Every website which has a purpose can be suitable for AB Testing. The question is always around where the practise is more useful.
Transactional websites, such as Retailers, Travel companies etc., have a very easy use-case for A/B Testing - one could look to any aspect of the journey, test improvements, and stufy the improvements in revenue.
Finance companies, such as Banks, Insurers, etc. also utilise AB Testing - for example if you have a goal to sign customers up for Savings Accounts or purchase Insurance, the principle of testing changes and studying the improvements in Conversion Rate are no different to Retail/Travel.
Content-driven websites, too, can utilise A/B Testing. For example, if a News blog has some goals behind the website - articles browsed, time on site, how far people read, if they see or engage with your adverts, etc. - you can again test changes and study the impact on these metrics (Key Performance Indicators / KPIs).
The websites where you'd struggle to A/B test the most are ones where it is extremely difficult to measure the impact of your work. For example, a personal blog where you don't care too much about how many people read it or who interacts with it - without a goal, there's little purpose behind A/B Testing.
There are also some limtiations in terms of traffic numbers where A/B testing is not viable. If you have fewer than 200 actions on your KPI per month, even a sizeable change is unlikely to make a measurable, and statistically-reliable difference.
Hope that helps.
Sandeep from the Webtrends Optimize36 team.