Theory:
It's easier (and more likely) for people to make choices when there are less options to choose from. If you offer too much, you risk Choice Paralysis.
Application:
Check your website. How many options does your customer have? Can you reduce the amount of options? If you can't reduce the total amount, could you divide the selection process into several steps with only a few options each? If you can't simplify for all customers, think about providing a customer segment with different/limited options, specific for that segment.
Example 1
Now you won't need to limit product choices for Group A since they already made their product choice. But it's very useful to limit the product choices to Group B. You don't want to overwhelm Group B with specific parts or product variations that only makes it harder for him/her to make a choice. In Magento, this can easily be done by changing the visibility of the product. Specific products are set to 'Search' only, general products are set to 'Catalog, Search'.
Example 2
You have a shoe-store and a customer bought some shoes size 40. After some months, that customer returns to buy another pair. Chances are pretty big that he won't be interested in sizes smaller or larger than 40. Use customer segmentation to put this customer in the 'size 40' segment and only show him shoes that will fit his feet. Of course other sizes won't fit, but if they're still presented as an option they're just making it harder for the customer. (Of course you will need to be careful with this and present the customer with the option to view all products. Maybe he wants another size, for example to buy someone else a present).
Most of my content is published on LinkedIn, so make sure to follow me there!
Recently I've seen some (often absolute) statements going around, generally in the line of "open source commerce platforms are a terrible idea". Now of course different solutions always have different pros and cons.
A hierarchy of evidence (or levels of evidence) is a heuristic used to rank the relative strength of results obtained from scientific research. I've created a version of this chart/pyramid applied to CRO which you can see below. It contains the options we have as optimizers and tools and methods we often use to gather data.