Logs tell you numbers but you need the visitors themselves to tell you who they really are and how often they visit your site.

    • Logs tell you numbers but you need the visitors themselves to tell you who they really are and how often they visit your site.

      My feeling is, if you are running any type of commercial entity and you don't know who your target market is, you shouldn't be in business.
    • This is not a troll and I'm not trying to offend you, but this makes no sense to me at all. While you think you might want to know how many visitors, eg. humans, you had visiting you, it really is quite meaningless because as others have pointed out, some people read ads and others don't.

      A more meaningful statistic, is how many times was an ad served, because at the other end of the served ad were some eyesballs. You could filter out spiders, who's behaviour is pretty simple to detect and you'd have a numb
  • ...amaze me. I recently helped a friend put together a website for his bakery. Why did he want a website? Because it was something to do that he hadn't done before. Will it drive customers to his place? I doubt it; most small companies like that survive on local ads and word of mouth. I guess my point is that I am still, after all this time, doubtful when it comes to the accuracy of usefulness of ads or site based on visits, click-throughs, etc. I don't think knowledge of the availability of a product is enough; a site must be informative and interactive above and beyond what other forms of advertising can do. While some companies do a great job of this, too many others are like my friend's site---little more than a billboard.
    • I've used the website of one local snackgrill back in my hometown many times to check when they are open, and to see whats the number to call 'em to make some pizza beforehand.

      as long as it's findable with google by typing the place's name like "jaskan grilli" it will be useful in such occasion if they have their own webpage that ranks high enough on google to be the first occurance of that.

      their site is JUST a billboard! BUT THAT'S JUST FRIGGIN PERFECT BECAUSE IT'S A BILLBOARD YOU CAN SEE ON DEMAND FROM
    • Will it drive customers to his place? I doubt it; most small companies like that survive on local ads and word of mouth.

      Not true. Maybe 10 years ago you were correct, but now - a HUGE number of people have access to the net, and probably darn near 100% of people with disposable income have net access.

      Out of those with net access, tons and tons have net access and use it for things like this.

      Now of course I have no statistics, but let me give you an example of how this influences how my friends and I
  • Like television, it's all page views. It's not complicated. It does not matter who viewed, or when, or repeated, or same computer but another person. Eyeballs, a pair of eyeballs come to the page x times a day. This is the number to go with. Now if you have a membership and can get demographics in the sign up form, that is another story.
    • Yeah, I personally think this whole thing is stupid. McDonalds pays to put an ad on TV at 6pm and 7pm, they don't get any sort of discount because I already saw the ad at 6.

      The fact is that just because it's the same person viewing the ad twice - the fact is they still saw the ad twice.

      If it's worthless to see an ad more than once, why do they still advertise Coca-Cola? I'm already aware of its existence.

  • The most interesting is Alexa's model (Score:4, Interesting)

    by Moryath ( 553296 ) on Monday June 21, 2004 @09:18AM (#9483975)
    Alexa's model is interesting - they hand out a "free" toolbar that gives you google search, as well as pinging Alexa and showing you every page's Alexa rank.

    Unfortunately, the toolbar also slows down your browsing (especially if you're on dialup). And the more tech-savvy a user is, the less likely they are to want that toolbar on their system. Thus tech sites are going to be depressed in those rankings, always.

    Alexa also can't tell a subdomain from a regular domain - so subpages of IGN.com or UGO wind up just increasing IGN or UGO's rank, and blogs hosted at X.BlogHost.Com just raise BlogHost.com's rank without being able to tell what the particular blog's rank might be.

    Finally, the biggest flaw in Alexa's ranking system is that it's based on voluntary input; rather than finding 'Net users and trying to get a representative sample (which is the goal of the Nielsen TV setup), they take anyone who'll put in their toolbar. Sure, they can get a pretty large number of idiots to install the thing, but they're still idiots - there are demographics that the toolbar just won't get adopted by in that fashion.

    The other sad thing is, there are companies that use Alexa's page rankings to decide how much they'll pay for ads. Go figure.
    • The other sad thing is, there are companies that use Alexa's page rankings to decide how much they'll pay for ads. Go figure.

      Of course, they do. If they find out that a site has a large number of idiots looking at it, they will want to advertise. That's their target audience.

    • Re:The most interesting is Alexa's model (Score:4, Interesting)

      by yppiz ( 574466 ) on Monday June 21, 2004 @10:34AM (#9484718) Homepage
      Moryath writes:
      Alexa also can't tell a subdomain from a regular domain - so subpages of IGN.com or UGO wind up just increasing IGN or UGO's rank, and blogs hosted at X.BlogHost.Com just raise BlogHost.com's rank without being able to tell what the particular blog's rank might be.
      I wrote much of Alexa's early traffic counting software (I worked there in the late 1990s).

      Your description is partly right. Alexa "rolls up" clicks on subdomains into the doman. So clicks on www1.foo.com, www2.foo.com, and www3.foo.com all count towards foo.com.

      Alexa does this primarily to deal with site mirrors, but also because some sites create subdomains for various functions related to serving pages. So someone interested in Google's overall popularity might prefer to see gmail.google.com, news.google.com, and www.google.com as one site, and not three.

      That said, the site counting software has (or at least had, I don't know if this is still true) rules for detecting home pages as stats-worthy sites independent of their domains. For instance, any URL with a tilde after the domain, like www.foo.com/~bar, has its own statistics. Similarly, there are special rules for recognizing "home pages" on domains like AOL and other big ISPs.

      It's a huge problem deciding what people consider to be websites - it borders on serious AI. For instance, is each Sourceforge project a separate site? How about several subdirectories off of someone's home page, each with a very different focus?

      If you think that your favorite domain should be divided into sites, and that it isn't happening correctly in the Alexa toolbar, you might try sending email to Alexa and asking them to take a look.

      Finally, the biggest flaw in Alexa's ranking system is that it's based on voluntary input; rather than finding 'Net users and trying to get a representative sample (which is the goal of the Nielsen TV setup), they take anyone who'll put in their toolbar. Sure, they can get a pretty large number of idiots to install the thing, but they're still idiots - there are demographics that the toolbar just won't get adopted by in that fashion.
      I am not familiar with Neilsen's current methodology, but I was unimpressed by their marketing claims when they first started their web metrics. At the time (late 1990s) I believe they were saying they had a representative sample of the internet, even though their sample size was: 1) tiny, and 2) made up of volunteers. I cannot say what goes on in Neilsen, or any other web ratings company, currently, but while companies may have very careful statisticians on the inside, often, the caveats and possible biases get stripped out by the marketing department. The moral of this story is, assume that any web rating (or television rating, for that matter) is biased, and understand those biases as well as you can.

      --Pat / zippy@cs.brandeis.edu

  • Webalizer, cookies, stats (Score:3, Interesting)

    by Lord Zerrr ( 237123 ) on Monday June 21, 2004 @09:18AM (#9483976)
    I use webalizer, cookies, and a two stats packages for my cms system (geeklog). One stats package only admin has privalige to, which gives me very detailed acurate info such as time, ip, which page viewed, referers, UID (user id), links followed, country browser, platform ect. All open source. Does the job for me.
  • Most websites have no idea how many people view their content

    We normally use our leaves to view content. Hope this helps the analysis.

  • This isn't rocket science! (Score:5, Insightful)

    by peterdaly ( 123554 ) <petedaly@ix[ ]tcom.com ['.ne' in gap]> on Monday June 21, 2004 @09:22AM (#9484006)
    Tracking unique visitors?
    Not that hard if small margin of error is ok.

    Charging for ads when you don't know how many page views you will get?
    What about CPM (cost per 1k impression) rates? Want 10k impressions? Pay for 10k impressions.

    Target demographics?
    How about track what article topics are popular, how many return readers per topic, etc?

    These are not that hard to do with the right people. The guy who writes the "techie column" in many cases is not the right person.

    I guess if you think like a newspaper, you end up with these problems seeming impossible to figure out.

    Have I lost my marbles, or is this really not that hard?

    -Pete
  • From the article: 'Most websites have no idea how many people view their content. This inherent fuzziness is causing problems for commercial websites, especially online publications desperate to make money from Internet advertising... How can you charge for ads when it's nearly impossible to tell advertisers how many people will see them?'

    Well, websites can just do things to make up numbers. Dead tree publications do it all the time. Ever notice how the the nation's most popular newspaper [usatoday.com] is probably so
  • > This inherent fuzziness is causing problems for commercial websites, especially online publications desperate to make money from Internet advertising... How can you charge for ads when it's nearly impossible to tell advertisers how many people will see them?' Then use performance based advertising - such as cj.com (or buyat.co.uk in the UK). You don't have to sell CPM (ie clicks) but instead get paid on results (eg. sales generated). This solution has been around for a long long time.
  • After careful review of our target audience, we have have begun work on our new bulk Prozac and Lithium banner ad campaign.
  • Back when idocs.com [idocs.com] was my hope for a rich future, I used to keep a sharp eye on the referer lines in my web log. I even wrote some nice perl scripts to summarize what searches people were doing to find me.

    One time the summarizer displayed a search string that consisted solely of pornographic terms: "pussy", "fuck", and the like. I was pretty confused because my site is just an HTML guide. Turns out it found me because of the word "maypole"... I still have no idea what that means in a porn context.

  • I wish Internet advertisers would learn... (Score:5, Insightful)

    by Xugumad ( 39311 ) on Monday June 21, 2004 @09:48AM (#9484267)
    I'm not going to click on your banner. Nope. Not a chance. Not happening.

    It's not that I'm not interested in your product. Online adverts I see actually tend to be:

    1. Something unavailable to me (wrong country).
    2. Something of no interest to me.
    3. Something I own already (this happens a _lot_ with Gamespy).

    But that's not the point. The point is, I'm at the web site because I'm looking for something, and it's probably not your product. When watching TV, I never watch an advert, and immediately decide to research/buy that product. At best I'll make a mental note to have a look out for information on it later, in most cases I won't think about it until I'm looking for that kind of product, at which point I'll probably remember your advert.

    An example might be easier. I frequently see adverts for car insurance. I don't drive, for a variety of reasons, but if I was going to learn and buy a car, I'd probably start calling around the companies whose names I remembered from adverts. Well, actually I'd Google for a comparison site, but lets pretend I'm too lazy to do that, okay?

    Oh, also, pop-ups/unders are a really good way of persuading me to avoid your company, your advertiser, and whatever site I got the pop-up/under from.

    • Re:I wish Internet advertisers would learn... (Score:4, Insightful)

      by Xugumad ( 39311 ) on Monday June 21, 2004 @09:55AM (#9484342)
      Oh, while I'm on a roll (just mod me offtopic, I'm ranting here)...

      If I'm on your site, you have my attention. Stop trying to get my attention with fancy tricks that break my browser or talk half an hour to download.

      Don't resize my browser. If I wanted my browser window to fill the screen, I'd be resized it myself. Equally, if I wanted a poky little window that happens to perfectly fit your site, I can grab that little resize widget myself. It's not like you're saving me effort, as I have to then resize the window back again later.

      Don't tell me your site won't work with my browser. Let me try. Chances are, you've mis-detected my browser, and/or haven't tested in three years, and it'll work just fine if you let me in.

      Okay, going to go get some actual work done now.
      • Don't tell me your site won't work with my browser.

        Instead, how about learning some fundamental HTML skills and crafting a page that any user can read. Seesm to me like the more folks that read it, the greater the chances your ads will have an impact.

        • Oh, had a great one for this recently. We're involved with a large Europe-wide project, and part of what we're doing is usability testing.

          We started looking into this, and then it occured to us to do the project's homepage. The page's content is generated using Javascript. It's a nightmare. It doesn't work on some graphical browsers, on text browsers (as close as I have as a comparison for braille browers) you get nothing to even suggest there's a page there. *sigh*

          • Here's a good one. Try visiting the KPMG homepage [kpmg.com] with Mozilla. You get . . . nothing. Not even a browser compatibility error. Just lovely whitespace and title bar text. There's not a single line of plain text anywhere on the page.

            And it's not just the U.S. site - the Australian page only gives you a single site menu and a couple of colored bars. It's absolutely unforgivable for a company that large to be that ignorant about web design.

  • Mind boggling stupidity (Score:4, Insightful)

    by danharan ( 714822 ) on Monday June 21, 2004 @09:55AM (#9484344) Journal
    Who cares about demographics? We're trying to figure out what people's interests are, what types of ads they'll respond to.

    Well, duh. If a visitor looks at the sports pages during work hours, you have a fair deal of information about that person already. Isn't that already enough to serve up ads that would likely be relevant?

    If these dead-tree publishers of yesterday's news got a clue, they might also realize that web-ads are actionable, and actions can be counted. Do people click on the ads? Do they generate leads or sales? There's this interesting industry called affiliate marketing they should look into (my guess is they'd make good money off personals and job ads).

    What they read, when they read it, and what ads they want to learn more about. WTF more do they need?

  • advertising (Score:3, Insightful)

    by mabu ( 178417 ) on Monday June 21, 2004 @10:02AM (#9484405)
    For those who haven't figured it out already, the web is not an advertising medium. Yes, you can find people who will pay for advertising, but it's a peripheral and unimportant element of the service.

    Hasn't the dot-com-bust taught us anything? Revenue models based on advertising are not going to work except for the rare few who have market share and a steady stream of gullible businesses that want to cheat and try to buy an audience instead of building one.

    Anyone who needs to know how many people are on his/her site and their nature, will already know, and will already have things in place to measure and qualify this. The most obvious of which is sales of their products/services. Traffic reports are amusing but otherwise irrelevent unless you're in the business of selling traffic reports (like Nielsen - another bottom feeder that is providing a crutch to businesses in an effort to continue to perpetuate the myth that online advertising is worthwhile).

  • Earing the right to count your visitors (Score:3, Insightful)

    by amichalo ( 132545 ) on Monday June 21, 2004 @10:19AM (#9484557)
    This is all about earning the right to count your visitors.

    There is NO WAY I am going to spend time giving up my privacy and demographic information if the site has not earned the right to waste my time.

    When you walk into any store in the mall there is a small laser that is counting foot traffic. Each person or close walking couple breaks the beam once to enter and once to exit. It isn't precise, but it is close enough and further the store EARNED THE RIGHT to count visitors becuase there is a reward - viewing the merchandise. Plus, there is a very low cost (exposure to a low powered laser).

    Compare this to a website that would require you to fill out a form, presumably with valid info (the article mentions 90210 as the most popular zip code on the web), and THEN you get to see the content. No thanks. potentially valuable content not worth the bother.

    Now if there was some technology [slashdot.org] that would allow you to store this reader profile and it would be transmitted when you visited a website without the need to fillout a form, I bet some people would use it.

    But no one wants to give their drivers license to the GAP store clerk before entering and there will never be a time that, no matter how valuable it would be for a web site owner, people provide valid, accurate data on who they are to view site content that has not earned the right to ask for that information.

  • How to track? Use Google AdSense (Score:4, Insightful)

    by Goldenhawk ( 242867 ) on Monday June 21, 2004 @10:36AM (#9484748) Homepage
    I have been pondering the same issue for quite some time, as my business depends heavily on internet traffic. I've found one of the best ways to both track traffic, and benefit from it, is the Google AdSense [google.com] program.

    With a relatively compact bit of javascript embedded into a page, the user gets hopefully relevant ads that are not obtrusive or flashy, same as the Google Adwords text-only ads you see on the right side of the Google results pages. And you can customize the colors and format to suit your own pages. Google, while they do serve the ads based on your site's content, do allow you to prohibit certain keywords, so you can block out competitors' ads.

    To make it useful to the host, Google allows you to create "channels", so within one AdSense account you can track different pages. You can get a detailed report of how many pageviews each channel generates, as well as click-thrus (which of course leave your site).

    To sweeten the deal, you get paid for click thrus. That means you get paid when someone leaves your site, but my philosophy is that if they do that, they weren't planning on sticking around anyway, so I might as well profit from it.

    In my case, my site generates about 3000 pageviews and 15 clickthrus, and that translates into about $1 a day in revenue. It's not much, but I roll that back into the Google AdWords [google.com] campaigns that I run, which generate inbound traffic. I'd rather have people coming to my site that want to be here, than those that don't, so I see it as a fair trade.

    And in the end, the reporting and tracking are handled by Google, and provide a tangible benefit to my business.

    Oh, and if you want to see an example in operation, look at the very bottom of our site's main page [worship-live.com].

    • If you're getting 3000 impressions and only 15 click thrus you're oversaturating your visitors with ads. The net result is that they see an ad they aren't interested in and then ignore the rest of the ads or are simply bothered by them to begin with.

      You need to figure out which pages are generating the most impressions and fewest click thrus and pull the ads.

      You should be getting at least a 1.0% clickthru rate. 0.5% is a big sign something isn't working. It could also be a sign you lack the content to
    • I think that might be a problem with the service, though. It encourages fake clickthroughs. I've used Google before. Their ROI is worse than targeted click sites like Overture.
  • My methods are:
    - The webalizer weblogs provided by my hosting provider. Disadvantage is that they mostly provide top-10s. So I get no data on the other pages. Also you count a lot of bots.
    - Google adwords and other advertisers with tracking pixels (like CJ). Problem is that if you compare them they give widely different values for the same page.
    - Nedstat. I like the referer information. But I find it too much work to give every page its own counter.
    - My own counter. Basically a piece of javascript t
  • The article is slanted to sound like it's bad that accurate visitor data can't be had for a website. They fail to mention that compared to other forms of advertising, the Web is a gold mine of information.

    Can a magazine tell you how many people saw a particular ad (without lying that is)? NO. Same for magazines, TV and even junk mail. They might have numbers that are reasonable as to how many people MIGHT see an ad, though take with salt. But how many people actually act on the ad? No sir. How many people
  • I run three tiny sites on one ancient system, and weblog analysis is good enough for my needs.

    For my personal site, don't care. With the on line high school yearbooks for my alumni group, looking for 404 errors from the hundreds of static html pages I hand editted from the initial template, and getting a general idea whether the alums are using it. For the old man's specialty CD-ROM site, just looking for a general idea what parts of the nation/world lookie-loos/orders are coming from.

    Lately, I've also bee

  • I have rolled my own ad-delivery package and I look at my raw logs every month. The metric that I find most useful is page-views. Why? Because 1) it shows the *actual* number of pages viewed per month; and 2) the concept of "page-views" is easy to explain to people who want to advertise on my site, after I get done explaining to them why "hits" is the worst metric. (doesn't discriminate between pages and graphics; the caching problem; blah, blah, blah.)

    Plus, I can divide page-views by unique visitors a

  • This is an old problem, and it's been solved by auditing.

    If site owners and advertisers care about whether the traffic on sites is "real" in any way, then they're probably best off paying for an independent audit of the site's logs. Organisations like ABC//e for example [abce.org.uk] here in the UK will do it, as will various other arounds the world. All use very similar methods and definitions (in fact they collaborate to define standard ways to audit metrics).

    Sure, it's not perfect, any more than ABC's magazine circu

Slashdot Top Deals

"It takes all sorts of in & out-door schooling to get adapted to my kind of fooling" - R. Frost

Close