Comments on: Do Ads Driving Firefox Downloads Affect Firefox Downloads?

By: Aman

Aman — Tue, 30 Dec 2008 13:37:41 +0000

Hi, We at a large portal wanted to determine exactly what you are. Whether ad campaigns cannibalize or uplifts organic traffic.

- Daily turning off turning on is imho not the best strategy as it doesn’t account for seasonality, news coverage, press releases, media presence etc.
- looking at the data at such a global scale may not tell the story, Id look at campaign level or keyword level to see.
- campaigns need to help uplift conversion for keyphrases where the natural rankings are not high on the SERP
-Any website traffic will tell you that weekly data has a certain pattern and attributing downloads to just days of the week is oversimplifying

By: Abhi

Abhi — Sun, 02 Nov 2008 12:53:45 +0000

Yes, things do make sense from the numbers. Most of the ad-generated clicks are made by those who are not really bothered to go to the site and download FF3 despite the buzz. If these users, amid the buzz generated are presented with an easy link to download the browser will take a shot and download the browser, 90% of the times either because they want to try out something new or something better. This ad driven marketing is almost like a last finishing touch or garnish over the already present marketing campaign in place. Hence the numbers!

By: Arya

Arya — Sun, 19 Oct 2008 19:29:59 +0000

I would say the results are expected. Think of this: You are not happy with your browser & want a better one. You either know that FF exists (power user), ask that tech savvy colleague/ friend or read some reviews (normal user) or call up tech support (novice). This 1% success rate is the users you are catching at a time when they are looking for a new (replacement) browser

By: Harry

Harry — Tue, 07 Oct 2008 06:21:38 +0000

Firstly, thanks to Ken, Alex and Blake for producing such an excellent blog!

The results of this study area really interesting. I’d love to see this study run again, over a longer period, and perhaps with a bit more detail regarding the actual ads and search terms.

Also, alternating the experimental condition according to fixed days of the week is a bit suspect. You probably need to counterbalance day of week since this is probably going to vary in a weekly saw-tooth pattern anyway.

I wonder if the last week of July accounts for a seasonal variation in a large country. E.g. students back to university, or a national holiday somewhere.

By: Web Strategy Guru

Web Strategy Guru — Fri, 03 Oct 2008 10:44:34 +0000

These figures are surprising, I would have expected a more significant difference. Can you post some of the ads here? Maybe the communication needs tweaking.