Sunday, October 20, 2013

Web Analytics - Reduces Ineffective Ad Spending

A sudden spike in web traffic and investigation into the cause and effect led me to revise a content ad campaign on Bing and Yahoo partner sites.

Traffic suddenly increased to several domains. Web analytics showed the origin of this traffic and the behavior of visitors revealed a problem with a #Bing content network ad. This traffic was alarming because of its magnitude. So much traffic would soon have some unpredictable impact on search rank or signal out of control ad spending.

There was no rise in ad spending anywhere. But reviewing the entry page tracking parameters, clearly the traffic was coming from one Bing ad campaign. The campaign was old and the link was obsolete. That provided insight to update the landing page.

The #bouncerate for this traffic was quite a bit lower than average so it looked like real visits. It looked as though visitors would arrive on a page, and continue within the site for several more pages, all fine from an SEO viewpoint at least tentatively. #Analytics showed that the referring domain was Clicking back to the referrers, yielded no insight at all. Our IBM NetInsight analytics showed the keyword was "railway rolling stock" on many of these visits. It's not a common term and the rise in traffic looked very suspicious.

A separate #analytics check on abandoned products – entirely unrelated to the spike in traffic - showed that many of the abandoned products were logged by visitors who came from and the entry page was always tied to the same URL even though the search keywords had nothing to do with the landing page.  One query especially caught my attention, “how to seduce married women” which really has nothing to do with “railway rolling stock” and the abandoned products did not relate to railway rolling stock either. Digging deeper in the abandoned products report, I learned the top abandoned products were tied back to visitors searching unrelated terms and had referrers from Analytics are great for this kind of insight, and they led me to discover and fix part of a content campaign with a #Bing #Yahoo partner.  Eliminating these ads from spammy pages by excluding from all campaigns saved money and squeezed out ineffective spending.

Thursday, October 3, 2013

Book Review: Social Media Metrics: How to Measure and Optimize Your Marketing Investment, Sterne

Sterne writes about measuring the impact of communicating via social media in his book, Social Media Metrics:  How to Measure and Optimize Your Marketing Investment. He makes good points, tells some personal stories and delivers practical advice on the tools and methods and importance of measuring the impact of social media.

Chapter one sets the stage with familiar information about early social media initiatives. By chapters two and three, experienced practioners will begin thinking they should take notes. I found myself thinking, when is the last time I checked Technorati or is Twitter Grader still operating? My colleagues agree when I say I cannot keep up with the amount of change in technology. Just when I thought I was safe with the blogs, LinkedIn, Facebook, dozens of Twitter accounts and YouTube, along comes Google Plus. Another mouth to feed another in-basket, more circles, more ripples.

Sterne quotes Jodi McDermott, talking about widgets and  the difficulties of measurement when things are changing on the fly, "This market is rife with start-ups in college dorms...They like changing things on the fly daily for fun."

I decided to try some of the tools Jim wrote about. I couldn't reach Twitter Grader. reported my main twitter account earned just a "casual" score, even though it has almost 5,000 tweets and 2,500 followers. I found the same account downright invisible on for the main topic of the account.

I got this message once on "You have been Rate Limited by Twitter
The technical explanation is that Twitter has changed their rate limit strategy for search, dramatically limiting Twitalyzer's ability to gather data on non-client accounts. We are working through this challenge now and hope to have a solution in the near future.

The non-technical explanation is ... Twitter won't give us the Tweets we asked for, the meanies!"

Getting back to the book, it provokes, it reminds, it will encourage you to think why measurement and objectives count. Most important what I walked away with, is that while Jim is explaining metrics, he very subtly reveals a strategic approach to using social media. Read the book.