At, Lots of Glitter but Maybe Not So Many Visitors – New York Times

At, Lots of Glitter but Maybe Not So Many Visitors – New York Times

The Times slams into this morning on the eternal subject of squishy traffic numbers. This is an issue endemic to the online media industry, one that harks back to the days of Time-Warner’s Pathfinder when Gerald Levin would boast about millions of “hits.” Now that the industry has settled down and focused on unique visitors, there is still a vast discrepancy between the external traffic reporters — ComScore, Nielsen, Alexa, etc. — and a site’s own server logs, ostensibly the only true measure of traffic, yet one wholly dependent on what filters and analytics are being applied to the raw numbers.

With no equivalent to the magazine industry’s third-party audit structure in place (BPA, etc.), online media has been able to play a game of squishy reporting since 1994. Take a good stat, lead with it, and let the rest of the numbers fall where they may.

“But a closer look at the numbers raises questions about’s industry-leading success. For its claim of a worldwide audience of nearly 15.3 million, it has been citing February data from comScore Media Metrix, one of the two leading providers of third-party Web traffic data.

“There are several problems with that statistic, though, and comScore has since revised the figure downward to less than 13.2 million as part of a broader revamping of its worldwide data for many sites. Jack Flanagan, executive vice president at comScore Media Metrix, said the new figures were released “a couple of months ago” after it changed its methods for estimating global audiences.”

While bragging rights are nice — “We’re the biggest” is always a nice marketing message — the advertisers are the one’s who are best placed to develop the metric that measures and that comes down to conversions. Forget CTR (click-through rates, forget reach (monthly uniques, visits), and focus on what happens to the referred traffic once it arrives in the form of generated leads, shoppers, etc. The notion that any media buyer would give more than a passing glance at gross tonnage metrics is risible. It’s their own metrics, how they measure what they’ve bought, that determines whether they’ll renew a campaign or drop it.

Author: David Churbuck

Cape Codder with an itch to write

0 thoughts on “At, Lots of Glitter but Maybe Not So Many Visitors – New York Times”

  1. “But ours go to 11…” – Spinal Tap

    Gross tonnage is an apt description here. Simple things like included pages, etc. can do amazing things to some lesser statistics engines. And they can do it by multiplying factors. And the myriad statistics offered in the big packages require a full time statisitician with a proven ability to read tea leaves.

    Many of the stat’s we’ve seen in the past simply don’t pass the sniff test. I’m totally agree, it’s got to be measured against revenue, quality leads, etc.

    Keep your eyes on the cheese…

  2. correct senor cahill. reading the runes of “negative correlation” makes even the best clairvoyant blush.

    look for “r” and square it.

  3. We’re currently setting up conversion funnels for a Fall campaign. By setting up a funnel for each channel (email, banner ad, collateral drive-to etc.)we’ll be able to measure the perfomance of each.

    The challenge comes when you consider the context of each lead and the expectations based on how the user arrived. It is important to provide the right info in each step of the funnel and ensure the design of the site makes this information easily accessible.

  4. Bingo. Which leads back to the original premise. Is bigger better when it comes to Internet advertising?

    I think not. Thinking big — in terms of “reach — is the model of the 70s when there were three networks and a hit mentality to all media. Skip forward into the vaunted age of the tail that is long and it ain’t the size that matters any more, just the ability to deliver. I remember a sales call when a marketer at a big electronics company said, “You give me 100 engineers responsible for specifying DSL chipsets and I’ll pay you a $500 CPM.” (this at a time when CPMs were depressed below $10 for the general media sites.)

    The only metric in the end is the one that the smart interactive marketer develops on their own: efficacy of placement. Ask any publisher who tells an advertiser that they received “bonus” impressions if that made the advertiser happy. Garbage impressions is more like it.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.