If you would like to become a successful affiliate marketer, you will need to understand the importance of testing your campaigns. Many people give up after their first unsuccessful promotion, or try something totally different, but making money takes time and experience. Instead, the initial failure should be looked at as a learning experience, and the next campaign based on a tweaked version of the original. To do this properly, it is necessary to thoroughly test all your campaigns with (A/B) split testing or multivariate testing.
Setting up campaigns
When first setting up a promotion, you will need your own landing page for the product. It is not wise to send people to the merchant’s sales page right away; you need to pre-sell the benefits of the product. Tell people the answer to their question: “what is in it for me?” You can write a review, which is a common method, or just set up a teaser page (a short review with a link out for further info).
For long-term benefits from campaigns, advanced marketers often first create a squeeze page with a PLR product or something else they give away and then redirect the visitor to a related ClickBank promotion after the person has downloaded their gift. These emails can later be used in email marketing to sell a number of products.
No matter what you do, you have to measure conversions, the number of visitors, the source of traffic, the clicks through the sales page and the earnings per click.
With split testing, two variations of the same page landing page (a page focused on selling a single product) are created and run to see which one is the most effective. Creating landing pages covers the basics of what a good page needs, but it is impossible to know what exactly will be best for any particular niche. Some of the different variants to try when split testing could include:
- Trying various graphics or positioning them on different parts of the page
- Eliminating parts of a form to see if doing so increases conversions
- Colors to include background and text
- Changing the very important headline, opening paragraph (as well as other content) and call to action
- Changing or repositioning “buy now” or shopping cart buttons.
- Logo design
- Price (if at all possible)
Basically, if changing something on a page could potentially get visitors to stay longer and interact more, it is possible to test it out and see how it works. It is often a good idea to try radically varying approaches as small changes may not make enough difference to be noticeable. Web users increasingly expect more, and it is important to be creative and put a unique flavor on what you do. Keep in mind that even modest increases in conversion rates can make a huge financial difference over time.
Split testing and tools to use for analyzing campaigns
Split testing is necessary to find the right campaign strategy. To do so, first pick different elements that you want to test. Create variations for headline, graphics, call to action, even buttons and graphics. Then, run these variations to see which bring the best results.
Google Analytics and Google Optimizer are free and some of the best tools for split testing. To use them, you simply need to create a Google Webmaster account and install a script or add a code to the page the promotion will be featured on. There are videos and the step-by-step tutorials with Google tools if needed. Users can test the different variations of the page and use the tools to see how visitors respond. In the end, you will get a report of the conversion rate of the alternate versions.
More coverage of Google as well as other programs for analyzing site traffic and ads can be found at traffic testing tools.
For faster testing, many advanced marketers use pay per click traffic to get people to offers and see which one does best. This can save you time, increase your earnings and teach you a lot about your customers at the same time.
Split testing tips
The data from split testing can get confusing if it is not tracked in a standardized way. Record results on a spreadsheet at least one set day of every month. Otherwise, results may be lost, misread or needlessly duplicated.
- Split test is not a one-shot deal. It must be carried out continuously.
- Always look at the numbers and do not go by feel. Many people settle on a particular page just because they feel it is doing better. Quite often, these feelings and memories are not accurate.
- Be bold by making major changes and seeing what happens.
- An example of some tests can be found here.
Picking and promoting a winner
It may take hundreds of visitors to find out which version of the landing page is the most effective, but eventually the one that earns the most needs to be selected. At that time, look closely at the sources of traffic and pick the top promotion methods. For example, how does article marketing do against video? After you have done this, you have to keep an eye on the refund rates for each traffic source, as it can be much higher for cold traffic than organic or blog visitors.
Once the testing is done, it is often a smart move to increase the amount of paid (PPC) traffic to offers that have good conversion rates and earnings per click. The same principles used in split testing can be put to work in PPC to find the best ads. At the same time, utilize free methods, such as bulk traffic (less targeted purchased traffic) or organic search visitors brought in through on-page and off page SEO. It depends on your budget, preferences and how much organic traffic the site gets.