My name is James Cooper and I have been the affiliate coordinator at ukaffiliates.com for half a year now. After much feedback from our members I decided that a report was needed on how affiliates could increase their websites' effectiveness (especially concentrating on their click-through ratio 'C.T.R.'). Having studied the statistics from ukaffiliates.com and searched the web for relevant articles I finally came up with a definitive report on the 'effective use of links'.
The first thing to note is that a good C.T.R. is going to differ depending on the individual. Each website is trying to lure a specific class of visitors, for example a website about ancient Rome is targeting a unique section of the browsing public while a site targeting garbage disposal is after another section of the public. Would the Rome site want to attract waste disposal fans? No. Likewise would the waste disposal site want the Roman buffs? I doubt it. Try to remember that your C.T.R. is something you earn. You can help it by continually providing updated material online and working to improve your site. But what it all boils down to is a simple fact: your C.T.R. needs to be acceptable to your website and its target audience. 
As a web-master you should already understand the importance of good web design. Web design itself can be placed into a number of categories ranging from graphics to content but most relevant in terms of this report, would be the placing of links (be they banner or text).

I cannot underestimate the importance of appropriate link placement. From my experience choosing and pasting the correct links not only effects your impression count (i.e. number of times the link is viewed) but also your C.T.R. and of course the aesthetic look of your site. If you are someone who wants maximum impression count but minimum C.T.R. then by all means paste hundreds of 'in your face' flashing banners on your front page - but don't expect a cheque above £3.25 (unless you are extremely well established). Over the next few pages I shall go into detail about how to increase your revenue by efficiently using your links. Please note - there will always be exceptions to the rule.

· Text

Links can be split into two categories, text or banner links. Banner links are by far the most commonly used links in our network but that doesn't necessarily mean they are the most effective. Text links are the subtler of the two links. They are generally the same colour and not as eye catching as, for example, a 120 x 60 banner. Affiliates tend to use text links in online directories i.e. a garden page would consist of numerous lines of text links allowing the visitor to choose where they want to go. As the links are relevant to the page they are (in the context of ukaffiliates.com) certainly clicked on more often and normally result in sales - someone wouldn't browse through a DIY page if they were after a mobile phone. Therefore the statistics for affiliates who use text links tend to have a lower impression count but higher C.T.R., which theoretically means more commission.

· Banner

Banner links have far more variables to them than text links e.g. animation, size, download time etc. and each factor effects the performance in one way or another. When used in conjunction with each other they can be very eye-catching and therefore extremely rewarding. Many tests have been carried out over the past few years on what makes a banner successful (and lets face it - it's the only reason why people bother with them); the tests are by no means conclusive but the figures speak for themselves. 

· Size

I have to agree with the Renault advert when it says 'size does matter'. It has been found that banners, normally those around 468 x 60 (pixels) are more striking thus clicked on more than the smaller pixel banners. This doesn't necessarily mean that bigger is better as you can aesthetically fit more smaller 234 x 60 banners on a front page without making it look tacky and amateurish. 

· Animation

Animation is an important factor in gaining the attention of the viewer and a recent report suggests animation can increase response rates by up to twenty five percent! Care does need to be taken though when placing these banners; flashing colours and fancy pixels would be noticed, but it would be unwise to place these banners on pages where a user would dwell over an extended period of time (the animation can become more annoying than beneficial and also be quite garish). Animated banners can also be too effective - yes they draw attention but do you want it to be at the expense of your site! Also try to remember that simplicity sells - the graphics on the banners should enhance the message and not distract from it. Another point to remember when choosing animated banners is that animation increases the time in which it takes the banner to download. You have to consider that a good portion of the world still surfs at 28.8 or slower so they will never get a chance to see your slow-loading animated banners because they've clicked off to some other page before the banners have even appeared.

· Wording

The wording of banner links is fundamental to their success rate. There are certain words and phrases that can grab your visitors' attention and result in a click-through. If you are seriously looking to make money from your traffic then in effect your site has become a virtual supermarket. Supermarket advertising shares the same issues to banners, given that they only have a few seconds to grab their visitors' attention and then make them decide to act on impulse. The word 'FREE' is the most powerful word in advertising; who wants to pay for anything if they can get it 'on the house'. You are more likely to get a click-through when your visitor sees 'free' rather than 'discount', 'save' or 'sale', for these words are selected for a targeted audience who want to buy those products. When it comes to freebies what has anyone got to lose … absolutely nothing, in fact they have more to gain than anything else (which is why the freebie directories in the ukaffiliates.com network are the most successful). But there is a down side to the word 'FREE' - yes you get a higher C.T.R. but on the whole the click to sales (C.T.S.) is not as high. 
On the other hand if your banners, actually "recommend" the products, services and program(s) your C.T.R. might not be as high (due to the targeted audience) but the C.T.S. ratio will be higher - and when it comes to pay-per- sale it means more money. 
Phrases, for example, 'WITHOUT CHARGE' and 'LAST CHANCE' give a sense of urgency - and we all know that the average web-surfer does not want to miss out on any bargains. Another important factor in the wording is a call to action phrase/sentence, for example 'CLICK HERE TO SAVE' and 'BUY NOW, ENTER HERE'. These phrases are in fact instructing the visitor to interact with the banner, which in recent tests has proved to increase the C.T.R. by up to fifteen percent (the most effective call to action banners normally have a little flashing icon on which the visitor can point and click). Questions in banners have also been known to raise the C.T.R. by as much as seventeen percent, 'WANT TO SAVE HUNDREDS OF POUNDS?' and 'WHAT DO YOU PAY AT THE MOMENT?' are two good examples of interactive questions that get your visitors thinking.

· Positioning

An ambiguous factor (as no one accepts anyone else's views) that affects the banner C.T.R. is its positioning on the website. Conventional 'Net wisdom' is that banner adverts perform best the higher up a page they are placed. I recently found a definitive report on the web, which challenged this view. In fact the results showed a remarkable contradiction to convention. The banners at the bottom of the page had nearly three times the C.T.R. than the ones at the top. The reason for this was that most of the visitors to the site were searching for a specific piece of information (i.e. highly targeted = less casual surfers) and were therefore very unlikely to be distracted from their search by a banner advert at the top of a page. However, once they had finished reading about their chosen topic they were more open to clicking on a banner. Another report I found on the web (based on webreference.com's traffic) showed that banners placed next to the scroll bar on the right (in the lower right-hand corner of the first screen) had a two hundred and twenty-eight percent increase in click-through rate than ads at the top of the page!!! They also found that banners placed one-third the way down the page, as opposed to the top, generated a seventy seven percent higher click-through rate (all adverts being the same size and based on the same material). 
There is never going to be a right or wrong answer to where the most effective position is for adverts on your web page (mainly due to the huge diversity of the web-surfer). Tests are carried out all the time - examples of results I found showed that a non-animated 'click-here' was far more effective than an animated 'click-here', an underlined hyperlink pulled better than a non-underlined hyperlink, a thin boarder was far more effective than a thick, dark border, the list goes on. I feel it is probably best for you to experiment with the placing of your links than to go with convention (remember that aesthetics are important to). Remember simplicity works, if the banners are not bunched up into one heaving, flashing mass of adverts then you shouldn't go to wrong. 

· Speed

There are a few factors that both text and banner links share to make them more effective. Like e-mail all links should carry the merchant's signature and they should be fast in downloading. There is nothing more annoying than waiting for a banner to download (especially if it is on your front page). Using fast-loading banners is critical - remember if your visitors don't wait for them to load, they won't get a chance to click on them (thus defeating the point of having adverts). In a recent test by Ad Resource (Internet advertising promoters) they found that the C.T.R. went from one percent when using banners of over 10,000 bytes to nearly three percent when the banners were around 3,000 bytes - the banners were all IDENTICAL!! So the loading time of the banner is fundamental to its achievement (ideally the banner should be 6,000 bytes or less).

· Specification

It is also very good practice to try and link the banners to a specific page (for example a product page) rather than the homepage. In reality this is not always that easy but with a little patience and hard work it can be done (for the merchants obviously want sales as well). Nobody really wants to plough through a corporate home page to get to what they want to see … why not help them out. It could result in a sale.

· Refresh

Another good procedure is to refresh your banners on a regular basis. Yahoo in fact claims that banners burn out after two weeks of use and a recent study found that after the fourth impression, response rates dropped from 2.7 % to under 1%. As frequency of exposure increases, there are both incremental benefits and diminishing marginal returns. At ukaffiliates.com merchants are constantly changing their banners (for example festive occasions) giving their affiliates more choice to change the look of their site and/or promote a particular product. 

· Brand awareness

As an affiliate you also have to remember that brand awareness is extremely important to your C.T.R. A few months back I wrote a report on the success of various clients in our database. I wasn't surprised to find out was that the better-known merchants were given more prominent positions on our affiliates websites. This is pretty much self-explanatory; the average web surfer of the web would rather place their trust (i.e. bank details) in high street brand names than newly established internet businesses (even though the new internet companies might be just as effective as the blue-chip ones). Unfortunately this is a way of life and for many more years to come surfers are going to go to the 'well-knowns' rather than the 'unknowns'. One interesting fact is that web advertising actually outperforms television advertising in its ability to build brand awareness that favourably influences purchasing behaviour. So in theory it makes sense for the affiliates to place blue-chip merchant banners on their websites for they will inevitably get more clicks (but not necessarily more sales). 
What I also found was that if a merchant was a 'blue-chip' merchant the links the affiliates placed on their web page would, on the whole, not be relevant to the page and instead be placed in a noticeable position for maximum viewing (rather than usefulness). I'm sure many affiliates feel that by having a 'blue-chip' merchant on their page it gives them a prestigious lift above others but in effect the only lift these affiliates have is a greater impression count and in general a lower C.T.R. (thus less money coming in). In our network, the affiliates with the highest C.T.R. were those whose banners were specific to the page. If a visitor were to come along interested in gardening he wouldn't click on an electronic retailer's banner, he would go to a gardening page. To sum up, if you run an online directory/portal or information-rich site it is worth your while placing relevant subject links onto appropriate content pages (due to the targeted traffic) because they will result in clicks. It is not good practise to place links randomly around a web page as the chance that you might get a click becomes much slimmer.

CONCLUSION

To conclude I'll let Dotson define what a successful banner should include by quoting the six definitive rules he came up with which make an effective link: colour, animation, key words, simple, relevance and call to action. When all these elements are put together you can bet on a significant rise in your C.T.R. In terms of page layout, it is my view that the page should be kept simple. I know it is tempting but try not to bamboozle your visitors with hundreds of adverts; they aren't stupid, they know you are trying to make money out of them and remember 'Once your visitor has clicked-through it's no longer up to you'.

James Cooper

UK Affiliates

Sources:

T. Stubbings
Webreference.com
Bannertips.com
And numerous others who I don't have time to mention