There are increasingly a range of acronyms being thrown around in web marketing that sometimes slip out when interacting with clients. Here’s a brief overview of some of the more common e-marketing acronyms to ease the conversation.
- SEM
Search Engine Marketing
SEM is a broad term covering the range of activities you might engage in to improve the extent to which you generate revenue via search engines
- SEO
Search Engine Optimisation
SEO focuses on the organic listings as opposed to sponsored listings, or PPC. The position of your site in the listings for a given search term is determined by a wide range of factors, each contributing to the search engine’s algorithm. They can be broadly divided, however, into two groups:
- Content Relevancy: The greater the extent to which the content of a page matches the search term in question, the better it will rank (subject to the engine’s spam protection measures).
- Trust relevancy The greater the extent to which a page is connected (by hyperlinks) to sites also well connected to other sites relevant to the search term, the better it will rank, based on it being a trusted resource.
- PPC
Pay Per Click
Many search engines present two listings when the user makes a search - the main, organic search results, and the sponsored listings. To achieve a sponsored listing, you must agree to pay a certain amount to the search engine every time someone clicks on your listing. Your bid compared to that of your competitors determines your position in the listing - bid prices can range from a few pence for less competitive phrases to several pounds for very competitive phrases. Most often however clicks will cost less than £0.50.
- SMO / SMM
Social Media Optimisation / Social Media Marketing
More recently, a new area has opened up in web marketing with the arrival of social networking and in particular social news websites, such as Digg, Reddit, and a host of others. These sites allow you to submit content to their ‘Upcoming’ listing, presented initially in reverse date/time order. As time passes, if your site receives votes, it maintains it’s high position and may reach the ‘popular’ listing. This is encouraging marketers to be creative and generate interesting content, as the potential returns of getting ‘dugg’ are substantial in the short and long term.
Following our post last year on Google UK’s market share in 2007, here are some figures providing an indication of Google’s Market share in the UK in 2008. Hitwise are providing the numbers again, and in a relatively short sampling period during March 2008, they found Google to be holding a market share 87.5%, 10 percentage points greater than the same figures for 2007. They found 73.7% to be using google.co.uk, and 13.8% google.com. Clearly Yahoo, Microsoft and Ask among others have therefore lost ground to their primary competitor.
Particularly interesting in the same report was the statistics regarding the proportion of users opting to use Google’s ‘Pages from the UK’ search option. The report confirms intuition by showing only 13.6% of searchers to make use of this facility. This will help you decide the relative value of a .co.uk TLD vs. a .com or alternative (a google.co.uk search for ‘Pages from the UK’ will only return pages with a .co.uk TLD).
For more details, click through hitwise.
I wrote a post before the weekend about some performance issues with MooTools, DHTML and AJAX, but having found the solution to a related problem this morning, I wanted to post a followup.
I’ve been working with a large, dynamically generated and AJAX populated table that features sorting, filtering and row highlighting, and was finding that my table took a long time to render, and was also very slow to close - that is, when I tried to close the window or refresh the page, there would be a delay when the browser would first hang for a few seconds. IE often (but not always) popped up the error/prompt Stop this script running?, and Firefox less often threw it’s Unresponsive Script warning.
I’d heard of MooTools’ garbage collection functions that reduced memory leakage, and figured that these were probably responsible. I then found Kevin Smith’s write up of his similar experience, and came to understand the problem. MooTools’ garbage collection takes time to clean up any element that has been extended. My table was executing the following code, effectively extending every table row in the table, resulting in some 500+ extended elements.
// MooTools Code
Element.extend({
…
getChildren: function(){
return $$(this.childNodes);
},
…
});
…
// My Code
this.tablerows = this.body.getChildren();
When it came to cleanup, this javascript processing took longer than IE’s configured script timeout, and thus prompted the warning.
To avoid this cleanup overhead, you’ll have to avoid extending the elements with MooTools’ extensive set of functions and instead extend the specific elements in question with the specific functions required.
For the life of me, I couldn’t figure out why I didn’t get different results, using SQL Server 2005, for a simple Full-Text Search query and a Full-Text Search query using the FORMSOF and INFLECTIONAL keywords.
What I stumbled across eventually was that when creating and configuring my SQL Full-Text Indexes, I had left the language parameter as the default selection. When I changed this to English, voila!, my FORMSOF (INFLECTIONAL, @query) queries started to produces fantastic search results!
Also, an example query showing how to parameterize the the search term, and how to rank your results:
@query
varchar (100)
SET @query = ‘FORMSOF (INFLECTIONAL, ‘ + @query + ‘)’
SELECT FT.rank, P.id, P.title
FROM Products P
INNER JOIN CONTAINSTABLE (Products, *, @query) AS FT ON P.id = FT.[key]
If you’ve any questions, I’ll do my best to help. Also, if you’re just starting with Full-Text, check out my posts on installing Full-Text Search on SQL Server 2005 and how to setup and configure SQL Server 2005 Express Full-Text Search the easy way.
I recently had to setup a search facility on a website driven by SQL Server 2005 Express. My natural choice was SQL Server’s Full-Text Search feature, but setup was not entirely intuitive, and the web was full of misinformation. I’ve put together a tutorial to save you time.
If you’re not sure if your have Full-Text installed, read my post on installing SQL Server 2005 Express Full-Text Search.
Once, Full-Text is installed, managing your index is actually very simple:
- Browse to the tables of your database, right click and select ‘Design’
- In the main window, right click and select ‘Fulltext Index…’
If you see the message, ‘You need to define one or more full-text compatible indexes on the table to create a full-text index’ and the ‘Add’ button is greyed out, it’s because you don’t have a column the database can use to uniquely identify each row. Make sure you have a primary key, and if you’ve no success, start your troubleshooting with this video - it discusses unique indexes about 1/3 of the way through.
- Otherwise, click ‘Add’, and you have a new index.
- To configure the columns in the index, click the ‘…’ button next to ‘Columns’
- For each column, select the column, and select the appropriate language
To understand why the language parameter is important, see my post on use of the FORMSOF and INFLECTIONAL keywords in SQL Full-Text Search
It’s that easy! Any questions, let me know and I’ll do my best to help.
