Clarifying questions to the data analyst:

1. Over what period of time have you observed this? – to find out if this is a problem generated after any new changes in the customer journey or for any external factor like other competitive offers?

2. What was the conversion rate for lunch before this analysis was done? – Has it increased for the others and decreased for lunch? – to understand if the conversion rate with lunch has always been this low

3. During which stage of the order journey customer is dropping off during lunch?

I.Now I would consider the customer journey and understand how it is different when someone orders a lunch vs dinner or breakfast using our app.

Usually breakfast and dinner would be ordered from residence while lunch would be taken at office..often in short time in between meetings. The item search + delivery time + path to be travelled inside office premises to get the parcel would be higher during lunch. ( Assuming food delivery is opted majorly by ppl staying outside home for work)

If this hypothesis is tested and comes out to be true..then the food delivery app might consider corporate tie ups inside canteens to solve this issue.

If there is issue with payment process time, a delayed payment option can be included during lunch.

Metrics to be considered:

%of active users who orders for breakfast and dinner but skips lunch order( doesnot open the app at all)

% of active users who proceeds for lunch order but drops off in the middle

Conversion rate of lunch during weekends

II. Next, are competitor brands offering something during lunch which seems like a tempting offer?

A temporary offer inclusion of similar sort ongoing in other apps can be a small test for this

III. Is food menu to be blamed?

%of item search during lunch showing no result

In that case an option to give suggestion for dish inclusion when not found can help

Tie up with restaurants which provide such items can help too

Out of these 3, we should first check with problem I and III, and based on the test result confidence value go ahead with one. The competition check is required but can be temporary as well based on offers..and if we are seeing an overall performance better than competitors, point II should be checked only if I and III Fail.

First, I would map the user journey and identify the factors that contribute to conversion in the food delivery app.

1. Search/Discovery: Users can search with different attributes restaurant, dish type, cuisine, etc.

2. Purchase: Users add items in the cart and make a purchase decision basis delivery time, cost of purchase, coupon discount available.

3. Payment: User selects one of the payment options/methods available in the app

Second, I would like to analyze the conversion contributor metrics associated with the three stages identified above, compare them across three meal times, and identify any discrepancies in the process. For eg. I can discover that the average coupon discount available during lunch is 30% lower than that during dinner or breakfast or the number of discount/promotional coupons available is 50% less during lunch when compared to dinner or breakfast or the number of food items available on my app during lunch is significantly lower than breakfast or dinner.

Third, if I am unable to determine any significant discrepancies from the app usage I would like to analyze some if there are any external factors that are contributing to this phenomenon or if there is a behavioral aspect that we may have missed in our root cause analysis.

– competitor conversion rate, delivery fee, food catalog(variety) vs ours
– canteens/mess availability in the concerned region during office hours
– customers intent (they could just be scrolling to decide what they want to order in for dinner)
– customers think it would be a hassle to get the food delivered in the office

Alternatively, you can formulate quick hypotheses and validate them using data.