How Not To Run an A/B Test

Ever wonder what you need to look out for when running a simple multi-variant test? Evan Miller tells you what to not do when running an AB test.

• Use sequential or Bayesian experiment designs
• Only run experiments where the sample size has been fixed
• Stick to a fixed sample size with near-religious discipline

http://www.evanmiller.org/how-not-to-run-an-ab-test.html

Recent U.S. Consumer Data

I recently stumbled upon some recent U.S. consumer data. Did you know?

• 300,000,000 – Number of U.S. citizens
• 200,000,000 – Number of U.S. citizens on the FTC’s Do-Not Call List
• 86% of all people skip TV ads.
• Technology trade show spending has declined 46%.
• 44% of direct-mail pieces are never opened.
• Three of four inbound marketing channels* cost less than outbound channels, like the above and other traditional avenues. The only exception is pay-per-click (PPC) marketing.

* Search-engine optimization (SEO), blogging and social media
Sources: FTC (July 2010); The Guardian (Aug. 2010); TSEA.org (March 2009); EPA.gov (Nov. 2010); HubSpot (2011)

Monitor Traffic with Quantcast

Looking for another lens to measure your website traffic? Quantcast will send you free reports that analyze your site traffic however they measure your audience. Who are becoming customers through your online channel? What do your visitors our your website like? Sign-up and try Quantcast and you can get another view of your website.

http://www.quantcast.com/