Taking into consideration time and cost restraints it's not feasible to produce optimised doorway pages for each search engine, especially if I have 10+ keywords to optimise the pages for.
Would it therefore make sense to produce a single keyword optimised doorway page for the most popular SE (Google?) with a view to getting reasonable rankings for other SEs.
If so how should the optimised page be produced, Keyword Density, Keyword Proximity, use of Headers, bold and italic text etc.
Everyone here has their own ideas of what keyword density to use and other techniques e.g. headers, titles, meta tags. Many of these things are fairly closely guarded secrets and no one is just going to tell you exactly what density to use.
What I will say is get rid of as much redundant code as possible out of your site, this will increase the density and allow you to promote your text closer to the top of the page. It will also give you more leeway in which to work on density.
Tip 2 - look at what density / page rank your opposition has.
Sorry I can't be more help ;)
I'd say that it does work, but not on its own. If you do good keyword research, and work your key phrases across the site (onto relevant pages) using the TITLE, BODY text and incoming link text then you'll usually do better than people who don't do those things and have similar PageRank.
If other people have better PageRank then you'll not beat them if they work the site design well, but you might well do if they don't. If other people have lower PageRank than you then you might beat them with a less optimised site.
The hardest part is getting the incoming link text to match your on-page content. There are some things that you can do to help with external links but an ODP editor (or other site that links to you) may prefer to use "companyname" than "red fuzzy widgets in mycountry". Links from your main pages to deeper ones can look odd if you use minor variations of the same phrases to link to sub-pages, so you can choose to compromise or cloak (I favour the former).
ODP and Yahoo! do help, but unless you get into a category with decent PageRank you're likely to find better links from elsewhere.
On the original question, I personally favour the 'optimise for Google' approach, partly because Google is a great driver of targeted traffic (especially for a large, well made and well linked site), but partly because the other engines have tended to move in Google's direction.
However, people who concentrate in Inktomi (usually those who pay for inclusion) seem to work aspects such as keyword density much harder and often report very good results.
Calum
So I take the easier route; keyword-rich descriptions in the major directories. Might have to check those category PRs, though. Great tip, thanks :-)
Red_Tractor, see this thread [webmasterworld.com] for general Google optimisations.
Personally I don't see keyword density as being critical; just having the target phrase in the title and body text has the biggest on-page effect.
The way you use internal links can help a great deal if you get some decent PageRank into your home page.
There are two things that people most often do wrong, IMO:
1. Failiure to decide to optimise each page (or section) of a site around different words. We know where we can find hints for words that people use (overture, wordtracker, competitors' META keywords, etc.). It's also important in Google to think about word order. A search for "hotel mycountry" tends to give quite different results in Google from "mycountry hotel" or "hotel in mycountry" (without the quotes).
2. Reinforcement. The title says one thing, the BODY says another, the inbound link text says another. This does not help Google to understand what the page is about.
Calum