Blogs seem to be all the rage. But do they always fit the situation? I would answer that by testing a blog internal within the company to first test if there are enough topics that could be covered as to create new and interesting content consistently and for a prolonged period, and two whether you can plan the keyword usage effectively enough to create an effective audience and appropriate search traffic.
I advise companies often to test blogging within the internal web space to test whether blogging is a good idea. If it is not, then no harm no foul. If it ends up being a good idea, then you have a populated blog ready to deploy. I cannot tell you how many people start out creating a blog, tweeting, and the like to only have it drop off or fall flat publicly.
Once created and before deploying, you can take measures to see if your blog will add value SEO wise. I always advise having an internal copy of your website that allows you to not only test new ideas but also run various SEO tools against. I have tested internal sites by DNSing the internal site using my DNS server and letting the SEO tools think it was the real-McCoy.
Generally, blogs do add value but you do have to commit to them. Have it on your own site and not on another site. Why drive traffic to another domain? I have seen companies use other blogging sites and have trouble converting the new users. All they really end up doing was creating value for a site they do not own. I tested a small blog some years ago and what a difference it made. It only took about 20-30 entries to really drive traffic. It was not my idea to test a blog, but to host some testing results that I was doing on an unrelated topic. I was stunned by the traffic even though it was not something I was ready to commit to. So after the testing was done, I removed it.
Sometimes you just have to try something to know if it is a good idea. You just have to see it before a rational decision can be made. Just do not do it publicly.