Summary: Learn how to use testing and web data to maximize your profits. Incrementally improve your reach and success rate by setting up small experiments and then evaluating them to learn what works and what doesn’t. With each new experiment you will be gaining valuable insight and compounding upon previous returns.
If you were the manager of a sales team, would you train your team so that your business can earn more revenue, or would you just let them work it out for themselves? In an online business, or when your website is a lead generation tool for your offline business, your website is your sales team. Testing your website and marketing is like finding out which sales message works for your customers, and then training your sales team to share that message.
In a technology setting this might sound daunting, but it can be a very simple and effective process. It can also be a very profitable one, as a website doesn’t forget it’s training. It will keep sharing your best sales message 24 hours a day, 7 days a week. It also prevents you from wasting money on something that doesn’t work.
Increasing your online sales is similar to doing a science experiment!
Do you remember the science experiments you did in school, where you grew one plant under red cellophane, one under green cellophane, and one under blue cellophane and then measured how fast the plants grew? Even if you’ve had no further science experience since primary school, this basic concept is all you need to know to get started with your website and marketing testing. Website and marketing testing is a little more sophisticated, but ultimately the goal is to find out what works the best and then to make that ‘best’ the new normal.
When you conducted an experiment at school, your teacher would have asked you to write a hypothesis at the beginning of your report. The hypothesis is the question that you seek to answer by performing an experiment. In the case of the plants’ growth experiment, you might have written that your hypothesis was that a plant under blue cellophane would grow faster than a plant under any other colour of cellophane. When you grew and measured all the plants, what you were doing was testing your hypothesis to find out whether it was a false belief, or whether it might be true.
Experimentally testing your website and marketing is the only way to find out whether your ideas or the ideas of your marketing consultants are working. Without testing you are actually only guessing. Your belief could be false and taking money out of your pockets.
Don’t rely on a HiPPO
What many businesses do is rely on a person who states they are an expert (otherwise known as the Highest Paid Personal Opinion – HiPPO). Experts get you part-way there, but the only way to find out if an idea truly works is to look at data to find out what happened. Fortunately for us, technology has advanced such that the data can be generated behind the scenes by software, and all we need to do is look for the answer. Some modern testing tools are so easy to use that they automatically do all the calculations and then give you a yes or no result.
Testing the outcome of specific sales goals
When testing your sales message there is more to consider than just whether or not your customers like it, it is also necessary to know what is the goal of the interaction. Some example goals might be:
- for website visitors to complete a purchase through a shopping cart
- for website visitors to provide their email address to a signup form
- for website visitors to read and share a blog post
- for website visitors to make a phonecall to your business
- for email subscribers to read your email
- for a cold lead viewing your advertisement to navigate to your website
Testing your website has the added bonus of giving you the ability to be very specific about your sales goals and then to optimize your sales messages in order for those goals to be achieved. By separating your entire lead generation and sales process into a series of goals and then testing for each one, you can create a fully functioning system without having to guess at how a lead is moving from first impressions of your business all the way through to the final sale. You can carefully craft and persuade your customers to take the courses of action that you design so that their purchasing experience is enjoyable for them and profitable for you.
Sometimes your experiment doesn’t give the result you hoped for. But that’s good news too
Bringing you back to the analogy of the school science project, if you grew enough plants under different colors of cellophane such that you had a reasonable level of certainty in the experiment, you might have discovered that your hypothesis was incorrect. So if you believed that the plants growing under blue cellophane would grow the most but instead the plants under the red cellophane grew to twice the size, you would have concluded that your hypothesis was incorrect.
You can find yourself in the same situation when building your own website and online marketing campaigns. Sometimes what you think is going to be a great idea makes no difference to the customers whatsoever, therefore wasting your time. Or even worse, a change could cause a significant loss in revenue. Even if an expert assures you that in their experience it has never failed, your customers could behave differently and be the first to buck the trend. If you didn’t test your changes you would not know until it was too late and the damage was already done. Basing your business on hope in this way is a surefire way to keep you awake with anxiety in the middle of the night.
The simplest type of test to alleviate this kind of worry and mitigate the risk of a business decision failure is called an A/B test, or split test. In this test there are two conditions: the website as it is now (the A in the term A/B test) and the website as it would be if the change was implemented (the B in the term A/B test, also known as the challenger). In an A/B test, your website traffic will be randomly directed to one of these two scenarios. When enough traffic has passed through you will have some confidence in which alternative to choose.
In the case where you were sure that a change was going to be profitable but it turned out to make no difference or to have a negative impact on business, you will have significantly reduced your risk by testing. Firstly, you will have reduced your downside by only testing against a portion of your traffic, and secondly you will be able to pull out the change that didn’t work, thus reinstating your most profitable implementation. In this case, learning about what didn’t work and coming up with a different hypothesis is a valuable by-product of your testing efforts.
Sometimes your experiment pays off big-time!
If, on the other hand, you discover that a change to your website or online marketing campaign results in a dramatic increase to the outcome of your sales goal (this is known as a conversion rate – i.e. the percentage of your visitors who convert into customers) you can quickly implement the change and take advantage of your increased profitability. It might take several tests to find this bonanza, but the rewards are great.
If you double your conversion rate then it is equivalent to halving your cost-per-acquisition. This gives you the ability to expand your marketing reach by entering new markets or the ability to dominate your current niche by increasing the amount and/or types of advertising. This could bring you a disproportionate number of additional visitors and catapult your business growth.
Similarly, if you sell your products via an affiliate network, doubling your conversion rate would enable your affiliates to earn twice as much from promoting your products, and then you may find that suddenly your products are being promoted much more than your competitors’. Even a small change could become a tipping point with a subsequent snowball in customer demand.
Essentially, you test so that you can be as certain as possible that the tactics you employ are helping rather than hurting your online efforts. In doing so, you incrementally improve your platform and by reinvesting your profits you can develop a streamlined sales process many multiples of profitability higher than your first version. Compare that to the business who never tests and 50% of the changes are marginally better but 50% are marginally worse, never quite reaching growth targets and not knowing why. There is a good reason why large businesses utilise testing so extensively, not just because they can afford to now that they are large, but because testing when they were smaller made them large in the first place.
So to sum up and conclude:
If you discover that a proposed change does improve your conversion rate then you will lower your cost-per-acquisition. If the subsequent increased profits are reinvested into growth your business may end up boosted by a tipping point that snowballs your efforts.
Testing your website and online marketing is like performing a science experiment. The goal is to find out what converts your visitors into customers and then to apply it to your business in order to grow your business.
Without tests you are guessing. Relying on HiPPOs and hope will keep your business small and your revenue variable.
Tests enable you to target the conversion of a specific goal – as you improve the conversion rate of each goal in your sales process a streamlined sales system will emerge.
The simplest type of test is an A/B or split test.
If you discover that a proposed change does not improve your conversion rate then you will have prevented the implementation of a failing business tactic and you will also benefit from the the learning that came from the invalidation of your hypothesis.
Large businesses test their website and marketing campaigns. They started when they were small companies and that is why they are now big.
Web Data Analytics specialises in growing your conversion rate online. Current and past clients include e-commerce in B2B and B2C industries, lead generation websites and corporates. Petra Manos is experienced in tracking and interpreting website visitor interactions, improving website conversion rate, and attributing online sales to marketing channels. Please call Petra on 0405 123 696 today, email firstname.lastname@example.org, or visit http://www.web-data-analytics.com.