“You can’t just extrapolate Google cars driving ~1.5 million miles under specific conditions (weather, topology, construction, traffic, accidents around it, etc.) to usurping the ~3 trillion miles/year under all conditions in the US. 1.09 fatalities per 100 million miles is the current non-self-driving numbers.
2014 had ~30k fatal crashes out of the 3 trillion miles traveled. We have to understand not how those crashes happened, but what makes the vast majority of them not happen. Luck is not a contributor, expertise is. Understanding human expertise is the key, not human frailty.”
Tech claims that security isn’t that big a problem and certainly not one that requires statutory approaches or regulation.
Two years ago Eddie Schwartz, vice president of global security solutions for Verizon’s enterprise subsidiary, said that self-driving cars will prove an irresistible target for hackers if they ever hit the roads.
Change if to when. Of course they’re irresistible; hacking and controlling a real car on a real road, with the potential of doing real damage, would be catnip to a large number of naïve kids (to prove they can), not to mention angry adults (getting even) and terrorists (creating chaos).
The cars aren’t yet able to handle bad weather, including standing water, drizzling rain, sudden downpours and snow, let alone police instructions (…) “I am decidedly less optimistic about what I perceive to be a rush to field systems that are absolutely not ready for widespread deployment, and certainly not ready for humans to be completely taken out of the driver’s seat.”
“When we designed the platform, three white guys, there were a lot of things we didn’t think about,” Chesky said to an audience at the conference. “There are racists in the world and we need to have zero tolerance.”
There’s no question that people of color, especially African Americans, have more trouble booking on Airbnb.
As of 2014 Hispanic’s spent $1.3 trillion, people of African decent $1.1 trillion, Asians $770 billion and Native Americans $100 billion.
That’s a whole lot of buying power.
There’s also no question that Airbnb has been slow to recognize/admit to the problem — as has the rest of tech.
In a serious effort to change, Chesky has hired former Attorney General Eric Holder to “craft a world-class anti-discrimination policy.”
“This process isn’t close to being over, but we want to be as transparent as possible along the way because I know we’ve failed on that front previously. I want us to be smart and innovative and to create new tools to prevent discrimination and bias that can be shared across the industry.”
Which makes it likely Chesky is serious, since guys like Holder don’t come cheap — nor are they easy to shut up or buy off if the company isn’t serious.
Hopefully it will help.
But it will more likely be a cold day in hell before anyone or anything changes racist MAP.
More recently, RMT (Riva-Melissa Tez,CEO @ Permutation AI and an active investor) wrote a superb post on Medium noting that Silicon Valley has lost its perspective on the difference between a ‘problem’ and an ‘obstacle’
— any obstacle that restricts our standard of living — is now framed as a problem. (…) Recognizing these obstacles or inconveniences and being able to avoid them are privileges — a special right enjoyed as a result of one’s socioeconomic position. They are perks…
I’m not a fan of a lot of AI, especially chatbots.
Most have speech patterns similar to human speech, lousy diction and rapid speech, which leaves most people with poor hearing our in the cold
And I find them relatively dumb.
Most of us have had run-ins with unhelpful customer service chatbots; the ones that are unable to respond to any but the most mundane quarries — which is why I usually just start by saying ‘representative’ until I get to a human.
I have no understanding why it is better to talk to your TV, rather than use the remote.
The first thing many of my friends and family do on a new iPhone is turn off Siri,
I know that many people love them, which is fine with me; whatever floats your boat.
An artificial-intelligence lawyer chatbot has successfully contested 160,000 parking tickets across London and New York for free, showing that chatbots can actually be useful.
That’s useful. And free.
Still more interesting is the fact that its creator is a 19 year old, with a history of using his skills creating tools for nonprofits, since he was 13.
I may be a digital dinosaur, but I’m not to old to learn and change.
Hopefully, this kind of usefulness is the future of bots.
And who knows. Perhaps by the time I need assistance the young developers will take into account the millions of hearing-challenged people who will be their biggest market, especially in healthcare and daily living.
And that favoritism usually results in more money, more introductions, more involvement, in fact, more everything, which results in substantially more innovation.
The data showed that companies tied to a competitor by at least one VC firm in common were indeed less innovative than those unencumbered by such ties; in fact, they were 30 percent less likely to introduce a new product in any given year.
It gets worse.
The UNfavored startups were 55 percent less likely to introduce a product.
Proximity mattered, too; those farther away from a shared investor were 56 percent less likely to introduce a new product.
What if your VC is part of the “golden circle?”
Companies tied to VCs in the top 25 percent of reputation indexes were significantly less likely to introduce new products in any given year.
And I’m willing to bet similar stats apply to super angels, regular angels, incubators and the rest of the funding world.
Arcade City Austin / Request a Ride is a Facebook group that has grown rapidly in the weeks following Uber’s and Lyft’s departures. The group, which requires approval to join, is currently populated by more than 33,000 members who use the group to find rides to and from their destinations.
Beyond that effort, there is Zipcar, getme, Fare, Fasten, Wingz, zTrip, RideAustin and InstaRyde riding into town (if not already there) and all willingly complying with the required fingerprint background check.
“So I say we are going to IPO as late as humanly possible. It’ll be one day before my employees and significant others come to my office with pitchforks and torches. We will IPO the day before that. Do you get it?”
Graham discounts the world, the people in it and innovation itself.
Kalanick plans Uber’s IPO with no consideration of the economy, competitors or the speed at which things change.
Graham’s words have already come back to bite him; Kalanick’s probably will, too.
I find it forever fascinating to try and decipher the minds behind the creativity that stretches the boundaries and adds unique beauty to normal, real-world stuff. Here are two wonderful examples.
It takes a rare mindset to see a utilitarian object, with its own shape and use, and turn it into completely different object with a totally different form and use. The beauty is found in the operational innovation, since each of the final forms looks totally normal.
Or the artist’s mind that takes something that’s been around for centuries and keeps it’s utilitarian properties, while changing it in ways so far beyond the normal decorative and stylistic features that it is almost unimaginable — except to that one mind.
Wouldn’t you love to share a meal (or a bottle of wine) and just talk? No agenda, no purpose, except to bask in the creativity that flows from a truly original mind?
For years, venture capitalists have been pushing hypergrowth over profits, at least though the initial phases of investment rounds. Investors told Lehmann to reinvest the company’s money in pushing more growth over building a sustainable business.
That advice didn’t go far with the Postmates CEO. (…) Lehmann argues that it’s the CEO’s fundamental job to have looked at the margins and made decisions early on.
“Companies that run for the last two years in hyper growth are now wondering how to make money.”
I completely agree — hypergrowth without a hope of unit economics that lead to profitability has always been a fool’s errand with precious few exceptions, and even those had their “come to Jesus” realization points that the investors were getting nervous and were anxious for at least a hope of a repeatable, profitable set of unit economics.
There has been a sense that pushing the bidding of sequential funding rounds at ever-increasing valuations would create a kind of de-facto “momentum” and crowd-out 2nd and 3rd and 4th place contenders, or at least amass a large enough war chest to drive pricing down as much as needed to push competitors out of the running (usually also by creating such a huge and dominant brand that customer acquisition in a noisy market is too expensive to make progress to catch up with the so-called leader).
This is ultimately as silly as the Texas and Miami and Las Vegas housing bubbles, that depended on “the next fool” to buy-in at a higher valuation, depending themselves on having a subsequent investor bail them out at a higher valuation, and so goes the escalator. The problem is, the escalator gets to the top at some point and there has to be a “destination” where value exists and with it, a hope of profitability.
The unsteady IPO market of last year and the continued bearishness of the IPO exit market this year has effectively called-out that “top of the escalator” and there are no more “next fools” (i.e. large enterprise buyers at the >$1 Bil level and no robust IPO appetite from capital market leaders that demand value and cash flow and a hope of profits).
So now, once again we are back to reality.
The great news about being back to reality is two-fold.
1) Sub-billion dollar valuations are no longer an “embarrassment” to VCs; and
2) Entrepreneurs can reasonably weigh a variety of capital structures that include bank and trade debt as well as investment equity and debt structures, all supported by revenue and that means free cash-flow.
With this in mind, the VCs and the investment community in general must start to become “reasonable,” because they are suddenly back in the traditional capital markets and will have to compete with other capital sources and structures for the hot deals.
Middle and nascent deals will have to become cash-flow generating, and for this reason they will also (wisely) become more reluctant to give up huge chunks of equity just to bring in working capital (at least not until the enterprise value pops to a higher tier by using bank debt, trade debt and other creative capital structures).
Savvy entrepreneurs and founding teams will also be less excited about creating an early and dramatic bump in valuation just to bank growth capital, because a down-round will likely wipe out a giant proportion of their equity. The giddy “we are a unicorn” has turned into “what happens in a down round?” reality check, that most people forgot about. Early venture investors have protected their downside with special preferred terms that founders and exec teams rarely consider or can demand. If this were real-estate, it might even start to look like over-aggressive venture investors that pump up valuations too early, only to have the market adjust to “reasonable” later, were “predatory.” It is an interesting parallel that will not be lost on founding teams, angel investors and early exec team members that hope to be rewarded via their equity stake.
The reticence to of many of the younger venture investors (those with fewer than 20 years of experience) having yet to bring in a 5X or 10X much less a Unicorn, to invest in early stage deals, is now balanced by the abundance of crowdfunding and syndicate fundraising at the seed and angel level. This is a great organic re-shaping of the investment and capital markets in favor of the early stage company and entrepreneur.
There is also a growing recognition that the early stage deals that do get picked up by venture investors have been in a long slow decline and “narrowing” of deals to known insiders and repeat successful (i.e. “brought a good exit to a venture fund’) founders. I think that this is largely common sense (bet on the horse that won the last race for you), and also based on the reality that it is a rare and elite breed of entrepreneur that can see an opportunity and execute a successful solution. That said, a close examination of the venture deals that have been funded in favor of known founders pales next to the stats behind the successful new ventures that have been founded by first time startup teams. The difference is largely that part of the value-add from the venture investors is the addition of those “experienced” startup executives onto the exec team as soon as the big money comes into play. Thus the risk of execution is somewhat reduced.
What does that mean to today’s startups? It means that the old concepts of cash-flow, repeatable and scalable selling and service delivery models, the idea of managing customer acquisition, retention and lifetime customer value, are again in vogue.
As they should have always been. While there will continue to be many good reasons for companies to temporarily sacrifice cash flow and profitability for raw user or customer growth, the days of “just get 1 million users and we’ll figure out how to make money later” are – at least for the time being, gone. And we celebrate that.
Unit economics always wins. This goes back to the days of “the lemonade stand” cash-flow exercise. It’s what built the world’s greatest capital markets. And it will always remain the best place to start. Water, sugar, lemons, cups and napkins. And a sign and a cardboard box. “How many cups of lemonade must we sell at what price to pay for the supplies, time and sign?” Simple. One does not need an MBA or to be a dropout PhD candidate to start with those basic principles.
In another parallel with the real estate (mortgage) market, today’s startup teams should be asking themselves the same questions that prudent investors will be asking them (kind of like the new mortgage market, where everyone has to go through “full documentation” to get a standard mortgage loan):
How can I make money? How can I do it at scale? What is my selling process and is it repeatable? Who will pay for my service or product and what will they pay, and why? How much money do I need in working capital to find my perfect product-market fit and establish the right selling model and price point/margin? What are the unit economics of my business? What drives retention and churn? What prevents others from copying me and disintermediating my base? Is there a brand value that creates loyalty, or is this market driven by other values and factors? What are my logical exits? Who are the logical acquirers? Is there a realistic IPO path?
Yes, we are back to reality. It sucks for some people. And that’s okay. Those people should get with the program or get out of the startup business. Disrupt and question everything. Be bold, revolutionary, even bombastic and disrespectful of the incumbents and status quo. But don’t ignore the fundamental rules of business that underly the path all companies must tread to go from small to large, and startup to profit and successful exit. After all is said and done, you have to make payroll. Sell to a customer a second time. Own a brand people love and trust.
Reality only sucks because it makes you work harder to win, and forces you to confront inconvenient tasks and difficult questions. Short cuts are nice but when they don’t work you end up falling off of a cliff. Better to work harder than run headlong at a cliff you can’t see coming.
Yesterday I described how I came to start my company Quant Price. Today is the story of what happened after that decision.
I first did the Founder Institute program that helps entrepreneurs create companies. They introduced many concepts that are hard for first time founders to grasp. Things of HUGE importance ranging from talking to potential customers to validate the idea to seeing how buying decisions are made.
There are many things that can make a startup fail. Each of these factors are risks. A successful startup has avoided each of these pitfalls in succession.
There is a step by step process to developing the idea and de-risking it. The founder Institute program helped me think through them.
One shortfall of the program was that it has a cookie cutter view to creating startups. It requires that everyone fit into a mold. This may not serve every startup well. I’d probably have taken it to completion if I were not “terminated” for having a hard time getting a sufficient number of video interviews.
Though to be rational, I’ve to admit that it’s also possible that I just have a very bad idea or am not really cut out for this. I may only know very late because “I’m drinking my own Kool-Aid”.
The risks are huge. There are a lot of things that matter to a startup being a success. There has to be value in the product. There has to be a way to reach a big chunk of the market. There has to be a way to convert all those gains into money. People have to believe in the dream and be willing to contribute for free for the duration that the cash flow doesn’t exist. It takes a lot to keep a startup together.
It was very hard to find my first customer. I didn’t even have my marketing material when I went to interview Katherine Krug the founder of Getbetterback.
Partially because I didn’t realize that I was supposed to bullshit my way through such meetings and, frankly, partially because I didn’t know how to bullshit about it if I had to.
Fortunately, Katherine is one of those people who tests things like prices. I think she was lucky to not have a mentor who had a strong opinion about pricing. (I still can’t believe that people don’t test their prices.)
Katherine first did a price test with Optimizely. Unfortunately, Optimizely is not really geared to do price tests. Additionally, putting Optimizely code on your web page makes it load a lot slower for the end consumer. That itself reduces conversion rates.
It made me wonder if people don’t know or just don’t care about the impact this has on the browsing experience.
Optimizely targets conversion rates, so once Katherine was done with the test it told her how many people had converted to buy at different prices. Obviously at the higher price, fewer people converted.
But the real question was which price lead to better margins?
And was the test significant?
That is when she realized she needed Quant Price.
I salvaged the Optimizely results to guess what the next prices for the next test should be.
We used our pricing engine and did an A/B test. It turned out that seemingly identical prices $49 and $59 were over 26% different from a revenue perspective.
We also realized that changes to the web design and the holiday season could change the optimal price. So much so that at some point $69 was 22% better than $59 because of the Christmas buying spree.
Hidrate, my second client, was easier to find after we put up a video interview with Katherine.
With Hidrate, we seemed to run into some bad luck. The $59 price for them was identical to $49 from a profit perspective. But as luck would have it, they were running out of inventory. Now, at the $49 price they were burning through 45% more inventories to make the same amount of money as at $59.
Once we told them about it, they raise their price to $59 to save inventory. They kept their customers happier for longer AND made money to buy more inventories without losing the company to investors.
We did all this by using technology developed for finance. Quants in the financial industry are usually tasked with pricing securities worth trillions of dollars based on market data. There are multiple factors that affect the “fair” price of a security.
I learned how this valuable task of pricing can be automated incorporating available data.
Many companies now quote prices and sell exclusively online. Computers can be used to make pricing decisions based on price sensitivity of individual customers, business specific parameters, such as inventory availability and market conditions and the current season.
Airlines and Amazon already do this on a massive scale and are very profitable as a result.
SaaS companies and large-to-medium scale retail operations could be next in line.
Fortunately, our market is easily defined. We are looking to help companies with more than 300 distinct customers. This is desirable because with more data we can get statistical significance on more complex models.
We are looking to help SaaS companies that want better pricing models. For them, we have a solution that reduces churn and increases revenue at the same time !
To justify building a company-specific solution, we need to be able to service a pool of revenue of over 10M$/year.
Quant Price’s free app Qbot is available to smaller companies in the Shopify app store.
In order to make our technology more accessible we packaged it into APIs and apps, that help us expose this functionality to more companies.
After I got my Masters in Computer Science and Optimization from USC, I worked for banks for a while as a quant (people who do quantitative or mathematical stuff such as build complex models to evaluate financial securities, risk and reward).
Contrary to what most people would tell you about working for a bank, I found the work very interesting—not surprising, given my math background.
But much as much as I enjoyed my work, I was bitten by the entrepreneurial bug and wanted to do something original.
I’m an engineer at heart and I like inventing things that makes life better or improve what people are already doing.
Innovation seems to happen overwhelmingly in small to medium companies and then big companies usually buy the promising smaller ones.
Doing my own thing would also gave me the ability to try out new things at an amazing pace that makes the process of discovery a lot faster.
Initially I had the bright idea of trying to beat the banks at what they do best: price securities.
I invented a way of making models that could learn more from purchase data about pricing than conventional models do and put those models to work trying to outguess Wall Street computers.
Eventually, I realized that even though my algorithms were smart, the networks that were affordable to me were too slow to use this information for the strategy I developed.
Even though I could predict prices, I couldn’t get ahead of computers that were closer to the exchange to make the profitable trades.
I had to make a pivot or bet even bigger and buy access to the necessary networks.
I paused to think for a while and it occurred to me that the world is full of things that need to be priced.
So, why stick to the business of securities where so many quants and fast computers were concentrated in solving a problem with a lot of history?
Why not solve a Main Street problem?
I began looking for a niche where there was a significant problem that I could solve.
About that time I heard a story on the NPR money podcast.
Two sets of people were asked to guess the weight of a cow shown in a picture. They first asked a bunch of experts what the weight might be and each gave a different answer. They also gathered answers from a crowd of non-experts on a website.
The median value in the crowd of non-experts was much closer to the true weight of the cow than the group of experts!
That led me to think about how the wisdom of crowds could be used to help small companies make better decisions.
Of course, I wanted to monetize the information in the buying decisions that people make.
How about learning the perfect offer to make to shoppers at an online store?
I’d recently built complex pricing models to value financial securities, so I knew I could do this.
I quickly noticed that small businesses were leaving a lot of profit on the table. They were essentially using rules-of-thumb, instead of measuring price sensitivity and making optimal offers for peak profit just like the airlines, Airbnb and Amazon are already doing.
How much of an advantage can optimization bring?
For example, the average retail store runs with a 5% net margin. What if the store could raise the sale price (after any discounts) of the average item by 1% without affecting demand?
That would raise their margin a staggering 20%! And that is what happens with just a 1% change.
Imagine what can happen with a larger price change.
In our experiments with stores on Shopify, we noticed that they were so far off the optimal price that they were making less than 70% of what they could be making if they could tune into the crowd.
That is beyond a big deal!
So I began working on an app with the vision of using optimization to create better offer management strategies.
Over the last year I created the company, Quant Price. You can try our first app for free on Shopify.
Join me tomorrow for a closer look at what Quant Price does.