If Uber Eats orders are down how will you track the root cause?

Carify scope:

• Ubereats is the intercity food delivery service where users can order from listed F&B chains and get their orders delivered to their doorstep within an hour 
• We assume that this service only delivers cooked food (no groceries etc.)
• We assume orders mean orders placed by consumers and not orders completion of delivery 
• We assure orders are down in the timeframe of a quarter (clarify this with the interviewer)

Chart user journey to articulate the funnel

• Users signup/login to ubereats->Views all the listings/nearby/offers-> searches for dishes/cuisine-> Views the filtered listings-> Clicks on an F&B outlet-> Views F&B detail page -> Selects items, adds to cart -> Reviews orders in cart -> Pays and completes the order 
• Note: We will not explore the journey of fulfillment that comes next. Let me know if that's ok?

• My assumption is that the #total users visiting the checkout/cart hasn't dropped off. is that the right assumption?ok
• We know that there is a drop at the bottom of the funnel. #total orders per quarter has dropped, which means either 
  • fewer users are clicking on the 'pay' button i.e. more are abandoning the cart
  • fewer users are able to complete the transaction i.e. payment failures
• Need to go a level deeper now:
  • Seasonality: The drop is compared to the last quarter. Were there festivities like Christmas, Easter or any special holiday in the last quarter?
  • Regionality: Are the drops concentrated in a particular region/market? 
  • Marketing: Were there any promotional campaigns going on in the last quarter that is stopped currently?
  • Platforms: Are the drops more concentrated on a particular device - mobile/table tor web?

Analysis:

• If all else is well, the drop has something to do with either design optimisation of the checkout page (this might not be indicated by a severe drop but by a gradual one) or it could be related to payment failures.
• In case it's a design-related fall off, the following steps need to be taken:
  • Use heat maps to check out user interaction on the checkout page
  • Use metrics such as average time spent on check out page (if it's higher, that's a problem) 
  • Use A/B tests to identify optimised designs with high impact and visibility/utility of call-to-action buttons
• In case it's a severe drop related to payment failures or lags, the following steps are taken:
  • Run A/B tests to determine whether users find it convenient to make payments. It could be related to:
    • Absence of real-time validation of addresses/cards
    • Absence of card-on-file features
    • Absence of convenient/good/popular payment methods 
    • Long time spent on the payment page to complete the transaction (>20 secs leads to bad results)
  • Involve payment providers (account/tech teams) to dig into technical problems that might have cropped up:
    • related to payment processors failing/lagging
    • payment gateways failing/lagging
    • problems with approvals coming from issuing bank/ acceptance contract related problem
    • fraud detection protocols rejecting genuine user payments

In conclusion, narrowing down the problem to a particular area of the funnel that's affected, followed by looking into all possible reasons UX and technical will help uncover the root cause of the issue.

Any feedback to my answer to this Uber problem solving question would be greatly appreciated.

• Orders down respective to what time frame ?

• Uber Eats operates in food orders, groceries orders/deliveries,etc; is this only for food orders ?

• What is the value of reduction in the orders(scale of impact)

• Orders count have reduced v/s last month

• Orders referred here are food orders

Uber Eats business model is based on successful food orders delivered and it is directly linked to the revenue metrics(north star metric), hence a P0 problem.

Customers- 

People who order food for delivery from app

Partners -

Restaurant partners who list their food on app

User Journey :

-Customer comes on the platform

-Searches for the restaurant on the portal

-Selects the dish/food item

-Checks rating, price, detailing

-Places order

Internal Factors

I would want to validate few factors for RCA of the problem

1. Instrument Error : I would want to check whether the tracking is working properly or not

2. Scale of Impact : I would want to check the absolute number of orders that have reduced v/s last month(as stated in assumption)

3. Product Release : I would want to check whether this issue corresponds to the last release or any bug/patch fix was done, to rule out issue caused due to interlinked features

4. Problem Finding :

   1. Overall

      1. I would want to check whether this issue is with a particular city /area / location

      2. I would want to check whether the issue is specific to a device, app version

   2. Partners : 

      1. I would want to check whether the availability(number) of restaurants has changed v/s last month

      2. I would want to check whether the average numbers of orders has reduced for my restaurant partners in a similar number

   3. Customers :

      1. I would want to check the user CTR, traffic, drop-off on the various parts of the user journey (as detailed above) v/s last month- to help to identify whether any part of the user order process is not broken

      2. I would also want to check whether the average orders per users(as per my segmentation) is the same or has changed- to understand churned users/activated users.

External Factors

I would also want to ratify external factors for RCA of problem such as :

1. Change in Market Dynamics :

   1. Entrance of new competitor-perhaps which caused my users to use other competitor

   2. Marketing campaigns of competitor/promos,etc

2. Social Media /News bad publicity, negative reviews

3. Market Conditions : Event which caused people to order less; for eg : COVID pandemic,etc

Overall I would do the following -

1. Identify the problem - Absolute value and affect(drop comparable to last month)

2. Brief the stakeholders on the issue.

3. Conduct RCA by diagnosing the health of -

   1. Internal factors - key metrics at overall, partners(restaurants), customers level

   2. External factors - such as change in market dynamics, negative social media ,etc

4. Share my findings with stakeholders and my next steps.

5. Based on the fact finding from above I would establish my next steps to fix the issue