So, you’re not testing.
And, I think I know why. If you aren’t testing, I am going to go out on a limb and say that it is because of one of the three barriers below.
Barrier 1: Traffic
This is a common problem. So common that I will spend the majority of this article discussing it.
It doesn’t take a statistics professor to tell you that it will take light years to get results when you have only a trickle of traffic visiting your landing page.
And, as it turns out, all traffic is not created equal. Here are just a few of the traffic variables that can affect the confidence level of a landing page test:
- Source – If you want truly unbiased data, you should avoid using traffic sources like your internal e-mail list, as they have prior experiences with your company. Also, the stability of traffic sources from tactics like SEO can be affected by algorithm changes, etc. Often, Pay Per Click is the preferred landing page testing traffic source.
- Seasonality – If you are selling sexy summer tees, the traffic you are getting during these hot summer months will react differently than they will in the winter months.
- Geographic Location – Site visitors living in warm locales will react much differently to your summer tee offer than visitors in frigid locations.
If you bust out your statistics text book from college you will find information about z-scores, t-tests and standard deviations. Standard bell-curves and confidence scores abound.
This next statement might draw fire some dyed in the wool conversion optimization specialists…
Testing doesn’t always have to be that analytical. The question is, what are consequences of being wrong.
Let’s take at two real-world landing page testing examples:
Landing Page Test 1:
Product: Paid webinar
Time Investment: Low
Monetary Investment: Low
Test: Two price points will be tested — one at $30 and one at $50
Risk of Being Wrong: Low
Traffic Sources: Email list
In this first example, the risk of the test leading to the wrong conclusion is low. Although traffic from your own email list is largely considered in the testing community as a biased sample, who cares?
You are simply trying to maximize the amount of ROI on your paid webinar.
In testing cases such as this first example, a large amount of the concern about unbiased samples, traffic sources, etc can be ignored. Certainly, it is better to apply best statistical practices in all cases but as a practical matter, it isn’t always possible.
Now, let’s look at another example.
Landing Page Test 2:
Product: New line of women’s clothing accessories
Time Investment: Very High
Monetary Investment: Very High
Test: A/B test of call-to-action button
Risk of Being Wrong: High
Traffic Sources: PPC
In this second example, the risk of making the wrong conclusion from the testing data is very high. Time to bust out the formulas and spreadsheets and get serious. Or, hire a conversion optimization professional.
A steady, high volume stream of unbiased traffic will be required to get the data necessary to make decisions. Therefore, PPC might be chosen as the traffic source as it provides an unending source of traffic, as long as you are willing to pay.
The bottom line on traffic is that it is important. You want a high volume, steady stream of unbiased traffic in all cases.
But if a lack of the perfect sources of traffic is stopping you from testing, consider weighing the risk of being wrong.
You just might decide that the risk is low enough to warrant testing a smaller sample of biased individuals.
Barrier 2: Inaccurate or lack of tracking
The beauty of online marketing is that it is measurable.
Website analytics programs can provide you with an astounding amount of data. That being said, it is remarkable how many website owners do not take advantage of this data.
Free programs like Google Analytics and Google Website Optimizer are really all you need to start making wise decisions about your landing pages.
There really is no barrier left here, get these free tools and off you go.
Barrier 3: Difficult or impossible to add/edit content
I saved the most common barrier for last (I think that is what my English teacher said to do.)
If you are going to test landing pages, you have to be able to build and edit those pages. And, as a practical matter, you have to be able to do it quickly.
The name of the game here is — Content Management System (CMS.) And, if you want me to get specific, use WordPress.
A good CMS will allow you to add pieces of tracking code (like your Google Analytics code from Barrier 2) with ease. With a good CMS (such as WordPress) you will also be able to add and edit landing pages easily.
If you have to call your “web guy” every time you need to add or edit a landing page, it won’t get done.
Migrate to a solid CMS system like WordPress and this problem is solved.
If you aren’t testing and measuring, you aren’t taking full advantage of the strength of Internet marketing. My guess is that one or more of these three barriers is standing in your way.
But these barriers are easily overcome.
How about you? What barriers have you seen that prohibit landing page testing?