I'm currently working on a new IA for a webshop. I've drawn out two examples.

OPT 1 (option 1) (more a flat(ter) hierarchy)

This 'could' be our new IA (it's restructured). More choice (of categories) at the beginning and it takes a maximum of 3 clicks in order to get to the PDP.

OPT 2 (option 2) (more a deep hierarchy)

Our 'old' version of the IA. Over-all less choice, but it takes on average one more click in order to get to the PDP.

Main difference between them: OPT 2 has an additional three (generic) main categories at the beginning, that were deleted in OPT 1.

QUESTION: What would be the better solution here: have less clicks but more (specific) choice, or more clicks and less (generic) choices?

Personally I think, OPT 2 starts with three too generic category names that are confusing for clients and they don't exactly know what to expect. But I'm not sure if starting with 8 specific categories, would be too much for users and impact their cognitive load because it's too overwhelming.

Sadly we can't test prototypes with end users at this moment.

Looking forward to your input.

P.S. the arrows stand for navigation possibilities to direct jump to (for example) PLP from the start.

enter image description here

3 Answers 3


It`s hard to come with an answer with the lack of details but here are the factors to consider.

What are the 8 categories in OPT 1 and what are the 3 categories in OPT 2 and how did you decide upon ?

For example in an ecommerce when deciding on categories we could put a broad category Electronics that includes PC and TV

or we could create two categories from the beginning PC and TV.

Clearly option two of my example is the go to because why would I determine my user that is looking for PC to first select Electronics and then select PC.

That is what you would need to determine. As I do not understand what your categories serve I cannot make a choice for OPT 1 or OPT 2.

The best way to solve this dilemma is to understand the users and as you said that you cannot do the user testing you can use the Free account on Optimal Workshop to do a small IA testing. https://www.optimalworkshop.com/treejack

  • Thank you Zasul for your answer. I think the example that you've used is with Electronics > PC and TV is pretty accurate. You can see it that way. OPT 2 uses three generic categories (like Electronics) and OPT 1 directly starts with 8 specific ones like PC and TV. Where in our case with OPT 2 (the names that are used, are not as clear as for example 'Electronics' is. So too generic.
    – Wishit
    Oct 18, 2019 at 8:03

I would say that in principle, and without having seen the details, I would go for OPT 2 but making sure there is a good visual indication of sense of place (e.g. breadcrumbs) so the user doesn't get lost in the deep hierarchy.

My reasoning is that it should be easier to make several choices from between a few items each than fewer choices but between many items. All of this is just an assumption and ideally you want to prototype and test it with a few people.


Choosing from more options slows down decision making and costs more cognitively than with less.

enter image description here

Included two comparison scenarios

A - more choice - fewer clicks

B - less choices - more clicks


  • easier conversion measurement/design optimization
  • choosing from more options requires more intellectual effort
  • theoretically shorter decision-making process, but more difficult
  • greater room for maneuver (more interaction in the designed architecture)


  • easier decisions, but more of them
  • there may be boredom, too many stages, increasing the rejection rate
  • less possibility of navigating between options
  • the first choice narrows quite widely

I would use option 1 (maybe limit the categories and did a little less in the beginning (?)) without forgetting about the search engine

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.