Most of the big social media and app companies are pretty light on hard-core technology. Happy to stand on the shoulders of the tech giants that came before, many focus instead on features, design and UI. This enrages the kind of hardcore math nerds that used to rule the Valley.
Well, they have a new geeky mascot: Uber. Uber only scales and survives with hardcore mathematicians on staff. Among its braniac hires are a rocket scientist, a computational neuroscientist and a nuclear physicist. (That’s an actual staff photo to the left.)
I have no idea what those disciplines have to do with predicting cabs arrivals and sorting cab inventory. But apparently, something.
A new chest-thumpy blog post shows that using Google’s ETAs for Manhattan cabs was leading to horrendous wait times for riders, about 3.6x off the estimates. That’s pretty much the worst possible user experience for a first time Uber user, particularly in a city where cabs are plentiful and users may never give it another try.
Uber dropped the Google API like a hot potato and developed its own algorithm. It wasn’t particularly comfortable about this, because it didn’t have much historical data to go on. But as some graphs in the post show, it immediately did better. How much better? Their quants crunched some numbers for me and found that Uber is on average 186.3 seconds more accurate than Google. On average, Uber’s ETAs were 42.50% more accurate than Google’s. And with every ride, Uber gathers more data and the estimates get better.
Do a few minutes make that much of a difference when you’re waiting on a cab? Well remember, this is the average. In some cases the differences between Google’s ETAs and Uber’s ETAs was 15 minutes or more. And if you’re standing in the rain waiting on a cab, hell yeah 186 seconds matter. Given that Seattle is one of the cities next on Uber’s launch list, this is a valuable algorithm to get right.
Uber CEO Travis Kalanick and I talked backstage at Disrupt about how his company lives and dies on its “Math Department,” as they call the team in house. The video is below. (We talk math at the four minute mark.)
Hi, Sarah Lacey backstage at disrupt with Travis Kalanick, or as I like to call him Conan T-bone.
Old school. It's at Travis K, now. Good story on that on Conan T-Bone. I still have it.
You still have it? It's waiting in the wings? I have Sarah Lacy even though I still use Saracuda. See I didn't sell out like you. I kept the original twitter net name.
I kept, I have the original. It's just for close friends. More risky.
More risky to go with the pink socks.
There we go.
So in the early days when you were Conan T-Bone, you were sort of this roving around angel and adviser. You know, I kind of thought you'd be this sakar, glad kid, running around not doing a lot. And here, you find this hot company and you step in and you become CEO. What made you want to really go for it again?
Basically what happened was, I had done ten years of up your startups. I had been sued for a quarter trillion dollars. I had gone through and did anotherthing called red swoosh, and tried to, you know, go the other way and go on the light side of P to P. When 5 years, was 5 years too early to market and then things start happening for those years, I didn't pay myself a salary.
So when I finally did sell off, I needed to like chill the frick out.
Wild card said worse.
I had to chill the fuck out. And so, I angel invested for a year and a half or so.
Was it hard were you just dying to get back in there or you tired ?
Well at first I was and then it was like OK I need a, I like to say, money will not make you happy, but it will pay for therapy. And so I just had to go through some shit.
That's my theory of working at TechCrunch, right there, in a nutshell.
I feel you. And so actually it was Garrett and I started talking about Uber, like way back in 2008 in the web. We started talking about. We were, we were throwing out ideas Aha.
All kinds of ideas. Actually a lot of them, sort of along the same line as far as brand, but he was doing some, he was was telling about some transportation stuff I was talking about, sort of other types of experiences. We just got this bad boy started.
Right.
And when we started it we're like, 'This is a limo company.'
Yeah.
And we're like, Derek and I are looking at each other and we're like...
Who takes limos?
'You want to start a limo company?' Or, sorry. We wanted a limo company in San Francisco. There's no way to get around.
Right But it was like, 'Do you want to run a limo company?'
Yeah.
Like, and I don't want to run a limo company. And so I incubated it, built the team up and after my recharge mode had happened and it became really clear to me that this was a product and a tech company. It just was the right match to come in and run it full time. And so that's pretty much how it went down.
So you're happy doing it again or did you get too old in the interim?
No, and that's the thing that's interesting. Some entrepreneurs, and I was one them, I was scared. I was scared. It's like an artist who thinks like at some point they're too old to do their art or to bring it and and I would look at like Woody Allen films and I'd be like, he's still got it!
I'm like, yes you can still do it.
Grandma Moses.
And I'm like, it was the same thing.
When the passion took over, I'm like, I'm better than I was.
I'm more intense; I'm more awesome.
I think the difference is, is that in the last one, I was afraid of failure.
Now I'm not afraid anymore.
So now I can just have fun and go kill it.
And you're having fun, you're challenging everyone.
Definitely.
I mean, there are so many things I love about UberCab, and the first is that...
Without the Cab.
Yeah, Uber, sorry. ..that I love about Uber, the first is that your're doing something in the real world.
You're disrupting real world, which to me it's the whole next wave of the Web.
I mean, I think all of the other things we've done up until now...
Yeah.
have been necessary in sort of laying the foundation.
But there's so many real world problems.
And taking on something like this is so ballsy.
And I think in addition to that, there is a lot of real hardcore math technology behind this company. And that's something we have not been seeing in Silicon Valley. This whole wave of the web, has been more about UI, design, vision, features, not about the hardcore. So we're standing on the shoulders of giants and building on what has been built.
So tell us a little bit about, for the geeks out there, what it takes to make this company work.
I know. I think that's a great question. And so, the high level is that we sort of look at it as a mix of UI and experience with sort of hardcore math. And what that means when the rubber meets the road is that it means efficiency with elegance on top. That is the wow experience of Uber and so where the technology comes in is that, we could put a thousand cars in San Francisco, and very quickly go out of business.
We need to actually predict what demand is going to be and then make sure there is the right number of parts out there every hour of the day. But you can't just say, "Okay this is what the demand is going to be, let me put out those cars." You actually have to position those cars. And so you basically got a moving heat map of where we expect demand is going to be.
And then we have what we call 'anti-heat' which is where those cars are at that moment in time and so the residue heat is under-served areas. We need to be able to dynamically respond to that kind of thing. So, everything from the demand prediction side to supply matching to supply positioning. And then you've got spikes, like rain, or shift change, or things like this.
Dynamic pricing is part of the equation as well. So here's just a ton of math which basically make sure that riders get a car in five minutes.
And making that elegant experience is very, very hard from a mathematical perspective. But once we do, once we have a huge network in a city, and huge efficiencies, and the pick up times are low, the efficiency is high, or the utilization is high, it's very hard for somebody else to come in and break that.
I have a reverse testimonial, for anyone who still is not a believer in using Uber Cab. Cause most people say, "Oh I got picked up, it's great, it's great, it's great. So time, I didn't use it. So my husband and I are going to the airport to go to Nigeria.
I'm sweating to here how this - oh Nigeria?Yeah , yeah.
No, this time we didn't take it and we should of.
We took an Uber Cab to the airport and we're like, this is a great experience.
About $20 more than we pay to get to the airport.
So this is good, but, you know, we don't really love nice cars, we're like we'll just .
Yeah.
So on the way back, we're like, let's just grab cab instead of calling.
There's this line of cabs at the airport.
This is a time when you would never use Uber.
There are cabs right here.
Why would we call a cab?
We get in, my Blackberry, my precious Blackberry is a like sitting in the pocket of my backpack.
I throw it in the trunk, pull the bag out, Blackberry's gone.
Second I step out, Jeff says, "Do you have your phone?"
I realize it's not there.
The cab's pulling away!
Write down the number, write down the license number. Call Yellow cab, say, 'Just left the house.
Blackberry.
Will give him a huge tip if he comes back."
"Never got the phone.
Never got the phone."
No way to get in touch with him.
That's right.
Had to call Yellow Cab everyday for a week and a half.
Never got it.
Never got it.
Wouldn't have happened with Uber.
No.
You know the driver.
It just wouldn't.
You would call him and he would come back.
We know the driver. We saw the route that was taken.
Drivers have star ratings.
It's all sort of a centralized reputation system there is no way that could not have been a bad deal.
So it's accountability.
It's accountability.
Not just convenience. That's what we learn the hard way, so, never again. We are the biggest dyed-in-the-wool customers now. TechCrunch is going to have to give me a raise to afford that extra twenty dollars for every cab ride.
I love it, this is good.
So next city, are you going to tell us?
Well, we have four cities on the short list right now that we are basically hiring in right now. So it's Seattle, it's DC, Chicago, and Boston. And where we basically get a general manager for each city, similar like maybe how a hotel has a general manager, for instance, they have to run the operation of the business, but also grow the top line.
And so, wherever that general manager comes in first is going to be our next city. We're spending a lot of time in Seattle right now, we did a happy hour earlier this week. Rain is a big deal. I think it rains there 200 days a year and our virality, like as far as how this spreads and sort of word of mouth.
We're pure word of mouth. We're old school word of mouth. Our virality doubles when it's raining. So one of every three trips, we get another registered user, and registered user means They actually have our credit card on file. When it's raining, its one and one and a half; so we get another registered user every one and a half trips that happens, because people need to get...
And it'll be a bigger math challenge because of the intensity of the rain. Alright, we've got to wrap. You've got to get to meetings. Thank you so much for joining us, Travis.
No comments:
Post a Comment