Search Engine Optimisation - Jill Whalen workshop at Webstock 2008

The following are notes from Jill Whalen’s Webstock workshop on 13 February 2008. I’ve mixed in some things I’ve found useful over the past year as well.

Jill’s definition of search engine optimisation is: “making your website the best it can be for users and search engines.” Her major theme was that you can construct your site and write your copy in a way that makes it both a good experience for the user (who can quickly figure out who you are, what you offer, and whether your information/service is relevant to them) and attractive to search engines.

Personally, I’ve found the SEOMoz blog really useful, and subscribe to their RSS to keep learning in small bites. The following are good backgrounders:

How search engines operate

The basics of search engine friendly design & development

An exhaustive list of search engine based keyword research data


Search engines don’t know you (and neither do the searchers)

Every page must provide specific information about what it offers, in plain language that uses natural keyword phrases.

Jill was particularly emphatic that homepages need to carry a concise summary of who you are and want you do, for rapid orientation.


Useful tools for researching and assessing key word phrases

http://wordtracker.com/ (subscription, has a free trial)

http://keyworddiscovery.com/ (subscription)

Not mentioned by Jill, but also useful:

Google Trends can be (freely) used to compare two potential keyword phrases.

Website Grader (haven’t tried personally, but recommended recently by TechCrunch).

Useful overview: Stoney G deGeyter & Jason Green, ‘Keyword Research and Selection’ (PDF)

And, of course, checking your web statistics and server logs to see what keywords bring people to your site, and what they search for once they’re in it.


3 types of keyword phrase

Worst: general keywords (real estate, financial services) are highly competitive and not specific.

Best: 2-3 word phrases which use well-researched language to draw relevant and specific traffic (free retirement savings advice).

Long-tail: uncompetitive keyword phrases; if you use these naturally when writing your copy, you’re likely to get this traffic. News items on a website are natural homes of long-tail keywords.

You can gauge competitiveness for a keyword phrase using this search in Google – allintitle: “your keyword phrase”


Where to use keyword phrases

1. Title tag (most important)
2. Link text
3. Alt tags (keeping accessibility needs in mind)
4. Headlines (according to Jill, these don’t necessarily need to be in H tags: search engines will pick up on the way they’re separated out from regular paragraphs on the pages; bolding and italics help)
5. Body copy
6. Meta descriptions (Jill not so fussed about these)

Important: words in an image file (e.g. a logo) are invisible to search engines. They also can’t see words in drop-down menus.


Search engines need to follow internal links

Do use:

• 'A HREF' tags
• in Javascript, 'no script' tags
• descriptive link text
• image links (with appropriate alt text)
• CSS menus

Not mentioned by Jill, but a relevant point: internal links in blog posts are useful for SEO in the early days, before you have many incoming links. See Aaron & Giovanna Wall’s ‘The Blogger’s Guide to SEO’.


Using keyword phrases without writing fugly copy

Good content:

• speaks to your audience
• solves problems, answers questions, provides information
• is descriptive and specific
• has enough copy to support the keyword phrases

Starts with:

• keyword research
• basing your text around the phrases you select. Put them in the title, the headings, some links, and a couple of times through the copy.

Targeted editing: if you don’t have time to rewrite screeds of content, try doing the keyword research and then making targeted changes to pages (e.g. if you use oenology and the rest of the world talks about winemaking, change that).

Multiple spellings: e.g. ‘health-care’ vs. ‘healthcare’; ‘optimisation’ vs. ‘optimization’. When this question was raised, Jill advised against using more than one spelling on a page, but suggested using alternative spellings on other pages. Person on the floor commented that this is where metatags would be useful.


See your site like a search engine does

Jill suggested using Google: drop your page URL into the search box, then on the results click ‘Cached’ and then ‘cached text only’.

Not mentioned, but which I’ve used before: SEOBrowser does the same thing in one step, but with a bit of a delay returning results.


The truth about page rank

Jill appears to discourage her clients from focusing too much on Google page rank as a way of measuring how successful a SEO campaign is: Her argument is that (a) SEO campaigns are medium/long-term investments, not quick fixes and that (b) you should be looking at your site traffic, not your page rank, to assess success.

Jill also pointed out that with widespread use of Google products (Blogger, Reader, personalised homepages) when you’re logged in Google is figuring you out and serving you up results you’re likely to want to follow. This affects page rank returns, and may explain why your site comes up 1st when you search on a certain term, and 9th when your colleague does the same search.


Observations

An interesting, but unanswered question: what weight do tags ( the social type, not the mark-up type) have with search engines?

Overall, I came out of the session thinking that as long as you follow good practice in terms of information architecture and write properly-formatted, user-focused Plain English copy, you’ve got the basics covered. After that, it comes down to how competitive you are, and that’s where a library or government department might be different from a real estate agency or financial services provider. The key is to use the language of your customer, not that of the business: check out this recent post from the NYPL blog to see how difficult this can be at times.