Instacart App critique
Check out Instacart for online delivery during this special time period!
Before we dive in…
Hi all! How are you doing these days? We hope you stay healthy! Here comes our first app critique sharing — Instacart. We chose Instacart because of the coronavirus outbreak and people started to interact with online delivery more than ever. Therefore, we are curious about how online grocery shopping experience will affect people’s daily routine! Our discussion encompassed their products on different platforms like iOS, Android and web.
What is Instacart and how does the business work?
Instacart is an on-demand grocery delivery service mainly served in the USA and Canada. It was founded in 2012 by Apoorva Mehta, a former Amazon employee and valued over $7.6 billion in 2019.
We concluded the main business model of Instacart:
- The markup on prices for specific stores
- There is a delivery fee for products that are ordered and
- There’s also a membership fee for actually using Instacart.
Topic 01: consistency & trade-off on categorization
We noticed that providing online delivery for multiple stores is one of the most important traits of Instacart. However, different stores have their own ways to categorize groceries. Would it be better to build a new but consistent categorization for all stores or allow more familiar but various categorization among stores?
Instacart chose the former one and pushed hard to build a consistent category system through the whole application. How does the new system work on Instacart? We analyzed its designs on the web version as well as the mobile version and found something interesting.
The balance between spaces and information structure?
Compared to mobiles, web-pages provide wider spaces for more information and therefore can fit more complicated information structures. On the web page, Instacart combined the 1st and 2nd category layers together to provide a detailed preview on categories. However, on the mobile version, with the limitation on space, Instacart unbundled 1st and 2nd category layers and moved the 2nd category layer to the top of the product list page.
The mobile design seemed reasonable and fit for the mobile ecosystem at first sight. However, the real experience is less smooth and it’s easier for users to lose track compared to the web pages. The main problem of mobile version targets on the 2nd category layer. It violates the Recognition Rather Than Recall principle in heuristic evaluation. It is easier for users to forget what they have seen when scrolling right farther and it is rather annoying to scroll back and forth. Moreover, the scrollable category layer fights with the clickable/expandable product lists on one page and multiple actionable buttons confuse users a lot.
What will be a better way to accommodate multiple layers of information on mobile? We researched other applications and found better alternatives —
Safeway is a good example of organizing different layers of information. It shows all layers of categories on the same page, which follows a similar scheme with the web version of Instacart. The simple expandable list makes it easy to navigate and track. We thought that maybe Instacart can follow the same scheme to separate the function of the category page and the product list page as shown on the right.
Topic 02: personalization
Personalization features are hard to ignore when using Instacart, such as my list, group cart, like button on the product page, substitute list…We believe that Instacart tries to personalize users’ shopping experience as much as possible. Personalization is a nice way to increase customer loyalty in the long run; however, too much personalization also makes it overwhelming, especially for new users. Having shared our experience, we agreed that personalization features on Instacart are a bit aggressive. The design for substitution is the most typical one.
Substitution
It is common that users need to select substitutes for products they choose when the stock is running low. Wholefoods provide the substitute option before payment and Lotus checks the substitute only when the stock runs out. However, things get a bit annoying on Instacart. Users need to choose substitutes when they put things into carts and the substitutes are not recorded for the next time. If users delete items in the cart and add them back again, the substitute they had chosen would not be remembered by the system, which means they will have to select it again. Moreover, the whole flow is not effective and users might feel time wasted if they don’t end up buying all of the items in the cart.
We thought more about why Instacart chooses such an aggressive way to personalize user experience. Instacart possibly follows the experience that “ask someone else to buy things for you” but not “personal off-line shopping”. It would be more natural to decide all possible substitutes before the person buys things for you in real life. Also, it can reduce the risk of disappointing users when products they bought run out of stock.
Can there be a better way for substitution? We consolidated the substitution process to the check-out stage. After deciding what to buy, users need to choose substitutes for low stock items. The original and modified flows are as follows:
Topic 03: Notification
The notification system in Instacart is unique. We noticed that the notification will show up at the bottom of the page. It is interesting that firstly we thought the notification is pinned on the page and cannot be dismissed because we tried to tape, wait, scroll up and down and the notification was still there, which is annoying. The design in Android is even worse and more distracting. (Hi readers, please stop here and think what would you do if you want to dismiss these notifications?)
It took us some time before we realized that it could easily be dismissed by scrolling left. It is an interesting finding that confirms how nuance in design leads to completely different user experiences. Scrolling left to delete/dismiss should be intuitive enough for users since it is one of the fundamental gestures in iOS. Why neither of us as users intuitively scroll left when we see the small card? We believed it is because users are accustomed to tapping notifications to view the detail, swiping them up to dismiss, waiting for seconds and they would disappear. Designers here should refer to the design guide specific for notification but not general iOS gesture.
However, another interesting thought jumped into our head — “will it be a good design if notifications (seem) stick on the screen?” It was hard to say “yes” at first sight but definitely worth a “yes” under specific circumstances. For users who just finish an order, the biggest reason to open the application shortly again is to view the process of that specific order. It would be nice that users, especially new users who are not familiar with the application, would not lose track of their orders on whichever pages they are on.
How can we combine the different needs of users to improve the current design? Compared to wait for notifications dismiss automatically and swipe, we thought that adding a direct actionable button onto the notification would be better and users can decide if they would like to view the notification according to their own understanding of the notification.
Takeaways
We unleashed our critique superpowers on our first discussion and it went crazy! It was difficult for me to reiterate every key point that we mentioned during the session. We will definitely hone our skills in both critiquing and writing for our upcoming sessions. That being said, here is a recap of some main points that we found during the discussion:
- We didn’t think about multiple stakeholders included in the application. For example, we didn’t talk about the delivery staff’s experience as we were not able to access it. It would also be an interesting topic to think of as it is an important part of the business model.
- Instacart sacrificed unique aisle arrangements across different grocery stores and decided to keep it consistent. We talked about it in the topic 01. However, as for the application, is it the best option?
- We noticed that the group cart experience was hidden and a bit messed up. It is supposed to be a feature that Instacart would like to promote or not? Why?
- Shall we have Re-order as the top choice over other options on the home page? Most of us thought it was too aggressive to have this wording choice as it might indicate to reorder directly without allowing the user to make any change.
- We mentioned that the whole user flow is a bit too long and users are easy to lose track in the process. What are the current user flow and how it can be improved to help users track where they are and how long it takes before they finish the order?
Up Next…
We are going to talk about Venmo next week! How do you like it? Feel free to share your comments on this app! Also, In the next discussion, we attempt to try different analysis methods such as heuristic evaluation and embed design guides to perform deeper analysis. Since we just started this initiative, we appreciate your input! Let us know your thoughts below or via other platforms. Last but not the least, stay healthy! :D
