You want your website to generate more revenue, right?
Of course you do.
If you didn’t, you wouldn’t be reading WiderFunnel’s blog. Generating more revenue is what the Conversion Optimization industry is all about.
That’s why I’m concerned about all the talk I hear about “optimizing for micro-conversions”, especially micro-conversions related to Social Media.
What are Micro-Conversions?
We define a micro-conversion as a visitor action that is either on the path toward a revenue-generating conversion or is not directly related to generating revenue.
In other words, a micro-conversion could indicate interest or intent, but does not directly represent a sales lead or revenue.
Your website probably has many actions that you could track as micro-conversions. For example, a visitor could trigger a micro-conversion by:
- Tweeting about your blog post
- “Liking” your blog post on Facebook
- Viewing the “Request a Quote” page
- Visiting the checkout page
- Viewing the whitepaper download page
- Clicking the ‘View pricing’ link
- Viewing the software trial download page
- Proceeding to the third step in the checkout process
- Viewing a product detail page
- Viewing more than three pages in a session
- Adding a product to the cart
- Subscribing to your RSS feed
- Commenting on your blog post
- Watching a demo video
- Using your search tool
- And many more secondary Calls to Action
So, should you run micro-conversion optimization tests to improve your results for these?
No, you shouldn’t. Allow me to explain why.
Why Some Marketers Optimize for Micro-Conversions
The most common reason people optimize for micro-conversions is to speed up their experiment results.
The length of time needed to run your A/B or Multivariate test depends, in part, on the amount of traffic and conversions you get. Most websites get many times more micro-conversions than sales or lead conversion, so your experiments may complete faster using micro-conversions as test goals.
For example, consider a typical gated form that a software company uses to generate leads for a sales team. The conversion funnel may look like this:
At each step, there is a certain drop-off rate. If the home page gets 100,000 visitors monthly, the counts for each page in this hypothetical scenario could be:
a. Download Page: 20,000
b. Gate with Form: 15,000
c. Thank-you Page: 5,000
Choosing Your Conversion Goal
Let’s say you want to run an A/B/n test on the Home Page above.
You need to decide whether your experiment goal is: a) visit to the download page, b) visit to the gate page or c) a download completion.
If you choose option “a)”, to use the download page micro-conversion as the goal, your experiment will complete faster since you’ll have 20,000 goals triggered rather than 5,000.
That’s good news, right? You can finish the test sooner and move on to the next test.
Not so fast.
The False Micro-Conversion Testing Assumption
You’ve made a big assumption there.
If you assume that there’s an equal drop-off rate through the funnel regardless of the Home Page variation shown, it shouldn’t matter which one you choose.
But will your funnel completion rate (from Download to Thank-you Page) really be the same regardless of the Home page variation?
In our experience, that is often a false assumption.
We have seen many examples where the first step of a funnel can dramatically change the completion of a subsequent step, even a step that is several steps removed.
Here’s a typical example A/B/n Test result through the funnel:
|Home Page||Download Page||Gate with Form||Completed Conversion||Conversion Rate Lift|
Depending on the micro-conversions you were counting as your goals, here’s what you would have decided:
|Conversion Goal||Winning Variation|
|a) Download Page||Variation C|
|b) Gate with form||Variation A|
|c) Thank-you Page||Variation B|
Clearly, the best performing variation is actually Variation B. It lifted the lead generation conversion rate by 46%.
But, if you had run the test using Download page visits as the goal, you would have chosen Variation C as the winner and could have seriously hurt your results!
What should you optimize for?
For Conversion Optimization, you should always set your test goal as close to revenue as possible. Optimize for direct sales, average order value and qualified leads generated.
Please continue to look at micro-conversions in your web analyses, but for Conversion Testing, stick to revenue-producing goals.
For examples of revenue-improvement tests, take a look at these Conversion Optimization case studies.
What do you think?
Which micro-conversions are important to your business? What do you use as your Conversion Optimization test goals?
Subscribe to get experimentation insights straight to your inbox.