Traffic Fraud Inflates Video Site Popularity

Dotnaught writes "A new study by spyware researcher Ben Edelman finds that spyware-driven traffic inflation is common, particularly at video sites. The study identifies Bolt.com, GrindTV.com, Broadcaster.com, Away.com, RooTV.com, and Diet.com as the beneficiaries of spyware-driven traffic. 'Our measurement systems are inaccurate for the amount of trust we'd like to put into them,' Edelman said. 'So that's the puzzle: How do you build an advertising economy when the number can't be trusted?'"

Comments

Who??

[Mr2001]

I've never heard of any of those video sites. Is this an actual problem affecting well-known sites, or just these no-names?

Re:Who??

[psaunders]

I care even less...in fact, I only came in here because the footer below the abstract said "7 of 9 comments". Ripped off, she's not even here.

Re:Who??

[Jah-Wren Ryel]

I've never heard of any of those video sites. Is this an actual problem affecting well-known sites, or just these no-names? I don't think it matters. With yoogle poised to do "revenue sharing" for videos hosted on their systems, the described abuse seems likely to become more popular. Both directly for profit and indirectly as "clumsy" joe-jobs to deprive the 'competition' from receiving valid income.

I believe the fundamental question about building an advertising based economy on untrustable numbers is indeed key. This attack is the equivalent of someone figuring out how to plant a remote-controlled tv remote-control in every Nielson living room and using it to fool Nielson's tracking into thinking the families where all watching certain shows - ones for which the producers had paid the remote-control controller a fee. If that were to happen, billions of dollars of tv advertisment revenue would be at risk.

The internet makes such an otherwise impossible attack relatively easy. I suspect the only long-term solution is to not base the economy on advertising. Find another way, my personal favorite being something along the lines of "global co-op comissioned work with release to the public domain." In other words, pay for actual creative work done, not unreliable statistics about eyeballs and promotion.

Re:Who??

[tinkertim]

I've never heard of any of those video sites. Is this an actual problem affecting well-known sites, or just these no-names?

Many web sites will gladly sell you a place to put one of your banners. They will charge you a fee that is justified by the amount of visitors who will see your banner, further justified by estimating how many of those visitors would be likely to click on your banner and why.

For sites that depend on selling advertisement space to monetize, traffic scores calculated by third party sites like Amazon make or break your ability to get prime bucks for prime space on your web sites.

Comapnies who need to show an instant boom in traffic sometimes employ the use of spyware that can be signaled from a remote connection to begin "surfing" a given site from the visitor's IP address proporting to be the user's default browser type. Instantly, millions of people start surfing the target site completely unaware they're even doing it. Its a booming business, building and renting these networks.

You may not have heard of any of these sites, but I'm sure you'd pay top dollar to advertise there once you saw their traffic scores. Its a new cottage industry that thrives on Windows / IE users.

Re:Who??

[owlnation]

How do you build an advertising economy when the number can't be trusted?

Find a way to get someone to post an article on your noname site on Slashdot -> get loads of visits.

No, I've never heard of the sites either.

Re:Who??

[QuickFox]

I don't think they can afford to mix-and-match Good Google/Bad Google Unfortunately they can. They actively promote and encourage domain squatting.

This is how...

[Duncan3]

How do you build an advertising economy when the number can't be trusted?

There's a sucker born every minute. Customers and advertisers both. Google proves it every day. Even the price comparison sites are becoming bogus.

Widget online: $3
Shipping & "handling": $25
Markup for the ad we had to buy to get you here: $47
On sale at the local mall: $2

Now that Google is taking over the entire ad space, it's one simple entry in the ad blocking software to eliminate most ads. Get Adblock properly configured and you'll rarely if ever see an ad.

It's a problem everywhere

[Evets]

This problem has been around since the begining of web stats in general. There was a time not long ago when people didn't differentiate between hits and page views or visits. 100,000 hits on a given site could mean anywhere between 1,000 and 50,000 page views.

Some people intentionally inflate their stats, others end up inflating them unintentionally. Drudge reports an absurd amount of page views in their advertising page, but if you stay on the home page for any length of time you see the page auto-refreshing. Does that count? If you are selling CPM advertising, it probably does. If you are buying it, you hope it doesn't.

In the end, advertisers either are doing brand advertising or conversion advertising. If they are doing conversion advertising it's simple - identify potentially good advertising locations and figure out the comparitive ROI with a trial run. If you are doing brand advertising, you can base your dollars on alexa or nielsen or some other marketshare stat vendor, or you can simply research the site niches yourself to determine the extent of their advertising power within the community.

Advertising has been wrought with snake oil vendors since the beginning. Nothing has changed and nothing ever will. Like anything else - if a deal is too good to be true, it probably is. And just because a deal is priced in congruency with the rest of the market doesn't mean that you can accept it at face value. PR firms don't just exist to put out a public image, they exist because they are supposed to understand the advertising marketplace better than most people would ever care to.

How about a "rate the ad" system?

[Opportunist]

When an ad is annoying (like all those in-your-face popups), at best they serve to get the ire of the user, get him to install popup blockers and other means of not having to deal with an ad.

How about a way to "rate" ads. Was this ad helpful? Did it provide information you actually wanted (I know, I know, but those things DO exist. But then, I also believe in the yeti)? Was it intrusive? Or downright nasty and obnoxious?

I'm pretty sure the advertising industry (the industry doing advertising, that is, not necessarily the industry that makes the ads) would be very interested in that information.

And as a nice side effect you get an immediate feedback about popup-abuse. Because they would most certainly be tagged "obnoxious", no matter how good the ad itself may be.

Re:Party over.

[king-manic]

Advertising does create value for someone. You need to know whats out there and no one has prefect knowledge. Advertising provides a base point to get more info. Suppliers too. Like it or not A great product with little advertising (Linux/stewarts soda) will always be beat by a okay product with good advertising (Windows/coke). With marketing being equal then other factors comes into play (VHS vs Beta). So while the consumer gets nothing out of it the retailer/manufacturer has to play the game.

Same problem for the Italian TV, no solution

[YA_Python_dev]

FYI we have the same problem in Italy for the number of TV programs viewers.

There is a corporation, Auditel, which is in teory indipendent but in practice is owned for the 66% by the two biggest TV networks (RAI, the crappy public TV, and Mediaset the crappier Berlusconi's TV).

Their numbers are used for the prices of the ads and the result is that they always greatly overestimate the number of watchers. An infamous case was when, due to a technical problem, the transmission of a big channel was interrupted for 30 minutes and according to Auditel millions of people (a big percentage of the Italian population) continued to watch it anyway, without any interruption and without changing channel!

Re:Same problem for the Italian TV, no solution

[Dogtanian]

In truth, while Raiuno is really bad - like UK tv from the 70s I should point out that the 1970s is considered by many to be the "golden age" of UK television; and that UK television is in general far better than many other countries'. So whether your comment was a valid insult depends on the context you meant it from; was it that of a modern viewer from another country?

How is that different from anything else?

[argent]

How do you jusdge the value of advertising anywhere? You look at how it effects sales.