Exercising my guerilla usability testing abilities - this two week project was a great opportunity to get out of my comfort zone and talk to strangers!


The objective was to identify some "quick" fixes that could be implemented tomorrow and improve the user experience.

But how could we make Instacart any better? I hit the streets of San Francisco to figure that out this week in my Usability Testing campaign!

"I thought this would be easier." — Roxanne

As you may know, Instacart allows us to order groceries and household items from a selection of local stores for delivery. Together with seven lucky strangers, I was able to gather very useful insights on how people are using the app.

Persona of Jack.


As I approached people I prompted them with:

"I'd like you to imagine you have a hot date tonight and you feel compelled to cook them dinner but can't make it to the grocery store. How would you use Instacart to get the ingredients for you meal?"

Within this prompt there are at least four tasks:

  1. Create a meal for a date

  2. Search for ingredients

  3. Add items to cart

  4. Check out

All my notes!

User profiles and quotes

I spoke to seven different people over the course of two days and gathered common lines of usability issues among them below. Each user displayed varying degrees of interest defining their meal; from the amateur chef who searches for obscure vegetables I've never heard of, to the user who just wanted to boil pasta - "batsa".

Dump and sort.


By taking common lines of usability issues, or pain points, and mapping them on a scale of importance and ease of implementation I was able to pinpoint the objective of identifying "quick" fixes that could be implemented tomorrow and increase the user experience.

Prioritize pain points.

No one knew there was more than one grocery store option.

Each store has a large banner with logo at the very top, but beginning with the store, Instacart Plus, is slightly misleading users to thinking it’s just the name of the app - not the store.

It’s difficult to see that you’re browsing departments rather than items.

There are 9 items above the fold which takes up the entire screen. It took some users a moment to move beyond the fold and enter each department rather than get the preview items.

During checkout, most people didn’t know you could “put things back” from your cart.

There is a hidden ability to swipe left to delete the item. Some users didn't find it at all and opted to select the item and manually reduce the item to zero instead of quickly deleting it. For example: if the item is in pounds, you'll have to reduce it incrementally in quarter pounds until it's to zero.



Indicators showing more than one grocery store option.

By including indicators on the banner and the menu, the user immediately has a mental map that allows you to circulate through stores rather than assuming the items are collated from all available stores.


Reduce previewed items in departments.

If the number were reduced to 3 items above the fold, then you would be able to see more than one department on the screen at a time - producing a mental model that requires you “enter” the department to see all available items.


Provide indicators showing we can “put things back” from our cart.

By providing an indicator the user knows that if needed, you can edit the item. By implementing these first three “easy” fixes, we can continue to refine the great product that Instacart is building.


What's next?

The "not-so-quick fixes" are more critical to the structural and business goals of Instacart. Some examples uncovered are the need for:

  1. Pictures for every item

  2. Availability of service in areas

  3. Filter by nutritional info

Please let me know if you have any other ideas!

Thank you for reading! If you want to read more, or read my blog post Architecture to UX.

Gavin Johns

Gavin Johns@gavinpjohns
Architect turned UX Designer