40 replies to this topic

[saschaeh]

(Kim's note, Oct. 9, 2007 - This thread was pinned because we felt it was so educational. The thread's originator later took the time to compile the posts into an article-like version, which you can find here.)

Im relatively new to this game but have a fairly good understanding of SEO. I lack some good hard practical experience which would lead onto my questions:

When you guys start on a SEO strategy for a website what are the initial stages that you go through to ready the site.

> Build the site using CSS and preferable DIV based.
> W3C standard compliant
> Keep main navigation to the top and stay away from JS menus
> Title tags relevant to page and other SEO semantics
> Linking structure - Make things easy for Google to access.
> Create areas for dynamic content and try keep it up adding content.
> XML Site Map
> Robot.txt
> Linking - (not so easy part)

Well im asking more then telling... I know SEO is an ongoing process of adjusting and tweaking but what are the bases you cover to get the site started on a good SEO path?

Edited by cre8pc, 09 October 2007 - 06:31 PM.

[EGOL]

Start by deciding what search engine space you want to occupy... then build a site (or new area of a site) to accomplish that. If I want to sell widgets or get traffic for that theme, I decide what turf I want to own, then decide if I can create highly linkable content for that theme, then build mutiple pages of that content to compete for brass widgets, wooden widgets, red widgets, green widgets. Many pages about widgets, all linking to one another. This creates content depth, authority in the theme and helps you get ranks for many widgets terms.

[Autocrat]

Well...

Document URL
Page Address - http://
Domain Name

• Try to get a Domain Name that has some of the Keywords/phrases in if possible, else go for the Company/SiteName.
• If the Name consists ofmultiple words, consider the primary domain using hyphens.
  : http://www.widgets.com
  : http://www.widgetting-engineeering.com

Page Name / File Name

• Try to make the Page Name / File Name match the PageTitle.
  : http://www.widgets.com/brass_widgets
  : http://www.widgettin...nd_made_widgets

Document Head

Title - <title></title>

• I place it at the very top of the section.
• I try to make it "sementic" = PageTitle + SectionTitle + Company/Site Name
  <title>Brass Widgets - Widgets - Widgetting Engineering Ltd</title>

Meta-Description - <meta name="description" content="" />

• I place it near the top of the section, (after Title, before CSS etc.).
• I try to make it "useful" = PageTitle, +brief detail inc. several key words/phrases, +Company/Site Name, + Contact details(Tel., Add.)
  <meta name="description" content="Brass Widgets - A sample of our wide range of hand crafted and machine milled brass widgets, Widgetting Engineering Ltd, Hampshire, Southampton, SO15 4PP, 02380 123123" />

Meta-Keywords - <meta name="keywords" content="" />

• I place it near the top of the section, (after Description, before CSS etc.).
• I try to make it "useful" = Single Word and Dual Words from the PageTitle/Brief +Company/SiteName
  <meta name="keywords" content="Widgets, Brass, Brass Widgets, Hand Crafted, Hand Made, Machine Milled, Widgetting Engineering Ltd" />

(NOTE: They may not be used by SE's any more - but they sure as hell help Clients to remember what the content is meant to be about - I make them right out Keywords first, then build the content around those focus points) ((How I used to right Essays too ))


Also, Put JS last - preferable in external files!




Document Body

Header Tag - <h1></h1>

• I place it as the first "content" if I can help it.
• I tend to apply bold to it as well.
• Even if the client has a logo in the design - I hide it with CSS negative positioning.
• You could make it a link if the Text is visible
• I try to make it "Meaningful" = Company/Site Name ( you may want to inc. SectionTitle + PageTitle)
  : <h1>Widgetting Engineering Ltd</h1>
  : <h1>Widgetting Engineering Ltd - Widgets - Brass Widgets </h1>

Header Tag - <h2></h2>
These are used in conjunction with the HeaderRule <hr /> and ID for splitting the site into sections and enabling certain user types to jump around the site.
(Header / Nav / Content / Footer)


Header Tag - <h3></h3>

• I tend to apply bold to it as well.
• I try to make it "Meaningful" = PageTitle (alternatively, PageTitle + Section Title)
  : <h3>Brass Widgets</h3>
  : <h3>Brass Widgets - Widgets</h3>

Header Tag - <h4></h4> <h5></h5>

• I tend to apply bold or underline to it as well.
• I use these to indicate sub-sections of pages
  : <h4>Hand Crafted Brass Widgets</h3>
  : <h3>Machine Milled Brass Widgets</h3>

Links - <a></a>
Navigation

• Do my best to keep them as Text.
• keep them in Lists.
• Main/Parent/Section Head Links I try to make Bold.

General

• Always supply Title Attribute.
• Always try to seperate them from standard text (Bold + diff. Colour etc.).
• Try to include "in-page" Links to other pages within the copy (sort of intro's to sub/child pages)
  : <p>We also provide <a class="bold" href="#" title="details on our custom, hand crafted widgets">hand crafted widgets</a>for those unique, custom or lovingly crafted projects.</p>

Page Copy - <p></p> <img>etc.

• Try to make the 1st Paragraph a general sumamtion of the Page/Section.
• Try to get at least 3 keywords/phrases/terms in the 1st Paragraph.
• Try to get at least 1 keywords/phrases/terms within the 6 words in the 1st Sentence of the 1st Paragraph.
  : <p>The widget you need, the way you want it - it doesn't matter whether you need them by the hundred or jsut those few unique ones, Widgetting Engineering Ltd can make the Widget work!<p>
• Follow similar patterns for other Paragraphs.
• in-page Images should have Alt Attributes, with a keyword/phrase in.
• in-page Images should have File Names that match (rather than img1003dd.jpg).
• Emphasis bits of text - don't over do it... but apply a diff. colour + Bold (Italic if the font is big enough/readable).

Additionals

• Visitor Sitemap (try to follow the Navigation example!).
• XML Sitemap (Make the bots life easier ).
• robots.txt file (inc. location of XML Sitemap).
• link in the <head> section pointing to the Visitor Sitemap
• link in the <head> section pointing to the XML Sitemap

Well - I think that's Just about everything I do...

[AbleReach]

Wow, brave Autocrat jumps right in!

If the Name consists of multiple words, consider the primary domain using hyphens.

I'd tread lightly on that one, because it assumes that the usability problem of hyphens will trump the word stemming boost of hyphens. Hyphens are not a nice thing for word of mouth or anti typo friendliness.

Also - my opinion - if the domain is made up of words that Google recognizes and separates in searches, I'd be surprised if there is still a SEO advantage to the hyphens. You'll need to get in there play with some searches. As a semi related aside, I have seen Google maps be a little less aware of some words - one example is a title containing [businessname keyword servicename.] Normally I like to experiment with putting the least common word first in a title. In this case, until I put [servicename] before [keyword] the site was not showing up on Google maps under [servicename]. The keyword in question was a niche-specific adjective. Title tags still pack a hefty wallop in ye olde algorithmic sorting and filing system. I'm sure other things came into play in this case, too.

Some of the other things you mention worked better a few years ago, from a SEO perspective but are still a good practice overall.

Though headings aren't the big bang they used to be, from an accessibility perspective you're still in good shape by thinking of headings as an outline and using them to signpost user orientation within page structure. I never get tired of pointing out that spiders and screen readers are both text-aware browser devices that end users can use to "see" a page.

Edited by AbleReach, 04 October 2007 - 06:38 PM.

[bragadocchio]

When you guys start on a SEO strategy for a website what are the initial stages that you go through to ready the site.

Here's one thing that a lot of people don't quite get - there's strategy, and then there's tactics. I appreciate Autocrat's list of tactics, because they are helpful to lay out in front of others who might not know how to go about different parts of preparing a site for ranking well in search engines. A strategy or a set of strategies are the bigger picture, and they have broader implications than the tactics that you might use to achieve them.

What are the business objectives of the site owners?
What is the marketing plan that they've developed, and how does the online aspect of it fit into that plan?
Who are the different audiences that the site can appeal to, and provide goods or services or information for?
What are the best ways to get those audiences to come to the site through online tactics, including placement in search rankings?
What words will those people use to search for what the site offers, what tasks or actions can they take once on the site, how does the site engage them, and keep their interest?
How might conversions be defined for the site? Purchases? Leads generated? Newsletter or service subscriptions? Downloading of specific whitepapers or pages or software?

A copy of the business plan and marketing plan, interviews or discussions with the site owners and employees and others, analysis of the market, competitive analysis of other participants within the market, indentification of gaps within the market, and an understanding of relevant trends may provide some other ideas for a strategy.

Once you have created one or more strategies, then you can start deciding upon some specific tactics. Then you know whether you want to create articles for the site, or add a blog, or rearrange the site structure, or add tools or build widgets or make it more interactive, or focus upon some offerings more and others less, and so on.

[AbleReach]

Well said, Bill. I have always admired your gracious and firm way of infusing detail-oriented discussions with a practical and interesting frame of reference. That kind of perspective is one of the reasons I kept coming back to Cre8.

[A.N.Onym]

Sorry I can't contribute as well as Bill, but I'd say that you need to create the website for the people, not for Google.

A domain name should preferrably be:
- short
- memorable
- an action word (a verb, call to action)

It can also be a new, coined word, but don't have to be (if you can find the right unregistered domain name).


You need to link to other pages on your site using the words that are relevant to the page. If you have plenty of pages, then consider using internal linking more intensively.


One thing that I rarely see is a site that is easy to use. If you can make it so people can find what they want and do anything on your site easily, you've done half a job.

You can do that by using the words your people are looking for in links, page titles, pages, call to actions, etc.

[iamlost]

I wrote an answer and took a break and found Bill answering for me - intelligent and psychic But you get my 2-cents worth anyhow:

I have always looked first at the project 'strategy': creating the RFP/Q (Request for Proposal/Quotation). Companies who would automatically have one for an offline project seem to go 'blank' when it comes to the web.

Back when I had clients my very first question was:
"Do you have a formal RFP/RFQ (Request for Proposal/Quotation) for me to review or shall I develop one with you?"

My follow-up questions were:
* May I review your formal business model, i.e. revenue streams and cost centres?
* May I review your formal strategic plans, i.e. implementation of the business model?
* How do you visualise your webpresence implementing the model and plans?

The penny-pinchers and tire-kickers had now run out the door

Without a clear understanding of the company and its goals, its suppliers and customers, products and services, there is no foundation upon which to build. Without a clear vision of a domain's purpose neither the owner, the designer, the SEM, nor all the King's men can determine requirements. Without a reasonable knowledge of the business's current and potential customers and services you can not create reasons for them to come or to return.

There are what I call the ?standard! best practices for site accessability and usability, traffic acquisition, retention and conversion. Then there are the extras targeting your client's desired audience. Arty or techy, young or old, male or female, business or play, local or international, etc. in whatever probable combinations.

Small and medium companies often have a poor vision of what they do or what they expect to accomplish online. Until they are clear on those basics it is impractical to expect them to grasp the tactical requirements of a domain.

There are two other absolutely critical important reasons to start any website (including SEO) project strategically:
1. it brings the company's owners/management into the conversation. They become part of the solution rather than obstacles.
2. it defines the scope and boundaries of the project. The tactical milestones are easier to set, the contract easier to write, expectations easier to manage, and billing costs easier to swallow.

Oh, and writing RFPs can become another revenue stream.

[Black_Knight]

Bill and I sing proudly from the same songsheet. In the opening list of things you thought would matter, Saschaeh, you'd gone straight into the building stage without having the architectural drawing, or even a building plan.

I don't know if you've ever been directed to the Marketing 101 thread, but it is a very important read, so if you skipped it, don't miss it.

Then remember that many of the guys in that thread are SEOs. Marketing is vital. You must know what you are to market, what the market conditions are, and firmly understand how your market buys before you even think of how to code or promote a site.

Otherwise you build a site that may be doomed to total failure simply because it didn't understand the specific shopping process of the customers, or didn't know who the customers were at all. A site that tries to sell the wrong things to the wrong people at the wrong time.

You need to know who the market are, and how they buy. You need to know if this is an impulse purchase or a carefully considered decision that will involve recommendations from others, or have to convince multiple decision makers. You need to know how long the shopping process is, how much comparison shopping is done, and map out all of the touchpoints that a smart marketer could use to address these potential customers (or their advisors) and help sway them for the mutual benefit of both business and customer.

If you skip any of that, you may as well not bother with anything else, because you'll have already missed the money, and be relying on pure luck to have the site make a dime.

The fundamentals of marketing are that it is far more profitable to produce what you can sell, than to sell what you can produce. That goes for the website just as firmly as for the product. Make what will sell, rather than sell from what you make.

As for the initial list, perhaps it will surprise you how meaningless in SEO terms most of those things are (though they may mean a lot in Usability and Marketing terms beyond SEO).

> Build the site using CSS and preferable DIV based. - Largely Irrelevent
> W3C standard compliant - Completely Irrelevant
> Keep main navigation to the top and stay away from JS menus - Helpful
> Title tags relevant to page and other SEO semantics - Important, and often misunderstood
> Linking structure - Make things easy for Google to access. - Vital
> Create areas for dynamic content and try keep it up adding content. - Largely Irrelevant
> XML Site Map - Where this isn't irrelevant, you have other problems
> Robot.txt - this is the brake, it never acts as an accellerator
> Linking - (not so easy part) - Most Vital

edited to fix link

Edited by Black_Knight, 05 October 2007 - 01:04 AM.

[saschaeh]

Reply in true Cre8 spirit! Excellent Response! - (ready to buy a crea8 t-shirts and mouse pads! )

Well I guess I had tactical in mind, thank you very much autocrat ill be keeping that as a good reference. Admittedly I had not really given thorough thought to strategy.

On reflection perhaps it is more strategy that i'm needing to understand. When I place myself at the beginning of SEO i find it hard to visualise my path simply because i have not yet got to a worth while end result.

Like anything you dont understand it seems hugely daunting but as you understand bits an pieces it starts to whittle down. hmmm i just need some more teething!


What would be great is an example/case-study of SEO strategy document/process?

iamlost - RFP/RFQ i have not heard of this perhaps i do but call it something different?



Black_Knight - thank you! i post shortly after you i did not see your response ill have a look at Marketing 101!

Edited by saschaeh, 05 October 2007 - 01:11 AM.

[A.N.Onym]

Though not entirely a checklist or a strategy, "A Quick Kick Start Guide to SEO" has plenty of info.

Though, I should say, Marketing 101 is more vital, than the SEO stuff (which only comes when you get to work on the site). So I'd suggest reading Marketing 101 first, then working on your business model and marketing plan and only then read the SEO guide.

[Autocrat]

Brave? No dear maiden... that is not bravery... merely the gift of ignorance makes me impervious to fear


Well, as we seem to be using some martial terms...


I never suggest for strategy, as all battles are different, (thus the standard battle routines and unit formations for site based improvments => tactics).


I tend to apply the same business principals as for warfare though (as lets face, you are to conquor the market and win the wealth of visistors and their money ).


So...

Know they enemies - Thats right - you have multiples...
* The initial/obvious are the competitors. Look at the successfull ones, and then the less successfull ones - mark down the differences.
This will give you valuable intelligence on how they work and how they respond to a given situation.
* The less obvious yet potentially more fatal enemies are lack of resources, lack of food and lack of will (client based information, user based information and money).
Make sure that your army (site) will march with leadership, not march blind and won't starve before/during or after the battle!

Know where the battle is to take place...
* The "lay of the land" is highly important - the highground has many advantages, but so do swamps and rocky fields!
So, you need to know what market sector you are tacklingand how it has behaved in the past/recently.
This will give you a good idea as to what units to deploy and how best to use them (advertising methods, contacts, associations, link-building etc.).

Know when to make war - Timming is important...
* It doesn't all have to happen at the same time. Deployment of scouts and skirmishers can greatly improve chaces of a success, (using certain marketing techniques in advance can get more support and remove a few ofthe enemy).
* Knowing when to make the major push and when to regroup will be more efficient and save your army unnecessary losses, (know when to really push PPC and Link Adverts, and when not too!).



So, not particularly "Strategies", but standards and methods to conduct a more successful "campaign" in general


[edit]
Sorry - I couldn't resist.

Edited by Autocrat, 05 October 2007 - 05:15 AM.

[kestrel]

I can't offer any more on strategy advice, only that you should always start with the big picture.

But if you're wondering which order to go about your 'tactics' i would recommend leaving the hardcore code optimisation i.e. CSS design and W3C compliancy until later on in the project if it's a small budget, unless it's an absolute must.

It's a nice have and should be done but sometimes small clients budgets don't permit. As long as the code validates at the block level then I would leave it if budget is tight and focus on getting the keywords, content, internal link structure right, building high quality links , building trust and meshing the site into the topical community.

This will get you traffic much earlier on in the campaign and your client will see a return on their investment sooner.

In many cases, clients get a more than satisfactory ROI without a CSS design. I'm not saying it doesn't help or shouldn't be done, but you don't want to scare your clients off with a hefty fee either.

Consider waiting until you've made them some profit first before redesigning in CSS.

K

Edited by kestrel, 05 October 2007 - 05:30 AM.

[Autocrat]

That raises a question (:hijack:)...

Do people make a distinction between coding a site, coding it for validation, coding it for usability/accessibility, coding it for SE optimisation etc.?
(I am not incuding page content, merely the structure.)


To me, it's all part of the same thing... do others do it differently then?

[A.N.Onym]

If you take all the aspects of coding at once, then it is one thing. If there are numerous people working on the code or it takes iterations to implement the differences between validation, usability and accessibility (and SEO), then the things are different.

It depends on the developer and time at hand, I guess.

[kestrel]

I guess they are different and they are the same.

By that I mean, accessibility and SEO overlap, but you might not need to make the code fully accessibile to acheive your seo goals.

Active and over states of hyperlinks are required for accessibility purposes but don't impact on SEO for instance.

But it's likely you'll want to make use of header tags and the alt attribute from as early on as possible.

K

[Autocrat]

I jsut find it odd that these things are not all done at the same - surely it is more efficient time wise to implement such things from the outset?

I can see the point of different contributors making a difference, but I would have thought that it was common practice to validate as you do thigns, and apply what ever "standard structure" you may use from the ground up... going back over things to make the adjesutments would only requirem ore time... but could also result in much needed revisions of code if the original was not vlaid/structured as required?

Even my designers know what to do when creating a design... I spent a week going over things with them and explaining what is doable/feasable and how I work - so I get designs that I know I can replicate without too much fuss - another time saver.

So others don't necessarily do this from the offset... interesting.
Amazing how there can be such diversity (I tend to think that people go through the same steps... that will teach me )

[businessservicesuk]

Sas I don't have the 3 hours it would take to read all this thread, all I will add is you are well on track with your fundamentals in regards to SEO.

Try not to get too lost in the weeds at your stage. I am sure the advice here is well intended but you stated yourself you are new to this game.

Having a system on a website that enables the user to adds unique content is a good idea as is finding ways to generate high scoring inbound links.

Go with your own system, reading your post you will not go far wrong. My only other advice would be to avoid the use of hyphens in your domain. Incorporating a keyword is a good idea but not crucial, it will certainly not be the defining matter if yours or any site is successful.

[kestrel]

I agree it may take more time in the long run. But i'm talking about cases where you're working on a project where budget is limited, which is highly likely if you're starting out.

If the client was after keywords such as Computer Repairs Norwich (very low competition) and they had a 5 page HTML site, i wouldn't recommend the additional cost of rebuilding the entire site in CSS.

Which is likely to be the kind of keyword weight you will be working with if you are starting out.

Yes 'ideally' you would want to do it all in one go, but with small budget projects, where the keyword is usually low competition, you don't need to go to that extent to get the desired results.

Get the right keywords in the right places, wack some header tags in, optimise the internal link structure and build a few quality links.

This usually gets my small budget clients the desired results in a month or so, which means they are earning from the website sooner.

When they've generated some revenue and want to know what else they can do it is at this stage i'd probbly suggest a CSS redesign.

If they are small budget then it may well be the case that they are starting out too and need the sales, leads or enquiries sooner rather than later to keep them afloat.

Of course if they have all the money up front then do it all first.

K

Edited by kestrel, 05 October 2007 - 10:18 AM.

[businessservicesuk]

I love clean CSS based websites, we do very well in this area by providing clients good quality SEO compliant CSS sites with a full database functionality for around $1000.

Some people however prefer to pay $100 so they are limited