How to optimize images for SEO

Quite basic tips I know but to get maximum results for image search it does require a strict process which must be maintained to achieve best results.

Image location

• Don’t save your images in an external location (like an image hosting company)

• Don’t save your images in your root directory (i.e. www.domain.com/photo1.jpg)

• If you have a small to medium sized website I would recommend keeping you images in a separate folder named “images” i.e. www.domain.com/images/photo1.jpg

• If you run or maintain a large website or an e-commerce website then managing images should be more categorised, for example:

www.domain.com/images/cars/blue-car.jpg

Image naming

• Quite obvious but make the image either gif, Jpeg or Png  format

• Use targeted keywords that give accurate representations of what’s in the image

• Using your targeted keywords as image descriptions may up your on page keyword density but this does not portray an accurate representation of what your image is.

• Common image naming structure:

http://www.domain.com/images/keyword1-keyword2.jpg

• Bad image naming structures:

http://www.domain.com/images/photo1.jpg
http://www.domain.com/images/cars/catid29876.jpg

Image placement

I have created a mock to show how I would choose my image placements for a page trying to optimise for Bill Gates + Microsoft. Based on experience image optimisation only really helps when you make the images as relevant as possible to the content you are writing about.

Image Tag

An example of a good image tag (Alt tag) would be:

• <img src=”images/bill-gates.jpg” width=”100” height=”100” alt=”Bill Gates”>

An example of a bad image tag would be:

• <img src=”images/photo1.jpg” alt=”Microsoft’s Bill Gates Microsoft”>

Like I said this is quite basic stuff but think it’s important to be reminded of it at times as I for one have been lazy at times and have not always followed this out.

Benchmarking is a necessary action when working on any aspect of SEO for a website. For those of you who don’t know what benchmarking is – Benchmarking is an action whereby you record specific metrics of success over periods of time. For example one of these metrics could be search engine rankings.

Without benchmarking how would you determine the success of a strategy? You need to capture results for whatever metric to be able to show how your SEO initiatives have affected your clients website.

Some examples of SEO metrics I use to assist with reporting and monitoring performance are:

1.    Keyword Rankings
2.    Search Engine Traffic
3.    Referral Traffic
4.    Direct Traffic
5.    Cache Dates
6.    Index Percentage Penetration
7.    Time on Site
8.    Page Views
9.    Time of Visits
10.    Day of Visit

How to perform Benchmarking?

The only accurate method of benchmarking in my eyes would be manually. Ok, this may be time consuming on resources but accuracy matters. Obviously some of the metrics I have discussed above can be captured effectively by Google Analytics which has the awesome price tag of free. However when tracking keyword positions over major search engines there are alternatives to doing the work yourself.

I have used the following three pieces of software, all with positives and negatives but in case you are unfamiliar with them I’ll give you a short piece about each.

Authority Labs

These guys are the new kid on the block if you will – I originally picked up on this resource by Raven SEO’s blog. It’s a web based system which allows you to add domains with each of the keywords you want to benchmark for very easily. The software will even provide you with recommendations of keywords if your own keyword strategy is not available. Then every 24 hours data is updated, and will show you any increases and decreases over Google, Yahoo and MSN. A great little feature which I love is the graphical representation of keyword progress over time, you can see very easily influences that you might be implementing taking effect. They offer a free tester with limited keyword tracking and they have a tonne of features in the pipeline so I suggest you guys check it out.

Rank Checker

Rank Checker is a Firefox extension free to download by SEO Book. This tool is great for estimates – you can add as many domains and keywords as you like and you can check over a number of different territories for Google i.e. you can see differences between Google.co.uk and Google.com. Depending on the number of keywords this tool can take a while to give results but when the checking is finished you can download a CVS format of the results which will display all data each time you have benchmarked. Like I said this tool is great for estimates, I find the ranking positions to be quite inaccurate at times, so I wouldn’t recommend using the data for reports – but for monitoring this is great!

Web Position Gold

These guys have been around for a long time and it shows with the level of detail that can be obtained by this bit of kit. It’s the most expensive (although you can take advantage of their free trial). You can enter virtually unlimited domains and keywords and get pre designed reports at a click of a button showing all kinds of keyword activity and recommendations. It’s a bit tricky to get started but once you get a complete understanding of how it works, this is a necessity for any SEO.

People know (or at least they should do) that implementing a number of SEO techniques and methods on any given page can influence the search rankings in a positive way. There are plenty of resources to help explain how you can create the “perfect page” in regards to SEO but are there any clear metrics for success? What can you expect if you change or alter page content, or perhaps the Meta data?

One thing we do know is that if a good SEO gets their hands on your website or specific page you will see positive results. What I’ve been doing is benchmarking when the changes take place in Google and whether the changes are positive and negative.

The Google update test

I optimised around 50 pages of a website I own that I initially setup around 3 months ago, the Meta data, page tags and content was not optimised at all. I created a strategy to optimise these pages, the actual content of these products were products i.e. one product per page. I changed the following:

1.    Optimised the meta data
2.    Included keywords and alternative keyword phrases on page
3.    Optimised the images on the page

I had read somewhere that updating large numbers of pages on a website all at once could lead to a possible penalty, although I have never seen this I thought this test would help determine this theory.

I benchmarked data over a six week period on Google, based on individual pages and their targeted keywords, which had been optimised.

Week 1
Around 80% of the pages actually increased rankings in the first week with around 15% remaining the same and only 5% dropping rank

Week 2
In the second week there were some more keyword increases and very few positions dropped – a good week all round.

Week 3
In the third week it was the complete opposite, just over 85% of the keywords dropped below their original ranking with 5% remaining the same and 10% increasing

Week 4
Huge increase of positions, now around 70% of the pages I originally optimised are ranking well above their previous position with many on page 1 or 2. Very few position drops from original positions but there were some.

Week 5
Not much movement between keyword positions but 30% of keywords have improved from week 4, 80% remain the same with around 10% dropping slightly.

Week 6
Final week and only one page has increased from week 5 while 2 pages dropped slightly, the rest remained the same.

Time for some graphs:

The first graph is has been taken as an average from over 40 optimised pages over the period of 6 weeks so visually you can see the update life cycle.

The next graph shows five randomly selected keyword behaviours over the 6 week period

The final graph shows another 10 randomly selected keywords and their position changes

Google’s Page Update Life Cycle

Yep, think that’s what I’m going to call it! Anyway I’m aware that this lifecycle of position changes probably goes on for a bit longer but the data over the 6 week period was the most active. This is something that I’ve seen many times before but have never benchmarked for such a test. It’s also worth mentioning that you do see position changes before Google has re-indexed the optimised page.

So if you go about updating pages of your site don’t worry if they go all over the place for the first month or so, if they have been optimised correctly then you should see some kind of improvement.