Business & Finance Advertising & sales & Marketing

Statistical Significance Brings Confidence to Your Business Decisions

Every business owner sometimes has to come up with serious "either-or" decisions about their website or online promotion. Smart owners conduct some sort of tests or surveys to make an informed choice. But how much test data do you need to draw reliable conclusions? How much resources do you need to allocate for your tests, and how long should you run them? That's where statistical significance can help you.

Statistical significance is a mathematical concept that allows you to determine if your experiment results reveal some underlying tendency, or if they are within the normal random distribution range. It is never truly possible to know for sure, but there are formulas that can find the probability that your experimental result actually means something. A 95% probability is usually considered good enough to base business decisions on the test results.

This concept can let you pick the best headlines, ad text, sales letters, sign-up and order forms for your online business much faster and with more confidence. This can be a major factor in making informed decisions and increasing your profits every month.

You may think that you can make conclusions intuitively, by just looking at the results. Is it worth the effort of calculating something when you can simply run the test for a month, and just choose the page with a higher conversion rate, for example?

But think about it for a moment. What if your results are just a random deviation? What if your test is actually inconclusive? If you base your decision on it, you can lose a lot of potential profit. And maybe with just two more weeks of testing you would get a scientifically valid result, and rest assured that your business is driven by solid science rather than guesswork.

Another possibility is that you could have reliable results in half the time, if only you could know for sure. You could spend half as much on your testing, receive the result much earlier, and perform twice as many different tests and optimizations on your website every year, each one potentially increasing your profit.

There are complex formulas and tedious calculations involved in finding the statistical significance. Luckily, this is a well-studied problem, and there are numerous tools that can help you. They can be either software applications that you install on your computer, or online calculators that you can easily use directly in your web browser.


Leave a reply