Conversion 101: Using the Scientific Method
One of the first things that has stood out to me in my Conversion Rate Optimization research is the emphasis on building hypotheses and testing. A lot of conversion rate experts discourage businesses from blindly adopting best practices from other websites because there are too many complex variables. Even if you are comparing to other websites in your industry, there can be significant differences in your customer base, price expectations, and UI preferences that will affect your conversion rate differently.
So you need to be ready to collect and analyze data from your own website.
Getting Set Up
There are a few tools that you will need to successfully track and analyze conversion:
Google Analytics: It can be assumed that most people have this set up if they're looking to improve conversion because this is the tool that tells you what your conversion rate is. However, if you're not set up with conversion tracking in Google Analytics, then here's a helpful guide to setting up conversion goals in your Analytics tool.
A/B Testing Tool: This tool will let you compare the conversion rate outcomes of different changes to your site. There a few main options here: Google Optimize, HubSpot WRO, and Optimizely. If you're using Google Analytics, then I recommend Google Optimize because they have a free version and it's compatible with Google tools.
Survey Tool: For some businesses with lower traffic (the general rule is <100,000 page visits in a month) it might be more effective to test your hypotheses with a survey rather than an A/B Test. Yotpo and Judge.Me are some Shopify apps that have surveys built in to their tools.
Generating a Hypothesis
Once you're set up with the right tools, you want to figure out which changes to your site are most likely to improve conversion. Remember, we are just generating hypotheses here, not making guarantees about what will happen if you change an aspect of your site. This is where Conversion Rate Optimization best practices can come in handy, because they can form the basis for your hypotheses.
Here's a short list of questions you can ask yourself about your site to see whether it meets best practice standards for Conversion Rate Optimization. While this list isn't exhaustive, it's a good place to start!
Is my Call to Action (CTA) Button effective?
Does my website have a good user experience? Does it load quickly? Is it designed for mobile?
Is my marketing content convincing?
Is my Marketing is targeting an effective audience?
Do I have enough reviews on the site? Are they convincing?
Do I have a live chat feature?
Am I using abandoned cart email campaigns?
Are my site's images and videos showing the product's value?
In order to generate hypotheses, you can evaluate each of these questions based on the ICE Framework. ICE stands for "impact, confidence, & ease", and in this framework, you assign a value of 1 - 10 for the three criteria:
Impact: How many of your site visitors will be impacted by this change? How big of a difference will this make to your site? How much does your site differ from best practices for this element?
Confidence: How sure are you that this will improve conversion? How certain do you feel that this change will not have negative side effects?
Ease: What is your ability to make this change? Will you need to hire an external consultant or third party? Will you be able to revert this change easily?
Once you assign the score to each criteria, add them all up and that's the project's ICE score.
When you score all the relevant best practice questions using the ICE Framework, you will have a list of hypotheses ranked in terms of which ones will be most effective. I recommend you start with those first! By generating sound hypotheses and evaluating them before you start testing, you can reduce the risk that you are wasting your time and money on a conversion change that doesn't move the needle.
Bennett
P.S. Want to read more? Check out this Shopify article and this HubSpot article.
