I first heard about 'The Sleuth Investor' (a book by Avner Mandelman) through the valuable value sites csinvesting and Oddball Stocks. Two days ago, I discussed a book by a value investor who believes in the superiority of a purely quantitative approach to investing. While 'The Sleuth Investor' is also by a value investor, it is in many ways the exact opposite; the author believes in the scuttlebutt approach to value investing, and offers some of his advanced techniques to the reader.
The author believes that investors who rely only on quantitative data are destined not to do much better than the market, if they best it at all. Though the author is a value investor, he invests in companies only when he has thoroughly researched even non-public data.
Mandelman straddles the legal line as he chats up secretaries, follows deal makers, and creates networks of "spies" from company customers to suppliers to its service staff. Buried within some of the more dramatic stories is some pretty good scuttlebutt; Mandelman becomes well-versed in the strengths and weaknesses of a company's offering relative to that of its competitors, mostly by getting to know customer preferences.
Unfortunately, a lot of the techniques lend themselves better to short-term traders rather than true value investors. For example, the methods Mandelman uses can help determine if shipments in a quarter are better than expected, which is likely to do little for the long-term investor. Nevertheless, every investor would probably be wise to learn what other investors are up to, even they choose not to pursue these methods.
Myself, I would say I'm partial to a more quantitative approach to investing. Partly, this is because I generally don't put a lot of faith in anecdotal information (perhaps to my detriment at times!) and partly because I believe behavioral errors (e.g. recency, loss aversion, fear/greed) are likely to play havoc with an investor who emphasizes the qualitative over the quantitative.
That said, I am not expert enough to argue against the methods presented here. I take Mandelman at his word when he claims to have earned superior returns with these methods. Unfortunately, he is not specific with his track record, nor could I find one through an internet search. Searching his name reveals a man more famous for his writings than his returns, which is a tad suspicious but not conclusive. If you're still interested, a direct link to the book is provided here.