OK, one more "startup economics" post before I go back to Anki!
rdearman wrote:The most realistic, and cost-effective advertisements for our mythical company is probably direct mail. It costs around about £15 per thousand address mailed in the UK, so if you mailed all 25 million homes in the UK it would cost you about £25,000 and you'd bound to find someone looking for intermediate language products. £1-3 million to directly mail everyone in the EU is a rough estimate.
Actually, the most cost effective form of advertising would word-of-mouth from existing, enthusiastic users. But there are two problems with this:
- Word of mouth is really hurt by language barriers, so you need to keep re-launching one country at a time. Everybody in France knows that Assimil is
awesome at least seriously worth considering for self-study. I only tried it because my wife is French. - Word of mouth works great for dabblers, because there are a lot of them. But successful intermediate students are rare, and any advice they give is drowned out by the dabblers. This is the classic problem faced by the fitness industry—they know how to reliably turn couch potatoes into lean, strong athletes who can run a 7-minute mile and squat 1.5 times their bodyweight. But the real money is in people that want to mess around for two weeks every January before giving up again.
So you can't really rely on word-of-mouth unless you target specialist communities online.
Adrianslont wrote:One of the biggest issues - as discussed in this thread and others - the large slice of the market is beginners. They are the low hanging fruit, easier to develop for and the segment where most potential profit lies. I’m just hoping someone comes up with a concept for intermediate and advanced learners that a business can see a future in. Not every business has to be Google or even Assimil. I’m just saying that the market for non-beginners IS also huge - if not as huge - and Cavesa has detailed where much of the market lies.
Adrianslont wrote:I don’t think Assimil have “failed” emk. Did you say they employ fifty people? I imagine the owners living comfortable lives in comfortable French homes, eating fine food and drinking fine wine. This is all my imagination of course but they have been around for three generations so I don’t think “failed” is an appropriate word. They’ve been doing well enough to keep going. I think you make good points for untapped potential, though, Cavesa.
This is a good point, and it's worth digging into it a bit. From an investment perspective, you can divide software companies into two major groups:
- Startups backed with venture capital, which need to "swing for the fences" and "hit a home run." These companies ultimately hope to earn hundreds of millions or even billions of dollars per year.
- "Lifestyle" businesses, which accept very little investment, and which grow out of their own revenue.
As one entrepreneur
puts it:
Corbett Barr wrote:What’s a lifestyle business? Some consider it a patronizing term that the VCs and startup ecosystem use to put down businesses that don’t consume your life.
There’s a debate between entrepreneurs who say chasing swing-for-the-fences startups wastes your life and VCs who call businesses that stay small on purpose “dipshit companies.”
To make this concrete, Readlang and Assimil are both "lifestyle" businesses. Duolingo is a venture-capital based startup. A company like Assimil is successful if it stays in business, meets payroll, and earns a few million dollars a year. But if you decide to accept venture capital like Duolingo has, then you need to earn a
lot more money. This
blog post describes how the numbers work:
Nic Brisbourne wrote:The headline is that they received 60%+ of their returns from just 10% of their capital which was invested in companies that achieved home run exits of 5x or more. The average exit multiple in this homerun set was 16x.
This makes it very clear that unless a business can achieve a 5x+ exit, and maybe up to 16x+ then it isn’t likely to contribute meaningfully to the returns of a fund. Hence VCs target these sorts of exits on every deal.
Exceptional returns on the winners are, of course, necessary because of the large percentage of losers. In their sample 50%+ of capital was invested in companies that returned less than the original investment, and the vast majority of that 50% was invested in deals where all or nearly all the money was lost.
So Duolingo has accepted $108 million in startup capital and burns $40,000/day. In order to get a 16x return, they would need to sell for for
US$1.7 billion. Actually they want more, because the most recent $25 million was invested at a valuation of $700 million, buying only 3.5% of the company. So the investors really hope to sell Duolingo someday for $3.5 to 12 billion dollars. As you can see, the economics of venture capital mean that every venture capitalist is always chasing the next Google. Most startups fail, losing all the money invested, and so VCs need the winners to win really big.
This, in turn, raises another question: Why couldn't you just get a VC to invest, say, US$1 million? Well, the problem is that the good VCs tend to have a
lot of money to invest, and even a really successful tiny company just won't make enough money to
interest them:
Allen wrote:For a $25 million return in a $500 million fund (even on a $1 million investment), half of one's partners might remember to mention it to you on their way to the coffee machine (it not being worth a special trip down the hall to your office), the second half wouldn't think it noteworthy enough to mention at all, and the third half would say to themselves: "why did he spend so much time on such a crappy little deal?").
So basically, unless you have a story about how you can become as big as Rosetta Stone ($250 million/year in sales), you're not going to get the VCs excited. But remember, it's not sales that count, it's
profit—and Rosetta Stone is currently losing money. What the VCs
really want is a story about how you're going to steal some of that juicy $35 billion/year "total addressable market" away from the language schools.
But if the best you can do is €5 million/year (like Assimil), then typical venture capitalists won't be interested. I mean, they might talk to you for an hour because they like to talk to people (and they might learn something), but they won't open their wallet. You could try to tell them, "Hey, if you invest in us, you can come to annual board meetings in Bordeaux" (and I've heard the French government actually push this idea), but that's not really enough.
But if you can't access venture capital, this imposes some very strict limitations:
1. You're going to have to work hard for years for small amounts of money. Rideout absolutely went through this with readlang. As I mentioned upthread, I remember an interview where he said he was earning something like $10,000/year after several years of very hard work.
2. You may not have enough money to build a first-rate app. I know of one software company in Boston that will happily build you a professional app or website. They're very good at what they do, and they won't just set a giant pile of your money on fire. It's perfect for a non-technical business founder with a good idea. But their
minimum project size is $80,000. If you complain, they'll patiently explain, "Well, it costs about $80,000 to open a coin-op laundromat. You've got to spend money to start a business." Readlang managed to get around this because Steve Rideout built everything himself—but he spent more than $80,000 of his own personal time to do so.
3. You won't be able to afford serious marketing. Several people in this thread have complained about Readlang's marketing. Honestly, I think Steve Rideout did a pretty
respectable job:
readlang-website-small.jpg
You can see at glance what the product does. It offers you a sweet deal: "Learn languages by reading stuff!" If you just want to use the browser version (not the ebook reader), you can be up and running in 2 minutes for free. If you scroll down the page, there are more detailed explanations, testimonials from users, and it's all nicely done. This is what marketing looks like when you're only earning a few thousand dollars a year. Yes, it would probably sell better if he added, "Learn
naturally, like a child" somewhere. :-/ Or if he turned himself into a public personality like Khatzumoto or Benny Lewis or Steve Kaufmann have done—Khatzumoto and Benny are
very good at Internet marketing.
So what does this mean for people who want to make language-learning tools?Well, first you need to decide whether you want to focus on beginners, or on more advanced students. If you want to focus on beginners, then taking VC money might make sense. But if you go down that route, you'll be pressured into a laser-like focus on attracting first-time students and keeping them happy. Any time that you spend worrying about intermediate students is a distraction, because it's a much smaller market. So ultimately this means either (1) you chase the "dabblers", or (2) you figure out how to break into the ESL "false beginner" market one country at a time.
Alternatively, you decide that your true love lies in helping students make it from A2 to C1. If you do this, you're looking at staring a "lifestyle" business and big investors won't touch you. And honestly, Readlang tried this, and Steve Rideout did everything about as well as he could, given his time and budget. Others have done somewhat better, including LingQ. I agree that there's probably a viable market here, but I don't think anybody has really figured it out yet.
A third option would be to start an open source project, and possibly set up a Patreon for pizza money (or a Kickstarter, if you're really ambitious). You'd still lose money, but you wouldn't have any illusions going in. The drawback with this approach is that just like with LWT, subs2srs and my own substudy, you'll probably build something just good enough to do the job, and save time by building a very basic UI.
I've been leaning towards the "open source" route for years. I've been approached twice by entrepreneurs who wanted to start a real company (and I still talk to one of them regularly). I think
somebody is going to figure this out. But I'm waiting until the numbers start adding up. And this means either (1) somebody figures out how to sell to serious intermediates, or (2) somebody figures out how to make the dabblers
care about the quality of the product they're buying.
You do not have the required permissions to view the files attached to this post.