If I was Product Manager at Uber
As any other Silicon Valley Professional, I keep coming up with next Unicorn ideas which mostly end up with my wife rolling her eyes.
So, instead of sharing this billion-dollar idea with my wife, I thought of sharing it on Medium with a bigger audience, at least I can’t see them when they are rolling their eyes.
I like Uber. And so do millions of people who use it. It has brought the Sexy Back in hailing a cab.
Uber has lot of smart people and they regularly talk about all the smart things they are doing at Uber on their Engineering Blog. Most probably, some Product Manager at Uber already came up with this idea and got axed.
But since I don’t work there, though I would love to, I can safely present it here. As my kids will say here, nana nana boo boo.
This is all inspired from the #MeToo movement happening around us and now gaining steam in India too.
Let me put on my Product Manager hat.
Persona & Problem Statement
Who is the target customer for this feature and what problem are they facing?
Women who avoid Uber rides for fear of getting sexually assaulted reading all these news articles.
Alternatives
What are these women doing right now?
They are opting for taxis or limos where you can request a women driver or asking for friends to pick them up.
Competitors
Who are the other ridesharing companies targeting these customers?
Value Proposition
If Uber also provided this feature of matching riders with drivers of their gender preference, it can get back these customers as it has upper hand in every other aspect like convenience and reliability.
Assumptions
We are assuming that riders really want it. This seems obvious but we still need to test whether they will be motivated enough to select this gender preference because from the usability perspective it will mean waiting longer as drivers of their gender preference might not be closest.
We are assuming that drivers would want to be matched based on gender preference of rider. This is a riskier assumption compared to first one. Some women drivers might not want to be matched like this as they will fear that some men may abuse this feature and request women driver to sexually assault.
We will be creating a new problem instead of solving the old one. Uber has strict screening for drivers to know their sex but can it give the same guarantee about riders?
We are assuming that it will be legal to sexually discriminate by providing this feature. One ridesharing company trying to enter this market seemingly had to shut down because of legal reasons.
This will be the riskiest assumption but easiest to test with internal legal team. I am sure Uber must have a big legal team now with all the sexual harassment lawsuits.
Assuming we get a green signal from the legal team as we are walking the fine line as Safr and not going all the way as Chariot for Women, its time to move to next step.
MVP
What’s the Minimum Viable Product we can build to test our hypotheses and assumptions without going all-in?
Uber sends an email after the ride for us to rate the driver and tip. It can add a feedback question at the end asking the riders if they will be interested in a feature like this even if it led to longer wait times.
Similarly, it can ask the drivers in a survey or during their registration if they would like to be matched as per gender preference of riders.
Product/Market Fit
If less than 20% of women drivers say yes to this, then its not a viable product as it will be difficult to match women riders to women drivers with so few options as only 14% of Uber drivers are female.
Similarly, if less than 20% of women riders show interest in this feature than there is not enough market for this feature.
If we get results somewhere in between, we can continue experimenting to see what riders and drivers actually do when its part of the product.
We can provide a option to specify driver gender preference right above the options to select type of Uber service with a note that it can lead to longer wait times. It will have options of Male, Female, Don’t Care with default as Don’t Care.
We can measure how many women riders actually select Female. We can even increase wait time of selected service by a fake 50% to test their motivation level. We can measure if they change the option back to Don’t Care after seeing the increased wait time.
It will not be a working feature yet. We can just show a message on selecting anything other than Don’t Care.
Thanks for showing interest in this feature. We are still working on it and will be available soon.
We can reset the option and wait times after showing the message.
We can do a similar experiment with drivers too. We can send women drivers a fake alert that a rider in their route has requested a female driver and measure whether they accept it or not. We can show them a similar message if they accept.
We are just testing this feature. Thanks for showing support.
We can continue iterating with our experiments to require riders to upload their DL or some form of identification in their profile to be able to use this feature so that Uber can confirm sex of rider for driver’s safety.
We can take this even further requiring riders to post selfie of them through the app just before requesting the ride to make sure no one else is using their phone to request a ride.
These last two experiments may force some women riders to change their mind about using this feature as it requires more effort from them. But that’s what we are trying to test. Their motivation level. Are they concerned about their safety enough to go through all this? Maybe, most of the women are not that concerned about their safety in Uber if they still make up 52% of Uber users.
But on the other side, maybe more women drivers will accept to be matched to a rider based on gender preference if in the alert that we send them we add that we have confirmed the rider to be a woman.
Maybe, this will pivot our feature from its focus on solving rider’s problem to solving driver’s problem. Maybe, only 14% of Uber drivers are female because they are concerned about their safety and this feature will make them sign up in more numbers.
Like the first experiment, we don’t really have to build any functionality to support these two experiments. Just some options with messages to measure their interest.
If the desirability numbers still look good enough to prove this feature as a Product/Market fit, we can persevere and actually build the feature into the App.
But if the desirability numbers don’t look good enough to be a viable feature or it becomes unfeasible because of legal reasons, we can just leave it here and take solace in the fact that at least our intention was good.
Conclusion
Regardless of whether this feature ever makes it into Uber, I hope one day I will be able to send my daughter in Uber without any fear.
There is time. She is only 4 now.
