Hacker News Reader

Comment by chatmasta | Hacker News Reader

Funny enough I just wrote one of my first blog posts in years, minutes before this article was written, about Magic. [0] My theory is that by positioning themselves as an aggregator, or a "platform for platforms," they retain the cashflow advantages of service based marketplaces, without the need to build supply because they piggyback on existing supply.

Some say this is unsustainable because of low barrier to entry, but I would say actually they are in a perfect position. With investment like this, direct partnerships with service providers are possible. Partnerships are a barrier to entry.

[0] https://milesrichardson.com/the-magic-of-service-business-ca...

View on HN

This.

I liken it to Segment.io who managed to move up the analytics stack and being the integration point for a product. This lowers the switching costs to using other analytic services and gives Segment all the leverage to create the highest value products in the future and offer one button migration. Additionally they have efficient platform economics with their emerging app-store. If I was an analytics provider I'd be really worried about Segment. Now Magic has something that could be similar, defaulting them to owning all the relationships with the consumer in the very hot on-demand industry.

To use a different analytics service, I have to change my code. To use an alternative to Magic, I have to text a different phone number.

apples and oranges.

Don't forget that they have a stored credit card with payment information as well (via Stripe). That's another barrier to entry - especially considering the effort in building up enough trust to hand over payment details.

that certainly appears to have been a barrier to their launch.

I haven't used magic, but their demos imply that they keep data on you.

Example:

  "Send a cake to my girlfriend"
  
  "That'll be $45.43 in 90 minutes".
  
  "ok, do it"

Who's the girlfriend? Where does she live? It's evident they've asked these questions the first time she's come up and then recorded them for later use. Since the service relies to much on personalization, the longer you use it the higher the switching cost will be.

Think they might have over simplified it in the demo. There is no way for them to know if the girlfriend has become an ex or if the person is still living at the same location. They would have to confirm it every time.

    > There is no way for them to know if the girlfriend has
    > become an ex or if the person is still living at the
    > same location. They would have to confirm it every time.

"That'll be $45.43 in 90 minutes to Kate at 123 Bubble St".

Big deal.

Yeah, it's so much harder to just type that the first time:

"send X to my girlfriend at 123 bubble st"

So I stand corrected. This is obviously a massive, defensible industry. Nevermind the fact that there are a dozen of these things in Europe, and none are notable businesses.

Actually the barriers to entry aren't that low. You need to have at least 6 figures to deal with the "float" problem (7 or 8 figures if you're successful), you have to hire people to deal with the orders, etc. Then you have the whole fraud aspect, which if not carefully managed could quickly bankrupt the business regardless of its bankroll. If I were a Russian credit card thief, I'd be hammering away at Magic right now. "Find me a Rolex submariner with 2 tone gold and mother of pearl face".

You need to have at least 6 figures to deal with the "float" problem (7 or 8 figures if you're successful)

What's your modeling look like here? They capture payment on day 0, are sent an ACH from Stripe on day 2, and physically have it in the bank on day 3~5. The worst possible credit terms they'll have with vendors is "We put it on our Amex on day 0, are invoiced by Amex on day 15, and pay the invoice on day 22." That lasts until they can tell without-loss-of-generality 1800Flowers "We did 100k with you last month. Give us 30 day net on a rolling monthly basis. [+]" My back-of-envelope math suggests they have ~15 days of float at present and will have ~40 days once they have direct relationships with suppliers.

They're almost certainly generating tremendous positive cash flow, limited only by their ability to generate transactions.

[+] That will not be the only ask in that conversation, believe me. I'd take pertinent note that 1800Flowers has fat, fat margins and is willing to pay affiliates 25% commissions, then ask them for their top-tier affiliate offer. Who benefits from that discount? Magic, because they're selecting for price-insensitive customers. If 1800Flowers doesn't want to play ball, you just update your CS docs to say "Our preferred flower vendor is $EMPLOYS_SANE_BD_TEAM."

> "We put it on our Amex on day 0, are invoiced by Amex on day 15, and pay the invoice on day 22."

Despite what people think, Amex is not unlimited [1]...in fact it has a limit, they just don't tell you what it is. Although he edited the comment, a very ignorant person posted below that they could just charge a few million dollars to their Amex, and pay it off at the end of the month. No, they couldn't. When you hit the Amex limit you don't know about, your card gets declined. That's why they don't say "unlimited," they say "no pre-set spending limit".

So that may have been good for maybe $20K or so (depending on the credit of the person whose card it was) but that doesn't get them anywhere near the finish line if they really had 17K requests in the first 2 days.

I agree with you that they could work out discounts/affiliate agreements with some large merchants, but given that this is a long-tail business, that won't be the norm.

[1] http://www.moneyunder30.com/american-express-cards-do-have-s...

A few Internet buddies of mine did affiliate marketing, which has roughly similar cashflow issues. If you need to charge $100k in a day, and you can demonstrate capability to pay $100k in a day, Amex will happily accommodate your business.

The main question is whether you can scrape together $100k~200k for your first 48 hours. I don't know what to say about that other than "It's clearly doable." (Anecdata: My personal credit limits total about $100k and have since BCC was my main source of income. That's the "No phone calls required" number for what a young professional with a fairly modest income and good credit habits can have just lying around. Do I have difficulty thinking that a team in Silicon Valley can float $X00k for N days? Not in the least.)

>A few Internet buddies of mine did affiliate marketing, which has similar cashflow dynamics to this. If you need to charge $100k in a day, and you can demonstrate capability to pay $100k in a day, Amex will happily accommodate your business.

Possibly, but I don't think that consists of calling Amex and telling the phone rep to log into your Stripe account to check out your mind blowing pending balance. Anyway all I was saying is that this was a weekend project of a YC company that had already (presumably) burnt through a good portion of their cash on their initial idea. You are right though, if you're a YC company, there are people you can call to deal with cash flow issues.

I would suggest that anyone who wants to run a business approach this and other issues with a mindset of "This is a problem which I can trivially solve." rather than "This is a problem which would be trivially solved for me if I were member of a special elect, but given that I'm not, boo hoo."

Chase bank didn't give me a $25k credit line because they're uber impressed with my HN karma. That's basically their get-you-in-the-door offer if you have good credit.

Amex has floors of people who do underwriting and convincing them to let you give them money is by far not the hardest sale required of a startup founder. (An early tactic to try: "Would you accept written confirmation from our accountant that we added $100k in accounts receivable in the last 24 hours and will probably add $100k in AR in the next 24 hours if we can get a credit limit increase from you? I can have him fax it to you on letterhead. This is time-sensitive and while I'd greatly prefer to work with Amex you are not the only bank with a fax machine.")

Stripe is full of friendly and accessible people, and should you be cursed with $100k of transactions on launch day (oh noes!), you have options including "Use the HN search function to find the CEO's email address and ask him for help, which he'll happily give you, because it makes his company very happy when non-fraudulent operations rack up $100k of transactions in a day. If he had a magic wand that could conjure up businesses like that he'd do virtually nothing other than wave it."

I was merely making a statement of fact - that if you are in YC, and you have a cash flow problem and receivables, odds are very good that someone will wire you some money or cough up a credit card that actually is good for millions of dollars. That isn't necessarily the case for everyone everywhere. Many businesses die because they don't have access to the capital to fulfill orders.

You're assuming $20k limit why? For my part I have several cards that individually have limits in the $50k range. Not so much because my income is all that huge, but because I have quite long, good credit history. I don't think that's all that unusual in places with SV level software developer salaries.

I'm not saying there are no cash flow issues here, but it would not take a very unusual team to be able to scrape together $500k-$1m in credit on cards alone.

It's something everyone can do, but it's also not something that's necessarily going to be all that difficult.

A large part of your blog post seems to be premised on the difference between incoming cash and outgoing cash. That only works as long as the people you're paying will accept net-30 (or have it imposed on them) -- Amazon was a great example of this but their suppliers also got wise to it.

In the current scenario, all those instacarts, eat24s, etc are also looking to get paid first, so this advantage should evaporate (leaving only those which are systemic -- ie banks, processors, etc).

If you receive money on a credit card and pay expenses on a credit card, you're virtually guaranteed to have positive cash flow. It's one of those services you're buying when you pay CC processing fees.

Example: Suppose I am in a "Buy things for $50, sell them for $60" like Magic is. Day 0 is the transaction day. I buy the good/service for $50 on my Amex and capture a $60 transaction from the end-user using Stripe. Stripe sends me 97% of $60 on day 2. It arrives in my bank account on day 5. My next Amex statement is, probabilistically, issued on day 15, and includes the $50 I spent to service my customer. On day 22, I instruct my bank to transfer Amex $50. (We'll ignore the relevant detail that I'm generating cash rewards from Amex up the wazoo once I hit scale.)

In addition to $8.20 or so of revenue I've also had ~17 days of use of someone else's $58.20.

Extremes in this area: insurance companies, pension funds. I think it's called "float".

OK, so in the 17 days I have the $58.20, let's say I can make a completely optimistic 5% APR return risk free on my float. I've now made a whopping 13.5c extra on top of my $8.20 I made providing the service. Excuse me if I'm not getting excited over squeezing out an extra 1.6% profit.

It's not the profit that you're trying to capture, it's the flexibility gained from having working capital. In that 17 day window, you can complete another sale without any additional outside money or pay for an unexpected expense. Having working capital is huge for quick turnaround businesses;

http://blog.mbaco.com/importance-of-a-good-working-capital-m...

That's not the main thing that float does for you. Massive positive cash flow as a result of operational dynamics means that you if you happen to have explosive growth you can fund it out of your cash flow rather than having to go begging on Sand Hill to pay for your rocket fuel, like many tech companies.

Might as well give two examples of how it looks like with different business models:

+ SaaS is often cash-flow negative if you're doing month-to-month billing, since cost of customer acquisition (marketing spend + sales commission) is routinely 4~8 months of billings. If you increase MRR by $100k in a month you're doing great, unless you don't have $500k in the bank account, in which case you just bankrupted yourself.

+ E-commerce: If you're in a model where you actually hold inventory, it is very possible to need to pre-purchase a large amount of inventory in advance, occasionally on non-favorable credit terms (like e.g. cash before delivery), prior to being able to sell it. Many of the great e-commerce businesses are built on various hacks which flip this cash flow cycle around. Amazon, for example, was famous for (ab-)using traditionally prevailing payment terms in book publishing which are hilariously in the favor of retailers ("We'll just give you free inventory and only ask you for money if you sell it." "OK. Tell you what, how about we sell it, only then even take delivery, and then pay you." "My, you tech people sure don't understand how this game works. OK, sounds great.") -- that's one of the reasons they picked books as their first base of operations.

Reply on news.ycombinator.com