The Most Critical Step For Any Business
You have probably heard this a thousands times. “Test, Test, Test Everything And Assume Nothing”
If you are not testing and tracking everything you do.
I can absolutely guarantee you that you are missing opportunities and you are throwing money away.
You must know the statistics of your online business, if you do not include testing as one of your primary strategies all you will be doing is trying to operate a business on your gut feelings and pure luck, not the most sensible way to build any business.
So what does testing involve?
You need to break down and separate the elements of your website including your sales process in order to conduct useful and profitable testing. This way you can track the results, modify them and improve on them.
There are four things you must remember when testing.
First, only test one thing at a time. Let’s say you are testing the headline of an ad you’re placing. You only want to change the headline and not the headline and maybe a word or two in the ad itself.
Because if your ad improves you will have no idea what caused the improvement. The headline change, the few words, or even just one of the words you changed.
Second, You need a control your tests can be measured against. So your control would be the last best performing ad, sales letter, email to your list or what every else you are testing. So if your next ad does better then your last ad, this new ad is now your control ad.
Third, Try to keep your testing within the same environment. What do I mean?
If you are testing headlines for an ad in April you should not be testing another headline in December. The seasonal change will affect your results. Keep that in mind for email testing; make sure you send your test to the same number of people and in the same month to get a more accurate idea of your test.
And fourth. Keep a record of your results. You would be surprised how many people do not keep accurate records. Also you may come across an old test and decide to give it another try in a different way and it could end up performing like crazy.
I was taught three different ways of testing and they are:
1) Linear Test
2) Spilt Testing
3) Multivariable Testing
First linear testing. This is the most basic method of testing. It will show you your conversion rate for your site. Lets say you are testing your web sites headline. Run that headline for a week or even two weeks and count the number visitors you get compared the sales you make.
Now divide the number of sales by the number of visitors then multiply that by 100 to get your conversion rate.
So if you have 100 visitors and you get 10 sales, you have a conversion rate of 10%. Next week change your headline and then track those results doing the exact same thing.
When doing this keep in mind of any other things you may have going on during this testing. If you have an ad going into an ezine during your second week of testing this could skew your results by driving more targeted visitors to your site. So the traffic could be boosting your conversion rate not your headline.
Second split testing. Split testing is great for people who do not have a lot of traffic coming to their sites and it also cuts your testing time in half.
Split testing also eliminates most external factors that could influence your results. Things like holidays and world events.
So how does it work? Basically what you do is run two different versions of the same web page on the same URL at exactly the same time. How do you do that?
So you want to test your headline of your web site. All you do is create the same web page, with the different headlines. The traffic is automatically rotated between the two pages with some code and those pages are flipped back and forth on the same URL. This is so cool.
Now you don’t have to be a programmer to do this. This process can completely automated with some very cool software. Vertster.com allows you to do split testing.
You can also use Google Website Optimizer; this is a free tool within Google Ad Words it allows you to do split testing.
Finally there is multi-variable testing. This testing method is great for split testing your whole sales process.. If your site gets a good deal of traffic and by good I mean at least 5000 visitors per month or it will take to long for you to get any useful information.
So if your split testing your headlines you can also test two different price points and you can now test the combination of your headline with the different price points and see which one converts better (sales).
As I mentioned earlier you could also use Google Website Optimizer, it allows you to also to multi-variable testing.
So what should you be testing? EVERYTHING!!
Your offer, your headline, your price, your layout of your site, your ordering process, the copy, the subject, the format, different locations of your opt-in offer, and so on.
I know testing is not that fun, but the people that spend the time testing and tweaking their websites are the ones who will profit the most.
I believe in you!