I noticed an interesting comment on my previous post on the launch of Google’s AdPlanner, from Gian Fulgoni, Founder and Chairman of Comscore, a company that has gotten hammered in the aftermath of Google’s launch. I asked if he’d elaborate, and here’s the interview:
In your comment on the Searchblog post noting Google’s Ad Planner, you noted discrepancies between publisher’s server logs and Google’s numbers. Can you say more? Why is this?
I suspect the main reason is that traffic numbers from server log data are inflated because of cookie deletion whereas panel metrics don’t rely on cookies and so aren’t affected by cookie deletion. As an example, Google Ad Planner shows mlb.com as having 9 million UVs in a month. comScore shows mlb.com as having 11.9 million UVs and mlb.com themselves have claimed they get 19 million UVs based on their server logs.
Separately, I’ve noted comments on the blogosphere from several site operators saying that their Google Analytic UVs are twice as high as their Google Trends UV numbers.
Are you concerned about Google’s new product? What are you telling your apparently startled investors?
We think that Google’s products and ours are designed for very different purposes. Theirs appear to represent a point solution aimed primarily at driving ad dollars to Google sites or sites in the Google Ad Network. In contrast, comScore’s products are designed to be used for media planning and analysis on a Web wide basis. We believe that ad agencies, advertisers and publishers (especially any publisher that competes with Google) will continue to insist on the use of objective, third party sources of data such as comScore’s. Nobody wants to see the fox guarding the chicken coop.
comScore offers a broad portfolio of solutions for media planning and analysis that include:
§ Home vs. work vs. university
§ Dictionary hierarchy of multiple levels: property, channel,
subchannel, etc.
§ Ad network view, custom views for companies that represent a
collection of sites.
§ Measurement of duration or time spent
§ Measurement of sessions
§ Measurement of day part
§ Segmentation by heavy, medium and light users
§ Segmentation by content consumption: finance, sports, health,
entertainment, etc.
§ Segmentation by Prism code, PersonicX code, customized customer
segments
§ Source / loss
§ URL level detail
§ Custom entities on the fly
§ Buying power index
§ Ad Impressions
§ Reach / Frequency
§ Advertising effectiveness, including branding, latency and offline
impacts
§ Video Metrix – reporting of online video consumption
Do you trust Google as an arbiter of where ad planners should put their money? If so, why? If not, why not?
My personal belief is that Google hasn’t built these products with the objective of entering the market research industry and being an “independent arbiter” of where ad planners should put their money. Rather, they appear to have built tools that help facilitate the movement of display ad dollars to Google and its Ad Network. Nothing wrong with that if you’re in the business of selling advertising. But, can you be that and be an arbiter at the same time? Perhaps Sarah Fay, CEO of Aegis North America, put it best when she said: “For an advertiser, the last thing you want to do is to have your adviser be the same person you are spending your money with.”
Hello, nice post and discussion. Its aomething inevitable. when a new product/service is launched by google it always shatters all its competition. Google’s ideology cannot be blamed..if its based towarsd there own company..every company has the right to present its offer and be biased towards its product.
But the fact that it results in drastic loss of business to others.
We can do nothing about it
“Separately, I’ve noted comments on the blogosphere from several site operators saying that their Google Analytic UVs are twice as high as their Google Trends UV numbers.”
I was also told a friend of my sister’s brother in law noticed the same.
More seriously, without names and data, this does not stand.
Plus most of the people will not use the correct value in analytics to compare with Trends (you have to search for daily uniques, most will stop at Visits.
In our personnal experience (non US, non english sepaking niche website with around 1 million uniques a month, something Comscore can not adress anyway) we found a 2% différence between Analytics (backed up by french mediametrie) and Trends. Pretty accurate.
Guillaume: ask and ye shall receive. Here’s a link to one of the comments to which I was referring:
http://tech.groups.yahoo.com/group/webanalytics/message/18013
I’ll look forward to receiving your acknowledgment that I have a valid point.
I can also assure you that the Google Ad Planner monthly UV numbers are way lower than either comScore or Nielsen published data. if you e-mail me at gfulgoni@comscore.com I’ll be happy to share the data with you.
Wished you had probed him to detail and elaborate on his theory: ‘I suspect the main reason is that traffic numbers from server log data are inflated because of cookie deletion…’
What evidence does he have to support his suspicions? Also are other variables being taken into account to compensate for this, such as: IP addresses along with proximity of visits?