The bigger concern I would have with applying to YC in your case is the existential question, "Should I take VC money or not?" What I don't see above is an understanding of the why behind financing. How will financing accelerate your business? If I give you $1M, what will you do that makes it worth $10M in X timeframe? While not necessarily a deal breaker for a lot of seed funds, I think anytime one is looking at taking on multiple hundreds of thousands of dollars of investment, they should have some basic answer to that question.
So it goes back to my earlier point about a baseline for getting accepted which my application would not cut since I don't have hard data to show CAC, LTV and other relevant metrics. It's certainly not set in stone as a small number of very early stage companies do get accepted, but in general products have to be much further along to have a serious chance of acceptance, or at least that's the impression I get.
Your solving a big problem: having commenting that protects user privacy. We are happy to pay if the product works in our use case.
Totally off the parent topic, but you do seem to be building something that is incredibly useful, at least for us.
This is the kind of reaction you want to see in your customers.
Most of Patrick's excellent advice can be lumped into these three buckets. Specifically:
1) You have to establish the credibility of the team: you've done impressive things before; you have a deep understanding of what you're working on now; you can read your audience and know how to communicate effectively; you can get a strong intro (nice-to-have); etc.
2) You have to establish the viability of the market: it's big; it has a real problem; the existing competitors are not doing a good job in a clear way; etc.
3) You have to establish the quality of the idea/product: you have a unique insight or approach relative to competitors; the prototype/early validation is strong; etc.
A lot of the pitches become mediocre when founders are handwavy in one or more of these areas. For example, if the founder spends a lot of time talking about the market and the product idea, but not enough time explaining why the team is uniquely/extremely qualified to succeed. Or the founder has good answers to product/team/market questions, but their answers show they don't know how to read the audience or explain their idea. (Example of not reading the audience: the investor is non-technical and the founder, who is productizing their PhD thesis, spends 90% of the pitch geeking out about technical details.)
Also, I'll add a few tips:
- Don't exaggerate or mislead. An investor will pass if they doubt one of your statements ("silverware is a $150 billion dollar market!") or realize that you're spinning facts (e.g. you say Dropbox is a customer, but later it turns out you meant that one of your free users has an @dropbox.com email). If it turns out that one statement you made is false, then investors will assume there might be more.
- Understanding risks is better than sweeping them under the rug. If your competitor landscape is missing key companies (mentioned in Patrick's post) or you dismiss some $1b+ company as a competitor without any rationale, your audience will become very skeptical. Admitting something is a problem and explaining how you will address is it much more compelling.
- Really know the ins and outs of everything about your company -- at least relative to the audience. If I ask a question or make a product suggestion that the founder hasn't considered, that's a yellow flag. Someone who has been living and breathing their startup for several months should have a much, much deeper knowledge of their domain than an investor who is hearing about it for the first time.
I've also worked in VC and would add that you really need to understand the motivation of the potential investors you are pitching.
Early stage Founders often waste a lot of time by pitching anyone who says they make investments.
This will save you a lot of time and energy focusing on funds that you believe can add more than money to your business.
Also, funds with a proven track record are important. I've witnessed outright fraud from a VC fund that claimed to have $50MM to invest and signed contracts to invest over $11MM when in reality they had no money at all.
Don't start hiring or otherwise committing your company to expenses just because a VC fund signed some paperwork.
Wait till the money is actually wired over to your account.
You want to vet your investors as much as they are vetting you.
The OP is significantly about applying to YC: From what I've heard, YC doesn't much like sole, solo founder startups!
Why should no VC need call? For my startup, a PC server from $1500 in parts kept on average half busy 24 x 7 should generate well over $200,000 a month in revenue. Thus there would be plenty of cash for more PCs at $1500 each. Even $10,000 a month in revenue would yield plenty of cash for more servers.
SUSA Ventures claims to want "technical founders". Alas, it doesn't look like the SUSA partners are very technical!
"Warm introductions"? SUSA and many want "warm introductions": But VCs and I do not have associates in common. So, I can't get a "warm introduction" to a VC, and no VC can get such an introduction to me. E.g., I might be able to get a "warm introduction" from one of my Ph.D. dissertation advisers, at one time President of one of the world's best known research universities especially famous for their STEM field graduate programs, including computer science and AI. However I doubt that that person knows any information technology VCs. E.g., recently Tom Magnanti, Dean of Science at MIT, gave a technical lecture on the foundation of the Internet at a lecture series named for Professor X. Well, Professor X was the Chair of my Ph.D. orals committee; maybe I could get a warm introduction from him; but likely he knows no information technology VCs. Net, my background is technical, but VCs are not qualified to be my technical colleagues or associates -- VCs don't measure up.
For the people I know, the best form of an introduction is a good peer-reviewed paper of original research in a STEM field, preferably applied math complete with significant theorems and proofs. It appears that there are few or no such information technology VCs anywhere in the US and similarly few who could accurately evaluate such a paper.
SUSA Ventures claims:
"We seek out highly defensible companies that leverage data, economies of scale, or network effects to build value and achieve longevity."
"Seek out"? My experience is that SUSA ignores such things even when they land in their e-mail inbox.
IMHO, for the OP again, all the advice on pitching information technology VCs is noise and filler and useless except just one word, "traction", preferably in the form of after-tax earnings significantly high and growing rapidly. But for a sole, solo founder startup, by the time the business has traction enough for a VC to write a check, the founder likely will no longer accept such a check.
1) $10k/mo is not even close to enough for most US companies that want to grow quickly. If it were enough, then way fewer founders would want to fundraise.
2) I'm a partner at Susa and I'm technical. And even if I weren't, being technical has little to do with being a good partner to technical founders. Most basketball coaches can't dunk, but they are still good coaches. Furthermore, if a founder is technical, then a non-technical VC/advisor/mentor/friend that can help in other areas can be more useful than someone who is technical. Most technical founders I work with have no problem with building tech, but many struggle with the business side.
3) Our website says we prefer warm intros. We don't require them, we just prefer them. I reply to cold emails frequently and often meet people after they cold email me. But the cold emails should be good/well-targeted, and unfortunately most are not.
4) Traction is not the only thing that matters. For example, I just did a quick calculation, and over half of our investments in the last 12 months had $0 revenue when we invested. Many of those investments hadn't even launched yet.
Source: personal experience
I got the $10,000 a month from the OP.
I'm a sole, solo founder. So, right, my startup doesn't have five founders, each with credit cards maxed out, each with a pregnant wife about due.
And, right, my startup has no co-founder disputes. And since I'm 100% owner, I have no BoD disputes and spend no time reporting to a BoD.
Since I'm technical, I have no need to search for a technical co-founder.
To me, $10,000 a month revenue is a LOT of money for growth.
E.g., for my startup, the main opex is just my monthly ISP bill. From my ISP, my upload data rate is 20+ million bits per second (Mbps), which for my startup is a LOT (if on average half filled would amount to ballpark $4 million a month in revenue), and when I go live the only difference will be a static IP address instead of a dynamic one (from DHCP or some such). The ISP bill is less than $100 a month.
For now and well into a quite successful business, my main capex is just servers, and there, again, for my startup $1500 in parts for a server kept half busy 24 x 7 should generate $200,000+ a month in revenue.
My first server will go at my left knee. When that fills up, I will be awash ($200,000+ a month) in revenue for more and I can put a second one beside the first one. Then I can start putting servers and higher end Cisco boxes (LAN and/or router with whatever network security Cisco recommends) on shelves and/or standard racks in any of three empty bedrooms. If I fill up those servers in all three bedrooms, then it's millions a month in revenue, all still from just one employee, me. Then, sure, go rent some space, say, much of a closed factory, warehouse, or shopping mall, and start slowly hiring, first, a good office manager to interface with the lawyers, accountants, auditors, bankers, real estate people, security guard company, janitorial company, business insurance agency, payroll firm, HR consulting company, interior renovation contractor, ..., etc.
I'm "technical"? I saw the market need, designed the first good solution, an excellent one, did the crucial, core, technological advantage, barrier to entry, proprietary original applied math research, designed, wrote, ran, timed, documented, and debugged the code. Some of the AI people would call my applied math AI; to me, calling my applied math AI would be a gross insult.
To the users, the solution is just a Web site. I used Microsoft's .NET with ASP.NET (for building Web pages) and ADO.NET (for access to SQL Server). I wrote the code in Visual Basic .NET (as far as I can tell, compared with C#, essentially a nicer form of syntactic sugar and otherwise close enough to apparently Microsoft's most important language C#).
I've done serious software, mostly scientific and engineering, for a long time, but this is my first Web site. My first code for my startup is also my first production code (no throwaway prototype code), and, from the user interface to the back end servers, it all looks fine -- no need for refactoring. The code as is should be good for a major business, but then I should likely put in some code for good real time system monitoring, sharding, etc. The server farm and software architecture were designed to be scalable, e.g., via simple sharding, and some of the sharding logic is running in the code now. If the site gets north of, say, 10,000 users a second, then likely will need some more work on architecture and some more code. Then maybe, as a good Microsoft customer (Windows Server, SQL Server), I'll call Microsoft for some high end consulting -- Microsoft has lots of people who long since have been there and done that.
The code is only 24,000 programming language statements in 100,000 lines of typing (lots of in-line documentation). I wrote the code with just a good text editor (KEdit), and there have a lot of macros that help the work. E.g., in my code, at a use of something tricky from, say, ASP.NET, one keystroke opens the relevant Microsoft MSDN Web page of documentation (from pages on my disks). And the code has lots of other references to my documentation outside the code. And the code has cross-references in the code -- for a key of such a reference, I use just time and date; since I'm the only one typing the code, those keys are unique. Then I have some macros that, given a key, will go to the location with that key. Simple. So, I never used Visual Studio or any other integrated development environment. I didn't need on-line debugging -- my bugs were too few and too simple.
The server farm architecture has a Web server (sure, using Microsoft's IIS), a Web session state server (my own code, TCP/IP sockets, object instance de/serialization, and two instances of a standard, likely at most O(n ln(n)), collection class), soon to have (instead of Microsoft's solution I'm using now) a Web log server, SQL Server, and two specialized applied math servers.
All the communications among the servers are just simple TCP/IP sockets. Right, each of these servers is just software and, thus, can run on anything from one to several physical servers or virtual machines. So, sure, will get to do some load balancing and resource allocation. Likely just use the bottleneck principle: Find the main bottleneck and open up that one. Else could do some applied math optimization, likely not necessary.
All the four servers from my coding are single threaded, and the queuing is just from the TCP/IP stack -- if occasionally that doesn't work, say, from just TCP/IP input queue stack overflow or a dropped packet from a TCP error, then I'll see the evidence in the Web site log file and tweak the logic a little.
In production the main data will be read as fast as possible but written only once a week or so so can make great use of the solid state disk (SSD) drives. Originally my back of the envelope arithmetic indicated that I could provide a mature solution for the work for the world with about 150 TB of such data, and now that much data can be in maybe just one server. Then one nice, easy, simple approach to scaling is just to have multiple copies of the 150 TB.
Business experience? Done that at GE, two small companies near DC, FedEx, IBM, and some consulting. Been a B-school prof -- so, can claim to have taught business.
The problem I'm solving? Pressing for nearly every user of the Internet in the world. Fully safe for work. Squeaky clean. Culturally uplifting. Can contribute significantly to better government, alleviation of human suffering, and world peace.
Initial users, likely the Chablis-Brie, BMW, opera-going, ski challet set. So, right, good ad demographics. Later may do own especially effective ad targeting (from the unique data from the site and some of my original applied math).
For the problem, currently there is no good solution deployed. My guess is that my first good, really, excellent, solution will be a "must have", for nearly everyone on the Internet -- smartphone, ..., desktop, workstation.
For the revenue, initially all just from ads just in standard sizes, initially likely just from ad networks, I'm starting with focus just on the US. Eventually I'll go international but, to get good ad rates, likely only in the more advanced countries.
So, sure, get good usage from a large fraction of all the Internet users in the world and will have a company worth from a few old peanuts up to $1 T or so, somewhere in there.
Users need only a Web browser up to date as of, say, 10 years ago. For JavaScript, I have yet to write a single line of it; Microsoft's ASP.NET writes a few lines of JavaScript for me apparently for some usage I've made of ASP.NET; but users need not have JavaScript enabled. My site has only two main Web pages, and they send for about 400,000 bits each (with the JavaScript). The page layouts are simple -- just tables with everything positioned to the last pixel -- so, no HTML div tags. The pages use just a few, standard HTML controls and are easy to read from large fonts and high contrast. Some of the colors have been borrowed from some famous, very successful Web sites. So, the pages load very quickly and don't jump around during loading. A user doesn't need a fast connection to the Internet. The pages are simple with no icons, pull-downs, pop-ups, roll-overs, or overlays. Simple.
My site makes no use of cookies, and there are no user IDs, passwords, or logins. User privacy is very well protected.
Net, for users, it's simple, dirt simple. It's so simple a child of about seven who knows no English should be able to learn to use the only version of the site, in English, in about three minutes of tutoring or just figure it out in about 15 minutes.
But, from all I've seen, VCs and YC don't like solo founders. And by the time my traction is $10,000 a month, no way will I get a Delaware C-corporation, a BoD, a term sheet, and an equity check.
By $10,000 a month, my startup is already well up in the air and climbing rapidly when it's too late to buy a ticket.
Besides, when I wrote SUSA all this stuff in an e-mail message, it was totally ignored.
I have to conclude, VCs including SUSA (A) ignore their e-mail and (B) really want to see traction, i.e., notice the startup from the traction, not a contact from the startup.
The OP mentions accomplishments: Sure, I've got a bunch. E.g., I can literally claim to have saved FedEx from going out of business twice, once from some software with a little math and once from some math with just some calculator arithmetic. Yes, my office was next to that of FedEx founder, COB, CEO, F. Smith who remembers me; but I doubt that now he knows any VCs!
Do big stuff really fast? Been there, done that lots of times: (A) In grad school, I took a one semester reading course to study a question; in two weeks I had a solid answer to the question, with some nice, new results, that I later published. One of my results solved a problem in the famous Arrow, Hurwicz, Uzawa paper in mathematical economics -- of course, Arrow won his prize long ago and Hurwicz got his a few years ago. IIRC Uzawa has yet to win his! So, one semester, in two weeks, and didn't just study the problem but solved it with a publishable solution. Strictly, my solution would have been a Ph.D. dissertation -- two weeks, no faculty direction, maybe a record? (B) Some engineers for the US Navy were processing some ocean wave data and were confused about some points in stochastic processes. So, I got out Blackman and Tukey, got smart in a day or two, and for the rest of the week wrote illustrative software. On Friday evening I called in one of the engineers, ..., presto, bingo, our software house got sole-source on a nice development project. (C) The Navy wanted an evaluation of the survivability of the US SSBN fleet under a special scenario of global nuclear war but limited to sea, in two weeks. I saw a continuous time Markov process subordinated to a Poisson process, designed, wrote, and ran the code, had my code pass technical review by a famous prof, and the Navy got their results on time. Ah, that prof likely doesn't know any VCs either! (D) I've learned and used calculus and advanced calculus, taught calculus in a university, and published peer-reviewed original research that used calculus, but I never took freshman calculus and, instead, taught it to myself. (E) For my Ph.D. dissertation, I identified the problem before I went to grad school and worked out an intuitive solution in my head on an airplane flight. I did the solid, original applied math independently in my first summer in grad school. Then I designed, wrote, debugged, ran, and carefully timed the software -- with some stuff I did to make the software fast, it ran in about 100 seconds instead of about 64 years; otherwise I'd still be in grad school. I never really had any faculty dissertation direction. I wrote the software in two months, heavily during Christmas at my wife's family farm, and wrote and typed the dissertation in six weeks. I stood for my oral exam, and graduated. (F) As a ugrad math major, I wanted to learn general topology, got the leading book, advanced, not easy, got a reading course, taught the material to myself, and gave lectures to a prof. One week I covered a chapter, and the next week, the exercises. The prof taught me nothing. (G) From doing such math as an ugrad, my math GRE score was 800. (H) I've taught computer science to ugrads at Georgetown U and to grad students at Ohio State, but I never really took a course in computer science. (I) A computer sciences prof had written a statistics package but during testing got poor accuracy from his polynomial curve fitting code. Well, he was using the normal equations which for polynomial curve fitting are numerically unstable. So, I wrote some code using orthogonal polynomials that gave very accurate results.
Do those accomplishments count?
I still have some interest in communicating with VCs if only to get their feedback. But the main feedback I get is just silence. Again, I don't have $10,000 a month in revenue, and, again, when I do it will be too late for me to accept an equity check.
Again, I wrote something like the above to SUSA, and it was all just ignored. Totally ignored. Sorry for no "warm introduction" -- we don't know the same people.
Good grief: Why would a startup, prior to equity funding, want to be a Delaware C-corporation instead of just an LLC?
Disclaimer: happy Atlas user who did previous C-corps the old-fashioned way and did not enjoy doing that.
What's the advantage of doing it in 2 (expensive!) steps if you know the plan already?
I thought revenue was a "protected" term, like how it's described in the books. In that case isn't GMV the same as revenue? Since that's the money you actually invoice. And your cut is "net revenue", profit or something instead.
The GAAP guidance instructs companies to make the determination whether they're the principal or the agent with the following criteria:
1. You are the primary obligor in the sales transaction. This means, are you responsible for providing the product or service, or is the supplier? If you’re doing the work or shipping the product, you can probably record at gross.
2. You have general inventory risk. If you take title to the inventory before you sell it to the customer, and you take title to any returns from customers, you can probably record revenue at gross.
3. You can select suppliers. This one is important, since it implies that there isn’t some key supplier operating in the background who’s actually running the transaction.
4. You have credit risk. This means that if the customer does not pay, then you absorb the loss, and not a supplier. However, if you’re only at risk for losing a commission if the customer doesn’t pay, then you’re probably looking at recording the revenue at net.
5. If you get to set the price, then you probably have control over the entire transaction, and you can record the revenue at gross.
6. The amount you earn is fixed. This indicates a commission structure, which is sometimes set up as a fixed payment per customer transaction. If you earn a percentage of what the customer pays, this is also an indicator that you report revenue at net. In either case, you’re really just an agent for someone else.
7. The other two guidelines for reporting at net are just the reverse side of some earlier guidelines. If a supplier has credit risk, or if a supplier is responsible for providing products or services to the customer, then you’re probably looking at reporting revenue at net.
There's a pretty comprehensive document from the 'Emerging Issues Task Force' of the FASB here:
http://www.fasb.org/jsp/FASB/Document_C/DocumentPage?cid=121...
It isn't mindblowing, but insightful enough to take a look. "Focus on nascent greatness" is particularly a great section, because tries to solve some misconception about bizplan and ideas
