Forecasting Techniques

Article Library

Start with a good repository of articles. You should create a significant number of articles in a variety of topics. Good article diversity and quantity will ensure you have a ready-made test bed for a variety of projects.

This article library can be an article directory (eZineArticles), one or more blogs, or your website copy portfolio. Using your own blog or website gives you the added benefit of driving prospective business traffic too.

Since Twitter users, like most social networkers, don’t like constant self-promotion I suggest you routinely test using others websites, blogs, or articles as well as your own. The other authors will love the traffic, may link back to you, preserves your credibility, and it still achieves your conversion testing.

Tracking Conversions

Don’t forget the primary objective–tracking conversions. If you are using your own website or blog this is pretty simple–look at your blog stats or Google analytics. However, what about Tweets that you point at websites you don’t own?

In the past there was little opportunity to collect measurable results from these tests. You could use Summize to track keyword discussion and possibly track re-tweets. Unfortunately, this only measures buzz and typically has more to do with the destination content, not the headline. It doesn’t give you the core metric–do people click because of my copy?

Enter bit.ly Tweetburner, a new tiny URL technology that lets you track source and traffic through your shortened URL. So, now simply shorten and attach a unique bit.ly Tweetburner URL to each of your headline or short copy tests. Then you can track sources and volume of click-throughs.

Respect Your Audience

The quickest way to damage this excellent focus group is to abuse it. So, here are a few cautionary etiquette suggestions to keep yourself from poisoning the water:

• Promote others, as well as yourself, with your headline tests
• Stagger test headlines over various days
• Make sure the destination content is interesting
• Do not link your headline tests to sales letters or landing pages
• Participate in the community too (ask questions and participate)

Happy Testing!