Trial and error has once again revealed a useful insight for affiliate marketing: there are two kinds of reviews, and if you write wrong kind for your product and target audience, you can kiss sales revenue goodbye.
Examples of this are my review of a cheap hard case for an old Macbook whose polycarbonate case was known to crack, a special kind of baits to kill the little black ants invading California, or collectible figures imported from Japan that are hard to find in the states. Here, the buyer has a problem or an idea what she's looking for, but either doesn't know the exact brand and model name (what to type into Amazon to find it), or it's not available on ordinary retail sites like Amazon.
With all of these products, I use the products's very specific name for the article to help it get picked up in search engine results, or, more often, ("hard case for old Macbook") I did some keyword research with Google's keyword tool to discover what people type when looking for that kind of product.
The "specific name" approach works for products that are difficult to find, or are sought by collectors. The "keyword research" approach works very well if you can zero in on a question or phrase people tend to Google when they're looking for the solution to a problem.
However, there are several examples where the "specific name" approach failed to get much traffic. See if you can determine a pattern:
What's the problem? People aren't liable to type out "review of Herman miller Aeron chair." Either they already have the Aeron chair and have formed an opinion about it, or they go straight to Amazon or some e-merchant they trust and look for product reviews on the site where they intend to buy the product. Why Google for a product review when there's already dozens right on the product page on Amazon? I still get some traffic on these kinds of lenses, but no more than 10-20 a week.
So think about your buyer. Do a simulated run of buying it on Amazon yourself. Is it an item lots of people know the name of? Are the Amazon reviews informative, and the product page tells you everything you need? Does the Amazon listing turn up at the top of Google when you search for it on Google? In that case, you probably won't have much luck, because the typical buyer is going to go to Google > Amazon > Amazon reviews > Buy as opposed to Google >Webpage Review > Amazon > Buy.
That doesn't mean don't do the review, because you might get lucky. However, for the most part, I recommend focusing on products that fit the criteria I mentioned above: they solve a problem, answer a question, are difficult to find, or are sought by collectors. Those get searched on Google, not just Amazon, and users may actually do a little research before heading to Amazon.
The Buyer's Guide
Often, web users trying to compare a bunch of products to pick the best one. Then they need a concierge to steer them to likely products. These articles take more knowledge and experience to write well, because you can't get away with reviewing just one product: you have to be familiar with the niche and give an honestly useful report and comparison of a bunch of different products. But these articles can be very useful to customers, because they can't just go to Amazon and read a few reviews to find what they're looking for.
I make a few sales a day on these lenses: Best Books on Greek Mythology, Great Books on Celtic Mythology, Best Books on Egyptian Mythology. I had discovered these kinds of searches in my old Greece-related lenses' traffic stats. On the same principle, I tried the Best Ice Compresses and got good results. So if you know a subject, and you've done your own comparison shopping to figure out which are the best, write a lens on it, with mini-reviews of the pros and cons of each and comments about what each one is good for.
I've also had tremendous luck with collectibles guides where I showcase all the items in a collection. People don't always buy those things, but they often go to Amazon to browse them, then they buy something— not necessarily any of those items, but those items lured them into the store like a shop window display!
The key with buyer's guides is that you're trying to attract people who are interested in a certain type of thing, they don't know exactly what product name to type into Amazon to find it. This kind of buyer's guide is especially effective for products that turn up a lot of four-to-five star rated products in Amazon search, so many that it would be bewildering and time-consuming to dig through all the customer reviews to find the best one (e.g. celtic mythology). If a search for, say, "eco-friendly Christmas lights" turns up only 5 products that are better than 3-star-ratings, then you're probably not going to get too many visitors to a webpage on that product; it's easier for visitors to skim those five products on Amazon, read the reviews there and make a choice. Again, this isn't an argument not to write such a review, but it's less likely to succeed. This problem is happening more and more, as apps and online retailers provide people many ways other than a generic Google search to learn all they need to know before making a purchase!
Here's a special case of the buyer's guide which is worth considering.
App Store Concierge
Right now, there's no easy way to SELL apps -- at least, I'm not finding an affiliate program with Apple that gets down to the app level -- unless you go for the Android apps available on Amazon. [EDIT: See below; Christene's got tips.] But since Squidoo rewards lenses that get lots of clickouts, there's an opportunity to make some money playing app concierge. There's a real opportunity here, at least for now, because the iTunes App Store isn't really good at turning up items unless you know the name of the app you're looking for. It's got no tags, too few categories, and only shows the top 100 in each category. It's a really inefficient way to find apps. For example, I couldn't really find an app that let me keep track of miles and calories for a treadmill, exercise bike, etc, because Apple's search tool is too inefficient.
So I'm creating a few guides to different kinds of apps. Right now I've got one on creative people's drawing apps, although I put that on Hubpages because I'm trying to diversify.
Basically, if you've had to do some research on, "What's the best app to do X?" make a lens about it, and be sure to make prominent links to the app's page.