If It Can't Be Said In One Sentence, It's Not Clear

The most important thing I’ve learned in my 2.5 years since starting Pristine is the power of clarity. Looking back, it blows my mind how unclear I was with all of our stakeholders - employees, customers, and investors - and even myself in my first year.


I see a similar lack of clarity when I speak with super-early stage first time founders. There’s a simple litmus test to determine clarity:


If it can’t be said in one sentence, it’s not clear.

So my advice to early stage entrepreneurs is to have a one sentence answer to all of the following questions:

  1. Besides raising capital, what are the 3 most important things that you need to accomplish in the next 6 months?

  2. What are are the ideal strengths of your first 5 employees?

  3. Based on the amount of capital you want to raise and assuming no revenue, how much runway do you have?

  4. Are you default alive, or default dead?

  5. What does your business do?

  6. Why does your business exist?

  7. Why are you going to be better than the next best alternative?

  8. Why won’t the next best alternative copy and kill you?

  9. How big is the market? Please provide a bottoms-up analysis.

Autonomous Cars Break Uber

As Bill Gurley (General Partner at Benchmark and board member of Uber) notes, Uber will be the dominant player in the ride-hailing business. Why? Because Uber’s business model is predicated on localized marketplaces of supply and demand for drivers. These localized marketplaces create strong network effects and “winner-take-most” markets.

The localized network effects are built on local liquidity of supply and demand for drivers and riders. This can be best summed up by this image that Gurley features in his post.

With this model driving Uber’s meteoric rise, tech pundits from Benedict Evans to Ben Thompson are asking the natural next question: What does the self-driving car mean for Uber?

The thing is…self-driving cars break Uber.

Indeed, with self-driving cars, we can just replace the “more drivers” element of the cycle with “more autonomous cars.” Drivers are intrinsically temporary. They are just people. They eat, breathe and sleep. They drive when they want to drive, and don’t drive when they don’t want to drive.

Autonomous vehicles are not temporary. Rather, they are permanent. Once they are on the road, they are available 97+ percent of the time to service riders (3 percent for gas, inspections, repair, etc.).

Autonomous vehicles break the “marketness” that makes Uber a market of drivers and riders. Supply will massively outstrip demand as vehicles become available 24/7 at dramatically lower marginal cost.

Autonomous vehicles will be much cheaper than human-driven vehicles. Today, humans are taking home 80 percent of the revenue (Uber the other 20 percent). Of that 80 percent, perhaps 30 percent is paid out for gas and vehicle maintenance, costs that autonomous vehicles must also incur. Thus, human drivers are taking home about 50 percent of the revenue they bring in; 20 percent goes to Uber, and 30 percent is the cost of servicing riders.

As Jeff Bezos of Amazon says, “Your margin is my opportunity.” Moreover, as Benedict Evans of A16Z notes, autonomous vehicles will be designed differently, with fewer features, thus making them even cheaper than cars that human drivers are using today to transport riders. With economies of scale, it therefore seems plausible that autonomous vehicles will be 60 percent cheaper than human-driven vehicles, even if Uber maintains its gross profit per ride.

Let’s return to Uber’s virtuous cycle. As outlined by Gurley, if we substitute “more drivers” with “more autonomous vehicles” and “lower prices” with “way lower prices,” it appears that the “market” breaks. You could argue that an aggressive company could accomplish the same task now by simply paying drivers a flat hourly fee, even if there’s no demand — but it doesn’t make any sense to do that because drivers are intrinsically temporary. Paying drivers now doesn’t mean drivers will service riders later.

On the other hand, paying for a car now, even if it’s underutilized, doesn’t mean it won’t be utilized in the future. Moreover, the marginal cost of a car sitting in park two miles from downtown approaches $0.

If it’s indeed possible to ever break Uber’s lock on the market, the shift to autonomous vehicles will be the disruptive force that enables someone else to win by beating Uber on asymmetric terms. Being first to market with the right vehicle will be paramount. Even a one-year head start could dramatically change the market dynamics now that Uber and its rivals have educated the market.

Google is best suited to productize autonomous cars for ride-hailing. Google invested in Uber and thus has access to useful, confidential information about the ride-hailing business; Google has the best mapping solutions in the world, the most advanced automated driving solutions and the first vehicles designed from the ground up to be autonomous. It’s hard to imagine any company that’s better poised to capitalize on this opportunity to break Uber’s impending monopoly.

However, as Benedict Evans notes, self-driving technology will commoditize. Given that Google is financing most of the R&D and their history with Android, it’s probable that Google wants to commoditize autonomous vehicles. At some point, it will not make sense for a single corporation to finance tens or hundreds of billions of dollars of assets. Rather, banks or public markets should finance these assets. It will be interesting to see.

P.S. Google named “Google Drive” a little bit too soon.

The Power of One

Over the last two years, I’ve learned an incredible amount about… well everything. Hiring, firing, raising money, selling, marketing, etc. But the most important thing I’ve learned is more meta than a specific functional area, tactic, or unique challenge. Rather, the most important thing I’ve learned is how to manage my time and the company’s priorities.

Jason Cohen from WP Engine does an excellent job explaining this idea with a specific emphasis on growth. But he’s right. A startup can focus on only one priority at a time. I didn’t absorb the magnitude of this statement until recently. But it’s really true. The business has to have a single, unilateral, unwavering focus, and everything in the business should be aligned around that. The opportunity cost of focus is tremendous. A single goal aligns the entire company and provides clarity to everyone.

What I’m proposing here isn’t novel. It’s just focus. But focus isn’t enough. Focus is ambiguous and soft. “One,” on the other hand, is concrete. There can only be one “one,” as the name would imply :).

Why Seed Stage Startups Should Join the Capital Factory Accelerator

Over the last few months, Josh and Gordon from Capital Factory have introduced a few startup founders to me to ask me about my experience as a founder that's gone through the Capital Factory Accelerator program. I'm not writing this post because they asked me to, or because I'm sick of talking to founders on Capital Factory's behalf (in fact, I would love to help out more if I can). Rather, I'm writing this blog post to provide written clarity to founders who are making a decision that will materially impact their baby.

But first, some history about my startup, Pristine. It's important to note our history and thus my bias in writing this. So here it goes:

Pristine builds software for Google Glass for remote collaboration in field service, training, and audit environments in life sciences, industrial equipment, and healthcare. As of the time of this writing, we've raised $5.4M, have more than 20 employees, and 30 customers. We are an enterprise SaaS company, with the caveat that we also deploy some unique hardware. The key challenge that the hardware presents is that because of it, we fundamentally change how businesses provide field service to their customers. Unlike most software companies, our customers cannot absorb 400 units in one day; they start with 5, then work their way up to scale.

My cofounder Patrick and I met in high school computer science class. Patrick and I started Pristine on May 15, 2013, a month after I turned 23, and couple of months before Patrick turned 21. Prior to founding the company, I was leading design and development of an electronic medical record (EMR) for hospitals. Patrick was doing freelance technology and data science work.

I started blogging on January 1, 2013 knowing I would start a company in 2013. My reasoning for blogging was simple. I had no money and no credibility. Why would employees work Pristine, why would customers buy from Pristine, and why would investors in Pristine if the cofounders were so young and inexperienced? We used blogging as a foundation to showcase to the world that we knew what we were talking about to raise money and attain our first customers (it worked; I highly recommend blogging). When Google announced Glass in February 2013, I immediately knew what I wanted to do - build software solutions for Glass in healthcare - and recruited Patrick to help me start what would become Pristine. Overtime we would change gears a bit, but we started in healthcare first.

Patrick began coding in May 2013, and I began looking for a beta customer and raising capital. We locked down our first $100,000 at the end of June, and raised another $125,000  by the end of July, bringing our total past $225,000 raised going into August. Meanwhile, I somehow (I can't remember) got introduced to Gordon in June 2013 and met him for coffee. I had no idea what Capital Factory was. He was intrigued by what we were doing and said I should meet Josh. So I came in and met Josh in early July 2013, and he offered us a spot in the Accelerator shortly after the first meeting. Apparently the application process to the Accelerator program has changed since then; I think it's a bit more structured now.

We actually considered not taking the deal for a couple of weeks. At the time, I had some doubts: I really thought the 2% equity-take was significant. I also didn't recognize the value of the Capital Factory workspace and network, or the signal that being in Capital Factory sends to prospective employees. I was really just full of myself and thought I could do it all.

But over the course of a couple of weeks, I spoke with some personal mentors. They made something painfully clear to me: the likelihood of failure at the "your product doesn't even work yet" stage is so high, you should do anything you can to maximize the likelihood of success. If you do fail, the 2% take doesn't matter because it's worth exactly $0, so just do what you can maximize the likelihood of success.

We joined on August 4th, 2013 right when our first hire, Mark, joined the team. Over the next year, we raised $1.1M in seed funding, and then raised $4.3M in our Series A led by S3 Ventures in September 2014. During the Fall of 2013 while working in Capital Factory, we grew to 9 people, all of whom worked out of coworking space. We were the loudest group in the room and we annoyed everyone else in the space; naturally, they all hated us. We moved out of Capital Factory on April 1, 2014 into our own space 1 block away.

Alright, so with my bias out in the public, here are all of the reasons you should join the Capital Factory Accelerator program:

1) Free credits for AWS, Azure, Rackspace, GCP, and more. These credits total more than $100,000 (upgraded to more than $500,000 since I joined). At the seed stage, every dollar is precious. Your mental state should be "holy shit, that's $500,000 for 2% of my company. That's a steal! That means my company is worth $25,000,000!" Obviously the valuation doesn't mean anything, but $500,000 is a lot of money at the early stages. This is reason enough. to join. I cannot conceive of any reason to turn $500,000 down at the seed stage. If you have an offer from Capital Factory right now, you should immediately stop reading this blog post now and accept his offer for this reason alone. I cannot overstate the importance of cash. Cash (equivalents) is/are king.

2) Access to the Capital Factory network. We did not go through the matching program by getting 2 partners to commit $25,000 each, although we tried. What we do is pretty unique among the Capital Factory companies and I think that, coupled with my brazen personality and arrogance, didn't resonate particularly well with the Capital Factory partners. We did however systematically go through the partner and mentor network. The indirect relationships that came from those meetings has generated several hundred thousand dollars of revenue for us, and will likely generate several million in the coming years as we grow. You should be as selfish as conceivably possible and meet with every mentor and partner. Going into each meeting, you should know exactly what you want - money, or a specific connection. Never ask for advice. They will give it regardless. Ask for what you need most: money and access to people who will invest in you or buy your product. The value of the network can be tremendous. Take advantage of it.

3) Work space. I know some people can work from home, but I cannot. Even if you can, that's no reason not work at Capital Factory. The workspace is phenomenal, and it inspires you to work harder. I loved working in Capital Factory. The workspace is a a place you should be proud to bring prospective customers and employees to. It's better than our work space today and we've raised $5.4M! Enjoy it.

4) Validation. Good luck hiring engineers as an early stage tech startup in Austin that's not in Capital Factory or Techstars. If you miraculously find an engineer crazy enough to work for a super-early stage startup, you'll have to convince them not to work for a Capital Factory startup and instead work for you. Good luck. Put Capital Factory on your side. It will help in recruiting.

And now for the downsides:

1) The classes weren't that helpful for me. The only thing that matters at the seed stage is achieving product/market fit. You should do anything you can to get there. You do need to "learn," to get there, though. I would instead recommend that you read like hell. The single best source of information in the tech startup community is the Mattermark daily email. If I could only read one thing per day, I would read the Mattermark daily email. It's absolutely the best source of tech information on the Internet by 10 miles. If you're running a SaaS business, you should also read SaaStr.

2) Capital Factory maximizes the personal value you can extract from your startup in the event of failure; if you fail but went through the Accelerator, you'll have a lot more friends and connections in the Austin startup ecosystem that you can leverage for your own future (e.g. work for a startup that is successful, identify mentors, make friends, etc). Do not discount this. Your odds of success are low; the least you can do for yourself is mitigate downside risk.

3) Parking sucks. Good luck.

4) Coworking sucks because you can't make it your own space, but that's ok. You're early stage. Deal with it. There are worse problems to have.

Recommendation: join Capital Factory. You won't regret  it. I think it's hilarious that for a time, I actually thought we would be fine without Capital Factory. In retrospect, I can say with 100% certainty that Pristine would have failed if we didn't join. So stop reading my silly blog and do whatever you can to convince Capital Factory to make you an offer. If you have any questions, you can email me at kylesamani@gmail.com. I'll be glad to help however I can.

What Infrastructure Do Manufacturers Need To Implement Servitization?

This post was originally featured on the Field Service USA blog.

This is the third post in a three-part series exploring servitization. The first post dove into the underlying problem that begat the need for servitization, and the second post dove into what servitization is and the impacts to manufacturers (OEMs) and their customers. This post will explore the infrastructure OEMs need to build in their organizations in order to bring servitization to life.

We’ll start with the implications of servitization on financial infrastructure, then dive into org structure, and lastly technology tools.

At the most basic level, servitization is about growing long term, recurring revenue while reducing short term business risk. More specifically, this is about moving from just selling capital equipment to selling equipment and services on top of that to maximize the value that customers can extract from a given piece of equipment. In the most extreme servitization models, this may even involve subsidizing capital equipment sales with service revenue, incurring a short term cash hit for even more long term recurring revenue.

The move to servitization has substantial impacts on cash flows. Traditional OEM cash flows tend to revolve around end-of-quarter as capital equipment purchases can take significant time to approve and typically revolve around quarterly budget meetings. Note that the numbers below are for a hypothetical OEM.


This stands in stark contrast to a more typical servitization based cash flow model, where revenue between equipment sales and service blurs and normalizes.


Although there is still seasonality in the servitization model, the month-over-month cash flow changes are tempered significantly. This reduces cash flow risk to the OEM, normalizes customer payments for the customer, and presents an opportunity for greater long term revenue capture per customer. Everyone wins.

Although every organization is different, it’s possible to paint broad strokes on the management structure needed to deliver servitized performance. The key to delivering servitized product delivery is organizational alignment. The lines between sales and service must blur: sales teams must learn, appreciate, and sell the value of service and customer-value-extraction. Service, on the other hand, must recognize that the company’s financial performance will depend on their ability to execute and ensure that customers extract the expected value from the OEM’s equipment.

If service fails, revenue will be negatively impacted. As a result of this interdependence, sales and service organizations need to work more closely together. These two organizations should report into a single unified head - with a title such as VP of Customer Success - whose two largest components of variable comp should be service delivery and sales, in that order. It must be clear from the top down that customer service is king.

Lastly, OEMs will need a new set of tools to succeed in a servitization world. Traditional field service management software will not be enough. Servitization is about customer service; as such, OEMs will need new tools to more effectively engage and support their customers. On demand support tools such as those offered by Pristine will become ever more important. So will tools that empower customers to diagnose and repair equipment on their own. These tools will become mission critical as OEMs won’t be paid if their equipment isn’t working as advertised. OEMs must find and implement the right tools so that their customers can service equipment on demand with OEM support and guidance.