[elelegido]

Just like a jilted ex-lover desperate to get her honey back, Uber keeps trying to get me to return. They're claiming $34/hour revenue. Well, they call it "earnings", of course.

There has been a distinct lack of post-your-boast threads about SF Bay area lately, and lots of people saying that business is slow. So, is this true or more Uber fiction?

 

[LoLo SF]

On occasion, it is true. But their earnings estimate is based on "booked time" and does not include your "online time" waiting for trip requests. On average, I'm booked 82% of my online time. That Uber's claim of $34/hr earnings is really $27.88 minus expenses which average $7/hr, without considering the self-employed tax.

Now mind you, I am mostly driving the best times, so during slow times earnings are barely minimum wage. If you're driving a sh!tbox hybrid, that's hit the bottom of the depreciation curve, you'll do a bit better. So there you go, Uber's statements are always a bit short of reality.

 

[elelegido]

I don't know if it's booked time or online time. The Uber propaganda above says actual earnings depend on number of trips given, but if they're quoting booked time then the number of trips given wouldn't affect the average.

So who knows. In the absence of any evidence from drivers to prove otherwise, I'll go with it being Uber fiction.

 

[ObeyTheNumbers]

You can make a lot, just not take a lot.

Uber and its drivers are cash flow businesses, just about everything coming in goes out to others. 😞

 

[NorCalPhil]

Median, not average, meaning just as many drivers are below that as above it. They don't split it up between service levels. Without divulging the data set, using the Median is an easy way to obfuscate reality - it may be the middle earner, but it may just be a number thrown out there knowing there is no way anyone can verify. If they used average, 5 drivers could get together and crunch their numbers and gauge how accurate it was.

 

[PukersAreAlwaysYourFault]

Everything is a marketing tactic. They'll look at their top earners and advertise to all the ants that it's a possibility they could earn that much per hour just to get all the suckers out on the road. Uber needs supply to outweigh demand, always and forever.

 

[Robert Larrison]

Everything is a marketing tactic. They'll look at their top earners and advertise to all the ants that it's a possibility they could earn that much per hour just to get all the suckers out on the road. Uber needs supply to outweigh demand, always and forever.

From a lifer in the peeps moving things/peeps around biz
A dispatch needs 10-20% more drivers to cover spikes in orders

 

[og bunky]

I'd say $3.50/hr is average

i wouldn't drive in sf if i were you.

stay in Manteca, Fresno, Bakersfield, Sacramento or whatever

 

[elelegido]

If they used average, 5 drivers could get together and crunch their numbers and gauge how accurate it was.

Disagree. The 5 drivers may not be representative of the driver pool. Reason: too small a sample size. They could be very much better-performing drivers than the median or average driver and earn more, or not as good as the median and average driver and earn less. It would not be reliable to take 5 drivers' earnings and assume that their earnings are representative.

 

[elelegido]

I'd say $3.50/hr is average

i wouldn't drive in sf if i were you.

stay in Manteca, Fresno, Bakersfield, Sacramento or whatever

No, the closest one of those to me would involve a 90 mile commute, each way, from Marin. It wouldn't be worth it, especially given the lower rates in those places compared with SF.

Did you commute out to these places to work them while you were driving rideshare?!? I hope not!

 

[og bunky]

yes all the time and i loved it

 

[elelegido]

yes all the time and i loved it

Well, that certainly could be why you were making $3.50 per hour!

 

[AgentSmith]

Uber’s new trick (they never seem to run out of them); when there’s a surge and I log on while hovering on the exact spot it drops the surge in half. Happened all night… the trolling is never ending. Shameless MFers.

 

[og bunky]

the app doesn't update the surge map real time when offline. there are many ways around this. you're doing it wrong

 

[NorCalPhil]

Disagree. The 5 drivers may not be representative of the driver pool. Reason: too small a sample size. They could be very much better-performing drivers than the median or average driver and earn more, or not as good as the median and average driver and earn less. It would not be reliable to take 5 drivers' earnings and assume that their earnings are representative.

You can disagree all you want but that doesn't change the case. Multiple drivers will give you an average. Of course its a small sample size. Get 10 drivers, or 200, or 2000. Uber purposefully uses the median because you can't gauge its accuracy without having their full dataset whereas the average would be useful.

 

[AgentSmith]

the app doesn't update the surge map real time when offline. there are many ways around this. you're doing it wrong

More unhelpful advice… unfortunately no one reads this forum so I can’t confirm the recent changes.

 

[elelegido]

You can disagree all you want but that doesn't change the case.

Regardless of who agrees or disagrees, a sample size of 5 drivers is too small to give an indicative average or an indicative median. For more info, read up on sample sizes and confidence levels.

Multiple drivers will give you an average.

Multiple drivers will give both an average and a median

Uber purposefully uses the median because you can't gauge its accuracy without having their full dataset whereas the average would be useful.

You're confusing accuracy with representativeness, i.e. variance and the uniformity of distribution of revenues. If the data given by Uber is the correct data, then it is accurate by definition.

What you are trying to say but are unable to express correctly is that you believe that the representativeness of a median is lower than that of an average, but we can't know that in this case without looking at the actual revenues of all drivers. As far as the usefulness of the median vs the average in terms of variance and the pattern of distribution, generally the median is more useful for elements such pay as it lessens the effect of outlier results at each end of the frequency distribution.

 

[NorCalPhil]

If the data given by Uber is the correct data, then it is accurate by definition.

Do you trust UBER to give accurate data? The reason they use the median and not the average is BECAUSE you can't prove accuracy of a median without the data set. They could say the median was $40 an hour and nobody can disprove it. Your assertion of what I'm trying to say is incorrect. Please learn to read.

 

[elelegido]

Do you trust UBER to give accurate data? The reason they use the median and not the average is BECAUSE you can't prove accuracy of a median without the data set.

One can't prove the accuracy of an average, either, without the full data set, lol. In terms of being able to tell if the numbers are accurate or not, there is nothing to differentiate median from average.

They could say the median was $40 an hour and nobody can disprove it.

Equally, they could say the average was $40 an hour and nobody can disprove it either. If Uber's goal were to give inaccurate data, they could just as easily give an inaccurate average as an inaccurate mean, and the inaccuracy would be as equally indiscernible without the full data set.

 

[NorCalPhil]

One can't prove the accuracy of an average, either, without the full data set, lol. In terms of being able to tell if the numbers are accurate or not, there is nothing to differentiate median from average.
Equally, they could say the average was $40 an hour and nobody can disprove it either. If Uber's goal were to give inaccurate data, they could just as easily give an inaccurate average as an inaccurate mean, and the inaccuracy would be as equally indiscernible without the full data set.

If they said the average was $40, and you get 10 drivers together and find their average is $25, then yeah you have a good idea the number is BS. With the median you can't test at all. You obviously don't get it. Moving on.