Hacker News Reader

Comment by hrdwdmrbl | Hacker News Reader

I still don't understand how 1. Companies are paying so much for their software. How much do they charge for licensing to make so much money? 2. They intend to defend their position from competitors 3. How they will ever grow to justify a PE ratio of 2,123.29!!!

I actually use their software and love it, but do they really have a secret sauce that won't be copied soon?

View on HN

The secret sauce for Zoom seems to reliability. For some reason it just always works, independent of how many people join the call (we have calls with more than 100 people calling in independently), which OS you use (most of us use Linux, but also Android, Windows, macOS and iOS). It just always works, in the last two years I had zero problems with it, no idea how they do it.

Last I checked Zoom avoids using WebRTC in favor of using media over websockets. This should obviate a lot of the reliability problems with enterprise networks that you would normally need to solve with STUN/TURN/ICE.

I think the other thing they do is decrease the image quality when multiple people are simultaneously connected. This is a nice hack since you don't really notice the difference, and users blame bad images on the network not the product.

The biggest challenge to their business IMO is if/when WebRTC becomes stable between all browsers. There are still many kinks that need to be worked out. Everything falls apart in the P2P version when you have more than like 5 people connected as the amount of information that needs to be transferred across the network scales exponentially.[0] So you need to add a gateway and connect everyone over a single pipe.

[0]: https://en.wikipedia.org/wiki/Metcalfe%27s_law

pg_bot covered the technical side, I'll take the social side.

That's the great thing. There might be an entire market of options out there for an idea, but if they don't get something like reliability right, then that's the easiest foot in the door.

I thought let's download and try to find out their secret sauce by reverse engineering. Their download is just a web browser client [0]. So, my guess is pg_bot is right. They might be using web sockets.

more details - https://webrtchacks.com/zoom-avoids-using-webrtc/

[0] - https://www.zoom.us/download#client_4meeting

If your product has no marginal cost (as much software effectively does) then you can justify any P/E with the prospect of increasing sales without increasing opex or capex. Of course that's just an approximation since you will realistically have some kind of customer acquisition cost, some of which is also realized in adding more developers to create features that are adoption-blockers. But still, that scalability is what makes early stage software companies such attractive investments.

I would be much more worried about competition. Video conferencing has a small moat especially since major incumbents are already in the space

Though in economic theory cost should approach marginal cost

Only if it's a very competitive market, and even that effect will only happen over time.

Product differentiation introduces pricing power for individual firms, and depending on the style of competition you think will occur (e.g., Cournot oligopoly vs. Bertrand), you may very well never approach anything close to perfect competition with the small number of firms there are in this market.

> Though in economic theory cost should approach marginal cost

...in a perfectly competitive market. In a situation where monopoly rents can be charged (which is what having a moat is about), prices should hit the point where unit price × sales - cost is maximized.

Well, good questions, but how is it that we’ve had 20+ years of video conferencing over the internet and most solutions are total crap?

Product management is a fine art... Maybe Microsoft and Google will manage to catch up, maybe in trying so they’ll dig themselves further into a hole...

I suspect because most are free. Pay for something good that just works. Zoom has taken my custom and money for years.

> I suspect because most are free. Pay for something good that just works. Zoom has taken my custom and money for years.

WebEx and GoTo charge (quite a lot), and both are seriously ropey.

Because imo, the decision of which video conferencing solution to pay for is rarely made on the basis of the quality of the software, at least in the corporate world. Cisco and Ms have very effective sales teams selling WebEx and Skype as bundled add ons to ip phone systems and windows site licenses. The fact that they're both complete garbage is never considered.

Att some point it is "free" or bundled with some offers

Not so much the windows site licenses as their office ones but yes. For anyone running office 365, base teams/skype conferencing is free and adding on call in lines and the rest is generally just a matter of ticking boxes and won't require purchase approval or the rest.

People tend to underestimate just how different it is at most companies to modify/increase an existing agreement vs. setting up a new one.

Probably better than the P/E of ZOOM, an unrelated penny stock which has spiked after their launch.

https://finance.yahoo.com/quote/ZOOM?p=ZOOM

More seriously for a very early stage company with a large market P/E is probably not a good measure.

I wonder if there's money to be made searching for penny stocks that investors might mistake for a larger one, and buying it.

Its possible that the people who bought it recently were employing that exact strategy. If the spike in price really had nothing to do with the underlying company you might be better off shorting it. That way you will profit when people who bought it (on purpose or by mistake) try to unwind their trade

From speaking to friends, getting everything to work reliably for videoconf is a really really hard problem given the number of variables involved (codecs, networking, firewalls, bandwidth, locked down enterprise environments), so any solution that provides even close to reliable service is in a position to charge a lot.

But don't you think it's one of those things, like the 100 minute mile, that once someone shows that it's possible then 10 clones will show up overnight?

Depends. If the 'moat' is a strong UX + a reputation for 'just works' + the tricky enterprise customers all have your connectors installed already, that could be difficult to overcome.

Where did you get PE 2123?

https://www.marketwatch.com/investing/stock/ZM

It's 2250 now! Weeeeeeee

Reply on news.ycombinator.com