Patrick McKenzie (patio11)’s First AMA about Japan, Technology, Business and Startups on the HN Tokyo Slack Community

Jay Winder:

Hi there!

Firstly, I’d like to extend a huge, warm welcome to Patrick McKenzie (patio11 on the internet) for kindly doing this AMA, and in general for being such a frequent contributor of knowledge, advice and interesting perspectives. For people who’d like to know more about Patrick McKenzie, please check out:

Twitter: https://twitter.com/patio11
Comments on Hacker News: https://news.ycombinator.com/threads?id=patio11
General Patio11 Overview and links to blog posts: http://www.kalzumeus.com/start-here-if-youre-new/

Patrick:

Howdy folks! Feel free to start asking questions whenever, I’m here.
I’m also happy to take questions in Japanese / 日本語でも構いませんので、遠慮なく聞いて下さい。


Today’s Q&A


Question from Sacha Greif

Q. If you were to start a new business from scratch now (assuming you wanted to do that), what would it be? What cool/interesting ideas have you daydreamed about pursuing recently?

Patrick: 

I think about the question of what business to start next reasonably frequently, partly because I think I’ll always be an entrepreneur and partly because my job at Stripe is making things to make entrepreneurs’ lives easier, so I’m always trying to situate a mental me somewhere in our product discussions. The thing I’ve been kicking around for the last ~3 years is SaaS Pricing Pages As A Service.

The core insight is that software people suck at pricing and that I am slightly less abominable. The first thing I’d deliver is a consulting offering where I’d charge $2.5~$5k a month to just own pricing and packaging for SaaS companies that were at material scale (~$100k a month or so). I’d sell that to 4~5 people in parallel. Then, while providing the service, I’d be building up technology on the backend to e.g. manage their Stripe accounts, extract per-plan churn rates, build up lists of people who were close to an upgrade point, etc etc, initially with the goal of making the service offering more effective. I’d then put a frontend on that and sell an offering which was materially similar but much lower entry point; $500 or so for “we do one pricing experiment a month for you”And then eventually build things up to where I could sell someone a hosted SaaS pricing page, with automagic conversion optimization etc, for $99.

Option B, which I have not been kicking around for 3 years but certainly deserves to exist in the world:Write the Stripe of email marketing APIs.

I think existing email marketing software pushes way too much work at the company, particularly with regards to lifecycle email. So if you want to have something like “If you add someone to a project they get a notification” results in a hidden requirement like:“If someone adds you to 80 projects because you just joined the company, you shouldn’t just get spammed 80 times.”

But a) nobody knows that requirement when V1 is shipped and b) nobody at the company prioritizes doing it, so everyone’s email notifications are just constantly broken. They’re broken in all sorts of other ways, too, often in ways which are obvious to email people but less obvious to regular developers or marketers. So you should be able to centralize “defeat the brokenness” in an email marketing company the same way existing email marketing companies centralize e.g. delivery management (another hard not-obvious problem).


Question from Matthew

Q. Are you working on any personal side projects?

Patrick:

My agreement with my employers mostly forecloses personal side projects, and between work and two young kids, I mostly am lucky if I have 2 hours a week to crack open a video game. But who knows what the future holds. 


Question from Jay Winder

Q. You’ve now had a fairly broad range of career experiences, such as working at a large Japanese system integrator, doing consulting, running software companies, and now working at Stripe. What are the major lessons you took from each career experience?

Patrick:

What I learned from being a translator: being a translator is a really difficult job which appeals to people in the relatively narrow range of a) being smart enough to do one incredibly difficult thing and b) being able to stifle their urge to offer an opinion when two people in front of them say inane things. I was poorly suited to this job.

After being a translator/interpreter for a while I joined a Japanese systems integrator and worked as an engineer for about 3 years. I wish that my key takeaway from those 3 years of my life was not “Do not, under any circumstances, work for a traditionally managed Japanese company.” But I did indeed burn 3 years of my life on that job and that was indeed my key takeaway. To the limited extent that I am ever jealous of anyone for anything, I am jealous of the 25 year olds in the Stripe lunchroom, who are going to spend 3 years of their early career doing work which really matters, learning a heck of a lot, probably having more of their nights and weekends to themselves, and having material impact on the future of the business/our customers/the industry.

I will say that I think 3 years as an SI under two senior engineers who certifiably Knew What They Were Doing (TM) probably gave me a better appreciation for proper engineering practice than my undergrad degree did, and to the extent that I later succeeded in a mission-critical application as a business, it was largely due to learning the mindset from them.


Jay Winder:

Q.For people who aren’t aware and who might be interested, what would be the major differences between a traditionally managed Japanese company, and, say, Stripe?

Patrick:

Excellent question. So, without entirely recapping my article regarding working/doing business in Japan :

The traditional expectation at BigCo in Japan is that your career will be spent substantially at the same BigCo and the first ~10 years of it are training you into being a proper BigCo-ite. You spend a lot of face time mostly as a form of hazing / “we went through it so you should go through it” / effort to show people that you are trustworthy, loyal, and hardworking the way BigCo expects successful BigCo-ites to be.

When I say “A lot”, I mean that a short week for me was 50 hours, steady state was 65-70ish, and crunch time was 80-90. During that _ten year long_ de-facto training period, you’ll be given very few decisions which actually matter, and will mostly be producing work exactly as directed by slightly more senior staff. This is an annoying state of affairs when those slightly more senior staff know what they are doing. It is not the case that you are guaranteed that the majority of work will be assigned by people who know what they’re doing.

Now contrast this to Stripe, or another well-run startup (Japanese or American or other). The company will, by default, not even exist in 10 years. You have to do meaningful work, today. People who turn out to be quite skilled are given projects which are incredibly ambitious as their first thing out of university, and when they succeed in those projects, they succeed massively. Stripe Connect, one of the most important things our company has ever built, was originally the work of a single engineer in her first job out of school. (This oversimplifies for clarity but it isn’t substantially wrong, to the best of my knowledge.)

Everybody new in the industry has to figure out what they want to do with their career, learn some valuable skills, and figure out what adulthood looks like for them. Far better to do it at a startup (or at AppAmaGooBookSoft) than at a traditionally-managed Japanese company; it will be less painful and you’ll get where you are going years faster. ~15 years into a ~45 year career I’m increasingly of the opinion that a year is a terrible thing to waste.



Question from Kornel:

QI’m running SaaS and sales-related tasks (queries/negotiations/follow-ups), take most of my time, and I’m most likely also awful at it. How can I improve efficiency of this? (Delegate/automate/hire/cast a spell to make product sell itself?)

Patrick:

Mind giving us a tad more context on this in a follow-up, @kornel? I have a variety of answers here but they depend on various factors. I can go totally broad for the benefit of everyone w/o specifics but some specificity would likely help with the advice helping you.

KornelB2B SaaS. 1-person small biz, positive cashflow [default alive] (https://imageoptim.com/api) My clients are all over the place (bunch of small non-technical ones, few larger less problematic ones, couple of big ones). I’ve got some interesting technology, and a free app with good reputation & seo (10k DAU) which brings me leads for the for-profit part of the biz.

Patrick: Ultimately most software businesses will hire people to sell the software, but in the short-run everyone running a SaaS business will be doing their own sales. One generally cannot simply employ someone to do sales until your business is mature enough to provide that person with a script that repeatably makes the sale, a process for provisioning the product (and dealing with contractual/etc bits), and a steady stream of qualified leads.

Knowing nothing of your numbers I would predict, from seeing your pricing, that you spend > 50% of your time on questions from the $9 a month folks and make < 20% of your revenue there. Feel free to contradict me if necessary.

If I’m directionally accurate, the thing which I would be doing is sunsetting further efforts to sell accounts to anyone who denominates their website spend in $9 increments. I’d try to double-down on the commonalities of my very best customers, where the best ones are the ones with the highest value added by the service, probably highish to high usage, and (subjectively) the sort of people I would not mind spending 8 hours a day in a room with. And then I would tilt both my acquisition strategies and active sales efforts towards that segment.

Kornel

The $9 folk tend to use a plug-in for an off-the-shelf software, so they’re not that bad actually. But they’re small. I have queries for $1000+ deals, but these clients are demanding.

Patrick:

What is the genre of “demanding”? How much of it is amenable to being processitized away by you?

Kornel:

Different ways of billing, bespoke technical integration, SLA it’s back and forth “can you do this? and what if that?”

Patrick

So these are great problems to have, because they’re Really Software Business Problems. Which means they’re still problems, but they’re the kind which suggest $$$ is indeed in the air. You should start breaking your sales cycle up.

First off: write down what your sales cycle looks like. I bet, given that I would have done it this way a few years ago, that you have a lot of these discussions before you have an agreement with the customer. Which is a fine way to get started, but you should stop doing that. For deals which are only $1k a month (which, despite what software people think, is a small number), you want to break the sales process into phases.

Billing questions: “If you’re buying, we can figure out a way to take your money. Are you buying?” Maybe not quite that direct. But the thing I’d probably try doing first is: Standardize your initial conversation with someone, where you have a 30 minute call, hear about their needs, and tell them about how your thing fits in.

Based on the call, you give them a proposal and/or demo, or you do not. At the conclusion of the proposal, you ask for a 1 pager LOI from them.You template this agreement. If you sign an LOI, we do a kickoff meeting with the technical team.

If you sign an LOI, we talk about feasibility of billing or contractual issues. If you don’t sign an LOI, fine, but we don’t talk endlessly about why you are not ready to commit yet. Integration work should follow a signed contract (next step), at which point a) you’re charging and b) impetus is on them, not on you, to move forward. Happy to chat about this in more detail some other time.



Question from Sonicrida

Q. As a budding entrepreneur, what can I do to navigate the battlefields of investment if I’m worried about not finding myself in a situation where I feel like I’m selling user data as a part of my business model.

That seems to be a trend among a lot of companies in the valley. I read your recent thoughts on creators relying on other platforms that you posted on Twitter that were no doubt in response to the recent patreon drama as well as little bit into public benefit corporations and the investors that are supporting those but I was wondering if you had thoughts about how to approach any business with keeping the privacy of users in mind?

Patrick:

So, regarding selling user data: I think that “most companies in Silicon Valley sell user data” is a folks belief among technologically proficient people and does not correspond to reality, in most cases. User data is an extraordinarily valuable thing at e.g. Facebook, and they don’t want to sell their valuable thing. They want to rent access to insights developed from it. That sounds like a distinction without a difference but it is a distinction that entrepreneurs really need to understand. There is literally no way to go to Facebook and say “I have a million dollars here, can I buy your list of Muslim Republicans with over $100k in financial assets, please.”

You instead get to target an ad against that demographic.

To the extent that entrepreneurs do not like the realities of ad-based business models, I’d encourage you to either a) do B2B, where ad-based business models are basically non-existence or b) attempt to be one of the (rare) B2C companies worth paying money for. Understand that if you are building for customers, asking for money up-front makes it materially harder for you do grow in the explosive fashion that investors want to see for companies that they back, so you’re likely going to be losing money (your own or someone else’s) for a while until you either get to that graph (and get money from investors) or don’t (and implode).



Question from Matthew

Q. Light question: Favorite vending machine drink?

Patrick:

I am a heretical Tokyo-ite; I strongly prefer not drinking things from vending machines. If forced, I usually go for Itoen tea. My beverage of choice is cold brew coffee or iced coffee; my favorite way to drink it is leisurely, at a cafe, generally while tweeting about Bitcoin.



Question from Garrett Mosier

Q. How do you typically decide on industries to enter (medical devices, supply chain management, etc.) and how does that relate to which entry point you’d recommend (consulting, product, or megacorp)? A lot of the projects you’ve worked on aren’t something most engineers would think of (the bingo app for example) but seem to have done well.

Patrick:

I read very widely and talk to people, which perhaps overinforms what I end up doing professionally. I don’t think industry is exactly the way I would think about choosing broad areas to work in. I would suggest thinking instead of “Who do I want to serve over the course of my career?” and “What sort of challenges do I want to work on?” I think that this is a more clarifying way to think of it than industry.

Spoiler alert: devs are in the highest demand we’ve ever been in and moving from e.g. social startups to payment startups to the financial industry/etc is mostly a matter of getting a few beers or coffees with the right people.

I have gone back and for over the years on whether doing traditional-dev-friendly-businesses is a good or bad thing for solopreneurs. I currently net out to this: a) Software people do write waaaaay too much software for software people. b) Businesses have much, much, much more money than consumers, and it is easier to get. I would almost always prefer doing a B2B company over a B2VideoGamers or B2RPGHobbyists company. c) Given that one is selling to businesses, feel absolutely free to sell to developers within businesses.

There are plenty of ways to make very happy businesses in software to software people, and that helps you understand who you’re selling to, where they hang out, what messages resonate with them, what they ACTUALLY need, etc.



Question from Sacha Greif

Q.Bitcoin?

Patrick:

I get ranty when I talk about BitcoinShort version: I think that, in 50 years, we will still be dealing with Satoshi schemes, that they will be called Satoshi schemes, and that Ponzi will have been forgotten except by people who enjoy obscure financial history.

My much longer version of this will probably be on my blog someday. Disclosures: I don’t own any Bitcoins, am synthetically short Bitcoins, and am looking for more ways to short 😉

As of 2/1/2018, Patrick currently holds no position in Bitcoin, but remains broadly skeptical.



Question from pjan (abbreviated)

Q. Even though Stripe is global by default, out of the US, it’s only sales(-focused) offices. Is the new US policy encouraging you guys to start non-sales activities out of the US, or how are you planning to address this “losing out on talent opportunities” in your current phase of (hyper)growth?

Patrick:

Stripe offices abroad currently do many things other than “just” sales; they also generally handle support and field engineering for their markets, and we experiment with other things as market conditions and opportunities dictate. But for any big earthshattering announcements from us on that score, it probably wouldn’t come from this Patrick 😉


 Question from Paul Delhanty

Q. Regarding pricing a SaaS for Japan, is there are an equivalent to the $499/$999 per month that can often be expensed to a corporate credit card in the U.S. without getting authorization from purchasing?

Patrick:

It is my general impression that corporate cards are far less widely deployed in Japanese businesses than they are in e.g. US businesses. I think the right answer appears to be “bite the bullet and assume that sales will be higher friction at lower price points in Japan than it will be at the US”, including things like e.g. in-person meetings for deals which would not justify it in Japan, invoicing / bank transfer for even relatively small accounts, and similar. Balanced against this you have 1/3rd the churn that other markets see (yay!) and relatively little competition in most verticals compared to other markets.



Question from Naoya Kanai

Q.What are some Japanese startups (or local offices) which have the coolest opportunities for tech-inclined folks?

Patrick:

This depends on what you want in terms of cool opportunities. AppAmaGooBookSoft all, to my knowledge, hire material numbers of engineers in Japan. Different strokes for different folks but I think Google and Amazon have really interesting challenges available locally; feel free to talk to people who work there to substantiate that. Financial companies have huge, huge, huuuuuuge challenges which are often underloved by tech people, but which basically hit everywhere in the tech stack.

The HN Tokyo meetup has a lot of people at interesting financial companies. Exactly which is the best fit for you depends on what you want to do; pip pip for Jane Street as one good option there. Stripe has a handful of engineers in Japan already and it is highly likely we will hire field engineers, support engineers, and related positions in the future. If that is potentially interesting, please get in touch with us.

The Japanese startup I’m most impressed with is SmartHR and their (meteoric) growth foretells all sorts of interesting challenges in their future. I also really like the team environment over at MakeLeaps; Jay Winder can tell you all about it.



Question from James Sinclair

Q. With your background in being an Entrepreneur and now helping aid them from Stripe, what are common mistakes you see younger entrepreneurs making?

Patrick:

Have you folks seen https://indiehackers.com? Courtland, who runs it, now works with me at Stripe. If this question is near and dear to your heart you should love that website.

Broadly speaking I see a few problems with proto-entrepreneurs and a few problems with entrepreneurs. For proto-entrepreneurs, I see folks spend years scheming and dreaming without actually making forward motion on a product. One week in the market beats one year in a notebook. Get into the market, as soon as possible, whether that is interviewing potential customers, attempting to pre-sell your thing to people, or building and launching it. Different strokes for different folks, businesses, etc.

Among entrepreneurs specifically: Making things the world doesn’t need. I see a lot of entrepreneurs in Tokyo specifically who are doing X because X sounds like a startup and they really want to be doing a startup. These are often “sounds vaguely like whatever the flavor of the month is” or things with a very, very, very niche use case but matched to the intention to do the VC rocket ship trajectory thing.

I would encourage you to force the discipline of talking to people, early and often, to verify that their business does indeed need the thing, urgently, and it is on their top 2~3 list of concerns. I would also encourage you to do some soulsearching regarding what one wants to get out of the whole running a business thing. I don’t think any one path is for anyone and I occasionally see people jump on the VC-funded train simply because it is the one narrative other than salarymanhood that they have heard, sometimes to their detriment.

You have a finite number of years; spend them with intentionality.



Question from Paul Oswald

QYou mentioned some differences between Japanese and non-JP companies from the ‘working for them’ side… what are some major differences on the buying/consuming side that stand out to you? What should people building SaaS for the Japanese market look out for that maybe they aren’t aware of?

Patrick:

Great question! Broadly speaking, and assuming you’re coming from “I have a reasonably good understanding of sales to non-JP companies”: Much longer sales cycles in Japan. More difficulty in determining who the ultimate decisionmaker is; it is often diffuse and relatively rarely “the person with the most senior title”, although that certainly can happen. Many Western sales reps report a certain level of frustration because they do not have a good calibration for disinterest, and they therefore assume that Japanese companies were stringing them along for months while having no intention to do a deal. The Japanese companies very rarely see things the same way; they were often just being appropriately professional and/or preserving option value.

Senior decisionmakers in many Japanese companies are not natively technologically literate, to a degree which is occasionally surprising, and are very influenced by trends, general prestige/branding/etc in the industry, and whatever a small selection of (print!) outlets says is the new big thing. There are presently any number of companies asking what their blockchain strategy is. They should probably be asking what their put-a-database-in-it strategy is, since that is a more reliable upgrade from paper, but life is what it is.

Could talk about this in more detail but y’all should realize that Jay Winder and @poswald are probably better positioned than me to answer it. Also feel free to talk to Daniel Heffernan, country manager of Stripe, for his thoughts sometime. I’ll try to finangle him out to the meetups.



Question from Jeff Madsen

Q. “Serious startups can’t succeed using remote workers” – comments?

Patrick:

Almost every question which includes the words “serious startups” is jockying for social position in an imagined hierarchy of coolness of business rather than a considered inquiry into how to make businesses more effective. There is no debate in the tech industry more boring than “who gets to call themselves a Startup?” This question makes no software and sells no software.

To the underlying actual question: can you make software with remote workers? You clearly can. Can you sell software with remote workers? You clearly can. Can you achieve an $X billion valuation while employing remote people?

*looks in the general direction of public reports of his employer’s enterprise value while noting that he works in Tokyo, not SF*

Any heat in the above answer was directed at the larger discourse of “serious startups” not the individual asker, who I understood to be repeating a sentiment they had picked up.



Question from Sonicrida

Q. Just as an alternative to seeking investment in the first place, do you have any thoughts on mastodon style approach that a lot of open source projects seem to be moving towards? That is, patreon or other platform funding + sponsorships by companies that believe in the project.

Patrick:

I actually don’t know what Mastodon’s business model is. I am very bearish on soliciting donations as a business model. (I am explicitly not commenting on Patreon here. Patreon is a client of my employer and I am happy to see them experimenting with interesting things, much like I am happy to see many companies experiment with different things.) If I were an OSS developer, the business that I would try to learn the most from is Sidekiq simple, efficient background jobs for Ruby.

Sidekiq is a sole developer OSS which does background job processing for Ruby. It has been publicly reported as > $1 million a year in recurring revenue, entirely from enterprise licenses to the batteries-included version.

Sometimes this model is referred to as Open Core. But I think Sidekiq is a great existence proof that one can monetize OSS as a small developer. (It is also truly fantastic software; rarely does infrastructure so keenly understand what developers, infrastructure engineers, and operators will want out of it.)

Question from Sonicrida

Q. This is unrelated to my previous question but do you have advice for junior devs with an interest in relocating to Japan from the states to work for startups over there? How would you approach that process?

Patrick:

I think one should be really honest with oneself about one’s goal is. Is one’s goal to work in Japan or to live in Japan, for example? If one really wants to live in Japan and the work question is mostly because you a) want to have a visa and b) don’t want to starve, then perhaps the easiest option is working at AppAmaGooBookSoft for 2~3 years, getting recognized internally for being a high performer, and then doing an internal transfer to their Japan offices, which are generally constantly hiring and not often on the beaten path for high performers.

If you want to work in Japan, then you should expect a bit of Playing Life on Hard Mode. Minimally, get up to business proficiency in Japanese, and then force the world to give you a job. If either of those two things is hard for you Do Not Work In Japan Life Will Suck.


Jay Winder:

Ok – and that’s a wrap. Patrick is currently out of Japan on a business trip but still found time to do this, so on behalf of everyone here, I want to extend a huge thanks to Patrick! Patrick as always, your answers are insightful, clearly battle-won, and also very enjoyable to read. Thank you !!!

Patrick:

ご静聴ありがとうございました。I really appreciate you taking time out of your mornings (or evenings) to listen to me. I’m patio11@stripe.com if you ever have something work-related, patrick@kalzumeus.com otherwise, patio11 on the Twitters, and you can find me at http://www.kalzumeus.com.

Stripe is hiring aggressively (including in Japan) and we’d love to talk; visit https://stripe.com/jobs or email me directly. And now I’m going to run off to dinner with the family. Later all!

Jay Winder:

More things you might be interested in:

Patrick works on Stripe Atlas which helps Internet entrepreneurs get up and running. (You can use the invite code patio11-2iff5zq to skip the line)

Twitter: https://twitter.com/patio11
Comments on Hacker News: https://news.ycombinator.com/threads?id=patio11
General Patio11 Overview and links to blog posts: http://www.kalzumeus.com/start-here-if-youre-new/


Transcript by Sonic Rida