Web Traffic: Counting Machines or Counting People

“Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: ‘There are three kinds of lies: lies, damned lies, and statistics.'”–Mark Twain

Numbers may be the purest language, but there’s some impurity when it comes to counting on the Web. A recent article on The Wall Street Journal explains that despite new amazing tools for tracking Web traffic on your site, the interpretation of the numbers can be flawed.

For instance, Facebook has more users than Yahoo, according to Compete, a traffic monitor. But other monitors say the opposite. Both are right; they have the numbers.

It’s not enough to count unique visits only. The article points out that a person may visit a Web site from home, work or on a personal device. A counter may see those three visits as three different people. Or a counter sees multiple people sharing the same computer as one user making multiple visits.

Technology gets in the way, as well. Web sites can assign unique tags–cookies–to each visiting Web browser, but many users delete cookies regularly. Thus each visit counts as a new one, because the site finds no cookie already in place. Bots inflate the numbers, too; search engines send automated browsers out in the Web for each search request, and these bots are then counted as people.

One marketing research company, comScore, is solving this problem from all angles, using what it calls a unified approach. Client sites use their own tools and a comScore cookie to measure users. Bots and international visits won’t be counted. ComScore uses its data to factor in deleted cookies and repeat visitors, also. The most common result for users is a traffic increase for their site.

But don’t insist your own client use comScore’s tactics. Some heavy-traffic sites refuse to work with comScore, saying the service–with a $5,000 fee over six months–is akin to paying for traffic increases. And other trackers have similar formulas, which give similar traffic numbers.

It’s a step ahead, even with the confusion still remaining. It’s good to count numbers and call it a day; but it’s better to count the right numbers and know what you are seeing.

This entry was posted in Godspeed Communications. Bookmark the permalink.

5 Responses to Web Traffic: Counting Machines or Counting People

  1. cnaughton says:

    The wonderful word of statistics…I think Twain says it perfectly. Numbers are an important part of this business because they help to provide proof and support in research and evaluation, but it’s how those numbers are framed that is more important. I agree that it’s better to count the right numbers to really know what it is you are seeing. If they aren’t accurate, what’s the point? They don’t mean andything, they just become statistics.

  2. aguido says:

    I agree with you that this is an issue with many Websites. At my internship, I do a lot of my work on updating our site–it is my most favorable part of my job. I am very interested in the site statistics for visitors, but our tracking system is very flawed. It does not give details about the visitors, like what kind of device it is from, or even unique visits! For all I know, half of those visits could be from me as I check every little change I make to a page, or perhaps the monitoring system doesn’t count our business computers. I have no way of knowing.

    Sure, I wouldn’t dare ask them to spend $5000 to count unique visitors and the like, since our whole business isn’t web-based (not to mention it is non-profit), but it would be great to know who is looking at our site. We ARE based a lot off of our target audiences so it would be great to see how many of what demographics are viewing the site. I know that is veering a little bit off course, but I think it all ties into the idea of counting you bring up. There is definitely a lot of confusion in such numbers and who are the real browsers and who are not.

  3. crandell says:

    A number isn’t just a number without context to it. Web trafficking is so important when it comes to public relations. If done correctly, it can be used as a form of evaluation. Clients need to know if their web traffic has increased because more individuals are seeking out the information from different outlets or the same person keeps returning. Both are equally as valuable. This topic goes hand in hand with search engine optimization. We can’t focus on increased the accuracy of web-based searches if we don’t have an accurate way of measuring web traffic. Also, the more we know about which site attracts the most people, we can begin to ask why. We can use that information to our benefits in public relations because we can tweak our clients web page to contain certain qualities and measure what works and what doesn’t.

  4. tburns says:

    Numbers can be manipulated to get whatever you want pretty much, which is why when doing research, numbers cannot be all that you rely on. This is where the importance of qualitative research comes in to back the numbers up (or vice versa). It’s the combination of the two types of research that make knowledge and expertise strong and accurate.

    With that said, I think counting the number of people who visit your site is only half of the battle. For example, my group’s client, the Human Tribe, is an organization based entirely online. If we counted the number of unique website visits, and it was a high number, and the number of tribes we had or the number of necklaces sold was low, what good is it knowing the number of website visitors if there is not an end result of our service being used? Sure, if the number of website visits is low, then the number of tribes or necklaces purchased will be low, and I know I have a problem with awareness. But if I have a high number of visits and a low number of tribes, then that means there is something wrong with either my website itself of the service/product I am providing.

    Therefore, I think it is important to always look beyond the number of website visitors to know what you can do to make your company better. Pair the number of visits with some other bit of information and knowledge so that you can figure out what is working and what is not. The Internet is a strong force of nature, and although it is a dominant feature of life and society now, we cannot just stop and look at the website— we got to look at the whole process and system.

  5. cmcelroy says:

    Last year I took an intro course on social statistics, and if I learned one thing it was not to trust them. It’s exactly like you said: everyone’s right because they all have the numbers to back it. The difference is the context. You can take any fact and twist it to convince someone of what you want. I remember learning both in my reporting and editing classes that as a writer you should always use the statistic that works in your favor. Sure, I can say that a budget increased .05% in the last year, but doesn’t it sound much scarier when I say it increased 4 percentage points? I’m not sure of the science behind web traffic counters, but I’m not surprised to hear that there are discrepancies in these stats too. I think in any situation it is extremely important to check where the facts are coming from. Is it a reliable source? Can they explain what exactly they are counting and how it determines unique visitors? I used ComScore this summer for research at my internship, and I found it to be one of the more reliable avenues to pull web stats from, but I am far from an expert.

Comments are closed.