A while back I did a post entitled “Segmenting Query Spaces by Audience and Behavior“. The research I’m discussing today would be the second step in my keyword development strategy for a site. The third step in the research side is gathering intelligence on the players in the query space. In between I’m getting rankings for the keywords developed in the previous two steps…. competition and content review and we are ready to go.
Keyword Research: The Foundation for Successful SEO
Just so we are straight, for the purposes of this exercise, site structure in search engine optimization is comprised of the HTML elements that contain values such as Title, , headings, strong, lists both ordered and unordered, italics and almost any HTML element with attributes and/or is a “container” of a value, for example, a URI. I would also include sitemaps (both XML and HTML) though I’ve never used either. Bottom line I’m not going to add something I don’t need… just cuz a search engine says so… I don’t do anything for search engines I wouldn’t do for a user. So IMO, sitemaps are crutches…I don’t build lame websites ( may look lame, but architecture wise..Mona Lisa baby! ).
Link architecture is the locations of links on the page, the link text and importance of the page it’s linking to, I have not necessarily put them in order of value because I think each persons style changes the order of importance. I know this works for me but YMMV (Your Mileage May Vary).
When To Change Site and Link Architecture
Let me re-iterate. I will never change all of these elements just to gain a little benefit for SEO. That said, any time a site is getting a major re-development a large part of the architecture and structure will be modified. At this time assessing the current degree of optimization, the phrases you are optimizing for and the importance of the keyword phrase is necessary to decide whether the benefits to SEO will warrant a major re-development or just small adjustments. It is good idea to do new research at this time because things may have changed from the original work. Many firms look for new keyword opportunities or channels few watch the value/importance of those terms to the business.
Personally, I believe if the keywords research is done right the previous site structure and link architecture can be appended to or deleted they seldom need wholesale change. Either poor keywords research was used or implementation was poor. In other words if an SEO says you need to pull it all apart just for SEO… run don’t walk. That said there are several reasons to pull apart a site. Many older sites have few or older community builders or the CMS may be dated. My point is don’t do it for SEO there’s not a lot of benefit for the effort
Optimize for “Keyword Themes” in Site and Link Architecture
I have an older site that I’ve decided to try and see if I can’t restore the prominence it had in the past. A few years back we quit adding content to or promoting the website. We had not done keyword research since before the site was originally launched in 03. We will be building the site out with new articles and a re-launch with fresh structure and link architecture so keyword research was the first order of biz.
The first step in the keyword discovery process is determining the important phrases which will be used in the “site architecture” or, as they are more widely known, folder and page names. It has been a known fact that Google analyzes not just the link text but also other attributes of the href element such as rel= or the src= among others. There is also the common practice of webmaster’s using a url as the link text so naturally site structures with important keyword phrases would benefit your optimization efforts immensely.
Now the fun part starts! For this stage of the work I like to use Wordstream’s NicheFinder to identify the importance of a phrase from a traffic standpoint and baskets of keywords that should be used in building a page or site. I like the NicheFinder because of the ease in building both lists of keywords to determine high level optimization of the site infrastructure but I mostly like the baskets of phrases it builds for page/or PPC campaign level optimization. I’ve found the Nichefinder does this better than any other product I’ve tried.
In this case I’ll be looking for SEO type keywords so first order of business is to determine if the acronym or the actual phrase ‘search engine optimization’ is the best keywords to optimize. I started with the shortest term ‘search engine’ which indicated both by number of direct results but also modifiers (keywords that go with ‘search engine’) that ‘search engine optimization’ was the primary term, one could assume in UK it is optimisation, but, that’s for another time. To verify this analysis I next did a search on SEO which indicated that indeed the money was in getting Search engine optimization not SEO. If this were a newer site with little or no link equity I would target the easier SEO term.
Now I have two baskets of primary keywords prioritized by traffic received. These two baskets of primary/category phrases generally contain several page level baskets of keyword phrases
Note that when the site this refers to was done SEO was just gaining legs. So in the first iteration I had changed the copy thinking SEO would eventually be the main term. I had also already been in the top 5 for search engine optimization and the conversion for the target clients ($10,000+) was getting a ton of traffic but these were mostly low budget/no budget. So I changed the copy so it was more slanted to SEO terms.
Keyword Prioritization for Relevancy and Business Goals
Now I add the business and conversion, target product goals to the equation by adding a dampener for keyword relevance to the site and negative keywords for a PPC campaign. I do this manually making it a very time consuming activity, however, the Task ROI is high since it is a way to further prioritize my activity going forward. I use an excel spreadsheet for tracking and add a column with a formula to adjust the traffic number down if the phrase is not relevant to the business or is less attractive for margin or other business reasons. If this were a PPC campaign I would research negative keywords. A phrase I might remove entirely is Search Engine, IMO, it is irrelevant without a modifier to target activity on. ;-).
Now we have a prioritized list of keywords to use in the site infrastructure, doc structure and link architecture! Now we can begin to develop and implement the keyword phrases.
Implementing Keyword Research In Your Site and Search Engine Marketing Strategy
The most important step in the above is the prioritization because this enables you to always make informed decisions as to importance and value of the phrase when writing Titles, link text and copy. In my next post I will share the process I use to implement the keywords in site infrastructure/structure and link architecture.




So you basically keep the major themes for site architecture and the modifiers for the pages and text with internal linking 🙂
Thanks for this step by step explanation Terry. Very useful information here. Definitely sharing it.
All the best,
Eren
@Eren thanks!
@Dries yes and no if the modifier is low priority yes copy and internal/external linking is what I’m counting on for these lower priority/long tail terms. This exercise is more about ensuring your primary terms are optimized to the max onsite.
🙂 am happy for like this ……
@Terry makes sense. Keep’m coming! I added your feed 😉
“When to change site and link architecture”
“… pull it all apart for SEO … Run Don’t Walk.”
I’d like to smack in the head with a wet sponge for that comment – not because I’m violent, but because I would find it funny and I’ll get your attention.
I regularly completely rebuild websites that have been built using tables for layout (yep, still loads out there), have no folder structure, no sensible page outline and are generally so nasty to work with that it’s quicker (and cheaper for the Client) for me to start from a blank page.
Granted, the majority have been built by people who can sell their website building services, but have no clue (or didn’t bother to try) how to sell the product the paying client was offering ( I feel a rant brewing, but I’ll stay on point).
So, respectfully, I disagree with that comment, however – I like the rest 🙂
@webecho ahhh note I said don’t do do it just for SEO. What you have outlined are maintenance and to a small degree SEO issues (no file structure means no keywords in architecture). Some even want to change the graphic design… for SEO?
To the others Thanks for the appreciation of the work.
@terry
The structure and, although to a lesser degree, the coding makes a big difference to the SEO.
The graphic design wont help with SEO of course, although if you’re rebuilding a website anyway, then it’s a good time to address the design from a marketing / sales / presentation point of view.
Thanks for the heads up on Nichefinder – hadn’t come across that one before.