Optimising a Web site

Tue, 20 December, 2005 – 1:43 am

Before I start I'm not intending to offend anyone, creators of software, people who have bought or set up AIS sites and had no luck etc. Nor am I saying that what I put below will work for everyone. I just decided that a couple of my AIS sites have proved worthwhile and I know a few others out there haven't had as much luck or else their initial luck died quickly, so I figured I'd just let you know what I did. Perhaps my tips and tricks may prove of some use to someone else

History:
I went to Burt back in October to see if he'd "build" me a web site using his Niche Portal Builder software. I had a domain that had laid dormant with a useless site for a few years so I figured I'd give it a go and see what would happen. I paid £20 for an almost full XHTML/CSS web site so my job wasn't too big to make it strict XHTML, accessible (as much as I could) and validate. I bought the site on a Saturday early afternoon and spent in total maybe 4-5 hours cleaning it up and then put it online.

What have I done?

The initial site was a nice simple design that didn't give the impression of your typical AIS/keyword adsense site. Burt's templates do not have the run of the mill look, which is good because I think the visual look of a web site gives people a better impression of it. The majority of the site was already set up in strict XHTML and CSS however not all. I'm guessing the template was but the bits inserted by the NPB software is out of the user's control. So I set to, cleaned up all of the code on the front page and then used that as a template. I stripped out all generic code and put it into PHP file includes to make life easier for updating late down the line. I shuffled the navigation around to get the more important links to the left side of it as well. I then set up file includes for all of the Google Ad types I wanted to use. I positioned a couple of 2 ad boxes on the front page, a horizontal one just below the title and above the page text and then a vertical one in the side bar. I then rewrote the page title of my front page and the main H1 header title.

I applied the changes to the layout throughout the web site which then just left me with the unique text and titles for each page. Starting with the non link pages I rewrote the page titles on every page and incorporated a different key phrase into the title and header tag. I also cleaned up any code, mainly where two line break tags (br) were used to create paragraphs, and used proper paragraph tags instead. I also removed my email address from hidden form inputs on every page – didn't fancy the spam that would come with that! I just hardcoded my email address directly into the formmail script instead.

I uploaded the site then as it was. For the first few days it made a few cents but nothing major. Google indexed about 5 of the new pages but that was all, and just refused to look any further. A couple of weeks later I decided to make some slight changes. I ordered the links directory into a few categories, stripped out some of the duplicate type pages or pages that sounded a bit daft to me, put the links directory into a list (previously just a paragraph with line breaks) and then created a Site Map. The Site Map is set up as an unordered list and linked to from my footer. I then linked to the sitemap from this blog and let googlebot have a nose. Within 24 hours googlebot had gone mad according to my stats, 2 days later my visitor count increased 3 fold and when the visitors went up, so did the adsense.

So for a couple of weeks I was on cloud nine! I was suddenly making between $2 and $6 a day for a web site which cost £20 (almost $35). Less than 2 weeks was all that was needed to cover the initial cost. However I wasn't happy (never am!). I was making less than 10 cents a click on average. On my own blog I could make over $1 a click, so if I got that onto the Swish site, well I could be pulling in 10 times as much. So I started to research and learnt a lot very quickly. Certain sites I was reading kept mentioning the 'footprint' of these web sites. Now whilst I knew that the main code had been altered extensively, the way the links were and the way the whole directory structure was didn't leave me feeling confident. I also logged into a client's adwords account (perks of the job!) and took a look at the top bids on related key phrases using the keyword analyser tool. By choosing phrases associated with my web site and putting in a high daily limit I can see how much it would cost for me to get the top spot on Google Adwords. Unfortunately the results were not great. The highest cost on a Google Ad was just over £1, so roughly $1.70. Of course we don't know how much Google pays out to the publisher from what it makes, whether it's 5% or 10% or more, but this gave me a rough idea of what I was working for. I admit at that point I did wonder if it was worth it, however I figured I'd use it as an experiment if nothing else.

On the selling side of things, I'm not so fussed on selling advertiser space as all I've had from the site is just spam anyhow. I'd already set up a shop page, selling templates from Template Monster, and the only book available on the subject from Amazon. I decided to push a couple of products more on the web site as I knew they would make more money than the adsense was. I put the Swish Bible book and Flash software (as you have to be a reseller to sell Swish software which I don't have the current inclination to do) on the side bar of every page in place of the newsletter sign up. By doing this I've managed to make 2 sales of the book, however I believe I can (and will) make more affiliating with Amazon US instead of UK.

Where am I now…
So I've started the job of moving the various pages out of their own directories and into the root directory. Each link page is getting (or has had) a short paragraph to target the key phrase for that page, thus hopefully boosting the worthiness of the page as far as google is concerned, they in turn will charge an advertiser more for a click. I'm also going through the links on each page, removing any that are out of place as some crept in, searching myself on Yahoo to see what else turns up, and also rewriting some of the link descriptions. The links are listed in paragraphs so I've changed this and am moving them into unordered lists, and also cheekily styling my lists to appear like Google's Adsense list (ie. the same spacing between each link/description). Each page takes maybe 5 minutes however I've seen my average per click almost double since starting this. Whilst my clicks have slowed down my total income each day (on average) has risen.

I'm making these changes slowly for a couple of reasons. The main one being I don't actually have the time to sit and do them all in one go. It's easier to do one or two at a time anyhow as I'm not the world's best copy writer. Whilst I do have distractions from my work, fixing pages doesn't spring to mind as a distraction nor does it always come to mind when I have a spare 5 minutes. The second other main reason however is because I want to keep Googlebot interested. By not making too many changes too quickly, and just the odd one here and there Google will think that the site is constantly having updates made to it and it will hopefully keep the rankings up. Since starting to break down the directory structure and edit the link pages my daily visitors, which went from about 25 a day to an initial peak of 140 and then back down to about 70 is now on an average of 80-90 a day, so I think my editing has given me a little more credibility with Google, rather than a new web site laying dormant.

For the Future:
I've still got a lot of link pages to sort out. Perhaps 30 more or so. I've also just joined up with Amazon US so I'll be changing the affiliate links to the US ones, hoping to make more sales as the bulk of my visitors are from the USA. I also want to add in other template options as i've signed up to a couple of other template web sites as well. However I'm happy with the web site as it is. There's only so much it can earn in Adsense, and there are only so many visitors that will come to the web site. I would like to set up a section to have sites set up in Swish for people to just see what can be done, to gain inspiration etc from them. This is what the web site was initially set up for back in 2001, however I stopped using the software and subsequently lost interest in it. Whilst I no longer use the software and cannot see myself taking it up again as I'm not a designer nor animator, I still enjoy seeing what can be done with this little piece of software, especially when you compare the pricetag to Macromedia Flash.

To date Swish World has made $101.56 in adsense earnings (today isn't over yet either!) plus £1.40 in Amazon fees.

• Bookmark this page at:
• Del.icio.us -
• StumbleUpon -
• Spurl