Measuring direct and indirect effects comes to the web

Comscore recently announced a new Internet metrics concept called View-Through.  I’m sure I won’t do it justice in explaining it, but through cookies it measures the indirect visits generated by banner ads.  Quoting their own data, “For example, a substantial lift in visitation to the advertiser's Web sites can be observed in the weeks following an exposure to a display ad, even though click rates are less than 0.1 percent. Specifically, there was a 65 percent increase in lift in the week following the first ad exposure and a 46-percent increase over the four weeks following the first exposure, underscoring the latent branding effect.”  In addition, search traffic is enhanced and now finally measured based on the placement of a banner ad.

This is now finally getting to measure the complete impact of web marketing.  The measurement of Marketing ROI for web marketing is now close at hand.   Now we need to add the impact (direct and indirect) of traditional marketing and we will finally have the true marketing impact across the entire marketing mix.

The article can be found at:  ComScore Weighs In On View-Through Proof.

Let me know what you think.  I would love to know. 

Comments

[Guy R. Powell]

John,

Yes, I have used the regression technique and another choice-based modeling technique to calculate the overall impact. It makes a big difference.

As to your question, I believe they are similar.

Glad to see that others are realizing that click-thru is not sufficient.

[John dawson]

Hi Guy,

Is this any different to "post view" vs "post click" metrics that are available elsewhere?

You're right to point out that this misses the rest of mix in terms of contribution. I've use regression techniques to successfully estimate these impacts in the past. It very much depends on the point of the campaign. Simply driving traffic isn't worthwhile in itself. It may well be the case that with certain campaigns you can improve the quality of traffic which would be a useful thing to quantify.

Interesting reading as always,
John