Ask yourself these three things before jumping to testing
During the last several years, a powerful hype has evolved around testing. Read this article (and its follow-up) before you decide to commit to testing.
During the last few years, we have had loads of discussions with several customers around the topic of split and multivariate testing. We have also participated in many projects that have had beautiful targets as well as great hypotheses behind the actual testing project. We have also seen many of these projects fail entirely.
For the sake of clarity: We strongly believe in testing as a tool and as a development method, but we don’t believe the hype that has been going on regarding testing during the last few years or so.
On some occasions testing is completely necessary, even a kind of “must” – and a really integral part of developing your site or (e)-retailing business. On some occasions testing is the key to a much better conversion funnel and often it is THE key to really get better return on investment from your online activities.
Still – what we ask is: never ever buy anything for the reason that it is a “must” to buy, because you “have” to be testing.
At least, before you commit yourself to expensive testing projects you should consider following questions:
1. Have you fixed the obvious pitfalls from your soon to be tested application, site, landing page or sales funnel?
If you have not, then why would you bother to test it? Fix the problem(s), in many cases you can very easily see which problem area to fix and even how you should fix it. For example if you think that your sales funnel really sucks, why would you test? You are probably right. Discuss it with your team, benchmark it, discuss about it, brainstorm about it, mind-map it.
If you don’t know how to fix the problem, you probably have some technical resources either in-house or in your platform provider who can fix it. These problems are also usually easy to fix – problems are related to calls to action and some obvious things at your end, such as poorly-sized fonts or similar small(ish) areas.
Fix those and see what happens, don’t pay consulting hours for big renewals before you fix the small things. Also remember to steal with pride from the biggest web services out there – they have spent their money already – why would you?
2. Have you checked from your platform provider what the cost would be if you do go ahead with testing project?
In many cases the price tag will be substantial when you are going to make changes to normal featured processes, so you will be paying two or even more vendors if you are unlucky.
The point is not to give you excuses not to test but to make sure that if you do go ahead with the testing, you ensure (beforehand) that the feature set that “wins” the test is possible to implement on your platform and that the platform provider/your internal team/your niece or whoever is doing the development actually has time to do the development.
In many cases they don’t have time or the winning formula is not possible to implement or would be extremely expensive to implement. Then you have spent reasonable amount of euros for, hmm, nothing. So, always check the pricing and or resource situation from your vendor(s) and/or from your internal resources and budget it before committing to testing.
3. Have you benchmarked your pricing compared to your competitors?
If you are retailer, and you are in agony that your site is not performing decently, the first thing to do is to check what the pricing levels are both locally as well as globally. I often see companies thinking that:
- Their site is performing badly
- Their agency/vendor is doing a bad job acquiring traffic
- They have to have tons of different landing pages
- They have to really push split testing to make a better ROI from your site
- They have to do all of the above and more.
What I often can tell from just a few hours of analysis is that there is for example a new player on the market globally and which is pushing their new product with a significant discount and free global deliveries. Often the answer is not the testing; it’s your selection, or the wrong categories on your site, or too high costs for logistics. Check your home base and have a solid ground before accepting the belief that your site is performing badly.
As many smarter guys than I am have said: competitors are global with (e)-retailing businesses, consumers are really calculating and they most certainly will order it where they will get the product with the best price. And please remember that the best price is not always the cheapest, it’s the best combination of excellent price, magnificent product and best customer service or user experience.
Hope you enjoyed this – having gone through the most common and easiest to fix mistakes I will continue next week with some more in-depth questions.