June 3, 2015
Growth is a result, not a strategy. Figuring out how to grow is the strategy, and really, the only way to figure that out is through testing. Hiten Shah, entrepreneur and one of the thought leaders behind Quicksprout, KISSmetrics, CrazyEgg, HelloBar, and Qualaroo, explained his thoughts on testing to designers, product people, marketers, growth hackers, and salespeople during Future Insights Live in Las Vegas.
Whether it’s A/B testing your button copy, creating exit surveys, or directly emailing your dormant customers – testing should be an ongoing process to inform your business decisions. Here are my top 5 takeaways from Hiten Shah’s Always Be Testing presentation:
1. Data doesn’t make decisions for you.
There is a difference, in fact, between data-driven and data-informed. Data should cause you to ask more questions.
2. Identify the problems your users have.
This seems obvious, doesn’t it? But how many times have you deployed a team of engineers to solve a problem that was more of a pain point for your team than for your customer? If calls to customer support are only bothering you, and not your customer, then you may need to spend more time figuring out what points of friction are actually preventing your customer from converting.
You determine the problems that customers are having by testing and gathering information.
3. Know when to run more tests and when to get more traffic.
You’ve got folks coming to your landing page in droves, but none of them are taking the action you want them to take. What to do? More tests. Pull out a graphic. Change a graphic. Move a button. Copy the tried-and-true 37Signals Plans & Pricing button. Scrap the Plans & Pricing button and create one that speaks to your own audience. Learn what users care about. Is their behavior effected by any of these changes?
You don’t have many people visiting your site or opening your emails, and your sales numbers are not going to impress anyone. What to do? Get more traffic. It’s not time to test yet.
4. Perform no test without a hypothesis.
Making an assumption is the only way you should be justifying creating any test. Create your best guess on how to solve the problem at hand.Utilize an equation with a [variable], a [result], and a [rationale] so you can assess the potential impact of each idea.
If [we modify the website in this way], then [people will click the call-to-action button] because [the copy will create an emotional response.]
Where will you come up with these hypotheticals? From your data. From your research. From reaching out and asking folks why they did not perform the expected behavior.
5. Perform no test without complete documentation.
A/B testing requires a lot of patience. Do you really want to go through all of this, only to have the new Head of Growth ask you in a year why you don’t have a action graphics on your site?
Collect dates that each test is run. Utilize control screen shots. Document screen shots for every variation. Include the decisions made from the testing, and track what you learned. Write up your project checklist before you create the test.
Create a discipline for this process so you have a very good justification for everything you do. What learnings do we want to carry over from the past? Avoid unknowingly decreasing your conversions. Be prepared to explain and justify your features with your documented testing.
“If you’re not testing, you’re not learning. Always have tests running, no matter how small.”
Image Credit: Flickr/clement127
Tech.Co is happy to announce our media partnership with Future Insights. Follow the discussion online at #FILIVE!
Did you like this article?
Get more delivered to your inbox just like it!