Using Data to Create a Winning Product Proposal- A Case Study With Flying Taxi

Published in

The Startup

9 min readFeb 17, 2021

This article is the third part of the series and a continuation of the last article.

Objective

In the last article, we tapped into the minds of our users and assessed the market needs for our flying taxi service through extensive data analysis. Now it’s time to come up with a product objective and KPIs, determine an MVP, and decide on how to assess the KPIs and gather feedback. Ultimately, we have to leverage the research from the previous article and create a unique product proposal to convince our stakeholders.

Key Deliverables

It always helps to map out your key deliverables at the beginning of a project and tackle them one at a time.

Let’s outline our key deliverables on a high-level.

Part 1: Product Objectives

Aim: Identify a key product objective for our product. The product objective guides our KPI’s, so identify what your product should optimize at this stage.

The primary aim of our product should be user acquisition at this stage.

Why?

Flying taxi is an emerging startup with an innovative technology that people are not used to. Because of this reason, it’s imperative that the start-up focuses on acquiring new customers.
From the user research, it’s very clear that women are more likely to use the service, but they have major trust and safety concerns.
It is critical for the company to address these fears through effective marketing techniques by offering some demo sessions, free rides, or discounts.
The company should also strive to reduce the frustration and anxiety with the booking experience by focussing on product usability and reducing the average wait time.

Is just a clear product objective enough?

Although a clear product objective can set you apart from your competition, it is imperative to define Key Performance Metrics to measure if the product is heading in the right direction.

In order to do this, let’s clearly define our objectives and its key results.

Acquire new customers and improve the user adoption by 20% within first three months.

This method is called the OKR framework. Objectives and key results (OKR) is a goal-setting framework that helps organizations define goals — or objectives — and then track the outcome.

Objectives are clearly defined goals at a high level.

Key results are target metrics that clearly affect the success of our objective.

Sample Example

Let’s assume you want to lose weight.

Your OKR framework would look something like this :

Objective: Reduce weight and get fit.

Key Result 1: Run 10 Km in three months.

Key Result 2: Eat a salad every day.

You might not achieve your objective in a week or a month. But you can definitely say at the end of the day if you achieved your key result 2 or not.

It is also important to tie your KPI metrics directly to your key results. How else will you measure its success?

Let’s look at our key results and KPI metrics.

What does success look like for our KPI’s

Defining just the KPI’s is not enough. If you do not know your target, how will you know how the product is performing?

It’s also essential to keep the targets within an acceptable range.

For example: If you know on average there are only 10 customers in your store, setting a threshold of 100 will be too hard to reach.

Now how do you arrive at these thresholds? Data of course comes to the rescue.

Setting Threshold for our KPI’s

No of customers per day:

From our EDA, we identified our peak timings as

7–8 AM
6–7 PM
1–2 PM

So it’s very clear that we are looking at the stats at an hourly level.

How many rides do we have per hour/per day?

Our dataset is bound between Jan -July 2016 which translates to 181 days. There are about 120,994 rides in total. So on average, there is 668 rides/day.

But we are only operating our taxi service during peak hours.

Looking at our hourly calculation, we notice that there are about 30–40 rides during peak pickup timings.

So let’s put our threshold for No Of Customers during peak hours as 30.

2. Time Taken for each ride:

From our descriptive stats, we can see that 98% of our rides take 19 minutes.
Our flying taxi should at least aim to cut this by half.

Threshold for total time is 10 minutes.

3. Customer Satisfaction :

During MVP we should aim to get a customer rating of at least 3.
Customers who even bother to give ratings have a higher chance of referring to others. So this metric is crucial for us.

Threshold for customer rating is 3

4.Avg Wait Time

We should keep the wait time minimum, ranging from 1 to 3 minutes.
There should be a fixed schedule for the rides, and we should give customers an opportunity to book in advance.

5.No Of Passengers

From EDA we can see that 98% of the passenger count was ranging from 1 to 2.
We should keep this metric realistic and set a threshold for 2 customers in the beginning.
However, for long-term profitability, it is imperative to improve this number to 4–5. This would require some user research to understand why customers don’t want to share rides.
Antagonizing customers during the initial phase of the launch by forcing them to ride with 4–5 people might backfire. So it would be wise to keep this number in line with the stats and iterate upon later.

Let’s summarize our KPI’s and it’s success threshold

Part 2: Establish the feature set for your MVP

Now that we have established our objectives, and KPI’s, our next mission is to decide on the feature set to include in the MVP to measure viability, while keeping operational expenditure under control.

Let’s look at some questions to guide us in this process.

What times/days of operation should the service run for?
How many pick-up/drop-off nodes should we have?
Where should the nodes be located?
Should we initially use copters or homegrown hardware?
Should the pricing be fixed or dynamic? At what rates?

Here is a quick summary of all the decisions I made based on data and its reasoning:

Part 3:Instrumentation Plan

Having all these objectives and KPI’s, in theory, is great. But without an instrumentation plan to collect the relevant data, measuring your KPIs is not possible. So, for each KPI that we came up with, we will need to identify the events and their corresponding properties.

One way to do this is to map out the user flow and think about the events a user could trigger.

Example: If you want to measure the total time, you need the time when the customer started the ride, and also the time when the ride ended.

So your corresponding events become rideStarted and rideEnded.

Let’s correlate our KPI’s with an Event and its properties.

Part 4: Qualitative Feedback

Okay, so far we have been laser focussed on the current goals of the product. Now, it’s time to think ahead by creating a framework to further understand and optimize the product for future iterations.

A qualitative feedback survey questions for users after their ride can be a great way to do this.

Let’s quickly put together some questions we can ask the users. Keep this to a minimum of 5.

Part 5: Creating a Product Proposal

A product proposal should summarize everything you have done so far in a readable format so that your stakeholders can give a go-ahead for the MVP launch.

Here is a quick snapshot of how a proposal should look like.

Who is the product for?

Your stakeholders need to know that you have conducted enough research to target the MVP launch to a subset of power users to control expenditure and measure viability. So this is an important point to address in the proposal.
The Target Population for this MVP will be women and men in the age group 30–50.
This was quite obvious from the user research that women in this age group and income group (above 40K) are more likely to use the product.
Men also showed a reasonable interest in the concept, however, they were more worried about money.

What do we hope to achieve?

Well, what’s the claim of the product? Is it fascinating enough to get enough users? Are we the only competitors? What is the vision of our product?

Our proposal should address all these key questions.

Our product aims to reduce the traffic congestion in densely populated areas such as New York, Brooklyn & Kings by offering a seamless booking experience.
Our product also aims to reduce the average wait time by 30% and total trip duration by 50%.

Do we have any evidence of the claim?

•From the quantitative analysis done, we know that on average taxis have a pickup rate of 30 -50 rides/hour/day.

• We also uncovered 5 pickup points, 3 in New York , 1 in Brooklyn and 1 in Kings.

•Our stats also show that 98% of the rides had only 1–2 passengers.

So assuming we can have 5 mini helicopters which can carry 1–2 passengers with an average trip duration of 10 minutes and 5 minutes to flyback, we can have 1 helicopter run every 15 minutes.
Trip Breakdown:
3 Minutes — wait time
6–7 Minutes –Total fly time
5 Minutes — fly back

So in the worst scenario, 1 helicopter will have 4 trips with 1 passenger. So 5 helicopters with 4 Trips carrying 1 passenger each time give us 20 customers in an hour. If we can get 2 customers, then this range can go up to 40 customers.
So our range of customers can be 20–40 in an hour, which is pretty good.

What is the impact of the product?

User Impact

Our product will directly appeal to the users by catering to their emotional worries, by providing a seamless booking experience.
Our product will also reduce the uncertainty of being able to board a taxi during peak hours by having a fixed schedule and pre-booking option.
Fair pricing will also remove the decision fatigue of having to carry cash all the time and worrying about a certain segment of customers getting a fair advantage.

Market Impact

If successful, ours will be the first flying taxi service in New York city giving us an unfair advantage over other competitors.
This also aligns with the personal mission of our product to solve a critical challenge of the people by directly addressing their pain points and concerns.

Risks

Since this is a novel concept with no historic data to look back on, people are going to be very skeptical. Establishing trust and safety is going to be paramount here.
Pilot training and certifications: Users might not be able to gauge the credibility of their riders and this might be a potential risk if not established early.
Air Traffic & Pollution: The product will have to comply with air traffic regulations.
Noise Pollution: The intruding noise of flying cars might irritate people, and the company could get caught up in environmental issues.
Cost: Flying taxis might have really high manufacturing costs since they might need to run on batteries for optimal efficiency and to avoid breakdowns. If there isn’t enough traction on the user acquisition in the initial period, it could be a potential pitfall.

That’s it for this article. That was an overwhelming amount of information,but hopefully, you learned something new.

Let me know in the comments if you have any feedback.