How to optimize as requests from Uber, since about 80% give up or wait a bit to actually request a driver when the dynamic price appears?
You can read the Portuguese version here.
Before returning to Brazil to work at Youse (An insurtech that is trying disrupt the insurance market here), I was on a sabbatical that earned me, among other things:
- Working for a Unicorn (Hi Meitu!);
- Volunteering to teach old people to use technology (miss you, @CTNbayArea)
- Study at Stanford and meet awesome people like Sophie-Charlotte Moatti (that was my teacher there);
- And be the first international student approved at the Product School (that allows me learned a lot with Carlos G de Villaumbrosia)
It was a very intense period that gave me many learning and desires, among them to start something similar in Brazil (which became the Product Mobile Course, which already has more than 230 participants in 8 editions), see the Brazilian Product Community getting stronger and much more.
And all this was only possible after joining my years working with product and breaking a little to do the final project of the class, thinking of a new feature for Uber.
What is it?
Well, I’ve spent the last few years devoting myself to creating apps that solve problems and make life easier and cool. I even won prizes like Apple’s BestApps (2x), Tela Viva Movel (3x) and I was listed as one of the TOP3 mobile professionals in Brazil because of that. But I did not get time to get my driver’s license either.
But in the era of the shared economy, having apps like Uber, 99, Cabify, Easy, it is possible to live well (and very well) without the CNH (as we called the Driver License here).
PS: And there’s still VAH! (A sensational idea of Leandro, one of my students at EBAC) That’s like a Google of the trips, that searches the price in the category apps without you needing to open one by one and even provides a discount too.
However, there is that offer vs. demand that creates things like Dynamic Price, which makes prices more expensive depending on the time and demand. If you have tried to use a Uber in Rio Reveillòn or in full rainy Friday at 5:00 PM in São Paulo, you will remember that name.
And then painted an idea: A subscription so that there is no price variation at @ny time.
Understanding the business
To structure the idea, I first learned about Uber and his clients (including me) a little better. Starting with the company, I looked at some data:
- Available in 58 countries; (Currently: 75 countries)
- Available in 300 cities; (Currently: 450 cities)
-
More than 1 BILLION of rides performed; (Currently: more than 2 Billion)PS: the numbers are from April 2016.
And I ran a research on my network where I discovered that:
- About 85% of the respondents had already been impacted by the dynamic price at least 1x;
- 82% did not complete the race request at the same time that the dynamic price appeared;
- 20% make more than 3 rides/weekly.
Ps: Just remembering that the survey was answered in April 2016.
Not to mention so much negative review because of this…
I thought of some people too: Aquis (inspired by my little friend Aíquis Rodrigues), a young executive who had no wallet, no car, and use Uber a lot, every day to go home and get to his meetings. He was very disappointed when for some reason (such as rainy Wednesdays and pre-holiday Thursdays) he paid very different amounts than he was accustomed to paying.And also Fábio (the one inspired by the good stories lived next to my childhood friend and roommate, Fábio Nogueira).
Gathering some data, doing some calculations and correlating with the research I rode, only in Brazil, Uber did about 20.ooo races/day. Taking that 20% of cancellations are usually because the price is higher, Uber lost about 3,500 races/day, something around $1,500,000/month.
But as Uber’s mission is:
Give access to reliable transportation for all people, everywhere
I kept thinking: How to optimize the requests, since about 80% give up or wait a bit to request a driver when the dynamic price appears?
The answer is: Uber@nytime!
How does it work?
Bezos Approves!
Uber@nytime is a $ 9.99 subscription model (with the first free month) that prevents Dynamic Price from appearing.The customer continues to pay the price of the trip normally, but with the advantage of not having surprises (like dynamic price), besides priority when request rides.
Metrics. How measure it?
In all my years on the road, I learned that it’s no use just putting a feature in the air with nothing to analyze. That start even before it goes into development. What is your hypothesis? Question itself!And in this case, I categorized between Activation and Retention, the main items to validate:
Activation:
- Which % of users, use Uber@nytime when the dynamic price appears?
- How many users tried the service?
- What % of rides (even with dynamic price) requested after the service launched?
Retention:
- After the first free month, how many users keep as subscribers?
- Did Uber@nytime subscribers increase the average number of races?
And in another calculation, only in Brazil, putting that the average price of races is $15.00, Uber would have more $8,640,000.00 of money.
Based on the comparative, 40% more trips equals just over 46,000 rides/month.
And there we have some data:
- If 50% of customers who use, keep the subscription after the trial period, are more $230,000.00/month, and almost $3,000,000.00/year.
- If Uber@nytime subscribers increase their number of daily races by 12%, they will be $216,000.00/month and more than $2,500,000.00/ year.
And some other metrics that could be analyzed:
- Daily trips growth per user;
- Fall of the skip of the requests.
How to show this to the users?
Discuss ideas with something scratched, helps in setting scenarios and even see flaws in flows. So, since most users give up calling when the dynamic price comes up, use a call-to-action to take attention and it does not cancel, raising a little curiosity and activating the trigger to reward with information/promotion that can be useful and facilitating payment with saved credit cards.
Sounds Good. And how about launching?
Using a 360º strategy for disclosure can give much more visibility to features and help you acquire users to test them. And when we talk about it in the apps world, we have some things that can strengthen the disclosure, like:
- Release notes;
- Push notifications,
- In-app,
- Special offers based on user behavior.
And of course, uniting with PRs, Ads, Social Media, the leveraged can be much bigger.PS: It is essential to have everyone well aligned. That is, before giving start in the actions of disclosure, it is great to take care of this alignment, not forgetting things like the training of the support team.
Next Steps?
Test the model in other countries, test with Uber Business…
Conclusions
In that one year, Uber did the redesign of the app and even hide the display of the dynamic price, which generated a lot of criticism.
However, I think the validation of this work was to see some apps in the category create subscription programs (each with its particularities), like:
- Uber himself, who started testing a subscription model in New York and San Francisco;
- And Easy, which launched Easy Club, ranges from more affordable monthly racing plans to unique advantages with app partners.
But if it was already cool to make a fictional feature, I do not even tell you how cool it is to do it on a day-to-day basis and see all these things happening really! The product management world is awesome place \o/