Keyword Research?

For me, that's just about the "deciding line". I'll normally do it if I think I can get to number 3; not if I can only get to number 4. I know the average "proportion of traffic" difference between those two positions is only small (and, of course, variable according to what's above them), but that's just my personal preference, given that the type of niches I find myself looking at seem somehow to have broadly similar quantities of "available traffic".

Yes, I do it very similarly.

I am going to agree with the two on the top 5 sites.

To break it down to you how my mentor taught me was use the Google Adword Tool and search by "Exact" and from highest to lowest on Global Search colum.

From there I can see if there is even enough search volume for me to want to get my hands dirty.

I will take on anything with search volume 5k and above for monthly searches. Anything under that is just not worth it to me unless I am sure that I can convert on those lower numbbers.

Now that I have a good amount of folks searching for those keywords I turn my attention to the Adsense payout (if you are going to promote Adsense). Anything above $1 is good for me.

I next will see how many people are fighting over those keywords in the Google adsense competion level. Anything 95% (3/4) full I'm all for it.

I head on over to Spyfu.com to see how many exact people are advertising those numbers. Anything over 5 people and I'm good.

Roll on over to godaddy or who ever is your favorite domain purchase site and see if you can get an EMD (exact match domain) for the keywords. If the .com is take try .net if that's taken last resort try .info. If all of those are taken then you can resort to placing another word at "the end" of the domain like blackrunningshoesreivews rather than blackrunningshoes. So your straight knowing that you can purchase the domain.

Now that you know you have enough people searching for the keywords, and you know that 5 people will pay you $1 per click in the number #1 spot in Google it's time to access the competition level of the niche.

Here is where a lot of people get things all murked up. It doesn't matter if there is 10 sites or 1 trillion sites competing for those keywords because you are only really concerned with the top 5 or top 10 of those results. The rest of them are not important because you are not trying to take their spot you are trying to take the #1 spot.

Access the competition to see if they are using an EMD domain. EMD only helps you slightly in the serps but then again it's really all those little things that put you where you need to be anyway.

Next check the page that is listed in the number 1 - 10 spot (not the entire site) and see why Google placed them there. Here's a few things to ask yourself.

1. Where they optimized for the keyword you found them with or did they happen to rank for the keyword by accident. (this matters because the backlink anchor text will not be optimized for the keyword and this works in your favor)
2. How many quality backlinks does that page have? This can be seen roughly (no search engine is going to give all the amount of backlinks because it's would punch holes in their security revealing their algo) using yahoo site explorer (this site reveals the most of the backlinks out of the other search engines). When your viewing those backlinks what's the pr of those backlinks? Are they relevant to the topic in anyway? Running shoes getting backlinks from one billion pizza companies isn't as strong at 1000 backlinks from shoe sites. Remember that every link will bring it's own strength over to the site so using some software to review those backlinks can help you determine how strong that page is. Also is majority of the links coming from inside the website itself? This can be because it's ranking off it's authority of the entire overall. Amzon, Walmart, etc.
3. How old is the site? Older sites are now starting to get a boost in the SERPS for some reason. Not sure as to why though.

Run those again the top 5 in Google and if you think you can do everything they have going for them and some then toss the site up and see what you can do.

UPDATE: Also remember that Google doesn't rank websites in Google they rank webpages so every page will rank indepently to each other but if you play your cards correctly and spread that link juice through the site then all your pages will rise faster.

Newbie here..
How can one ascertain the SEO of the top few..?

Unfortunately, it's not quite as straightforward as that, I've learned.

The problem is that there are some sites "at the top" that actually have pretty small numbers of backlinks to them, just because the backlinks they do have are mostly from real authority sites, and you have to try to take this into account, too. Factors like how well-established they are come into play, too.

The good news is that there can be others with huge numbers of backlinks to them which can be really easy to beat.

As a general rule, if I see article directory articles in the top 5 results (which you easily can do, when people have made the mistake of doing backlinking campaigns to their EZA copies instead of to their own site's copies, and so on), I take that as pretty encouraging, and I know I can beat those without worrying too much.

A high-PR "health site" registered years ago that has 50 - 100 traceable backlinks including stuff like WebMD, well-known medical schools, Pub.Med., Wiki and so on is going to be really hard to beat, whereas an EZA article with 20,000 "crap backlinks" pointing to it can be trivial.

It's really "quality assessment", not "quantity assessment".