February 02, 2008
Jim,
As I was trying to understand where the angles might be in mobile computing, I came across this (cell phone soft), which says it is very difficult to start such a business. Thought you might be interested plus I wonder if you have a counter argument. –Stu
Cell Phone Software: The Billion-Dollar Sand Trap
We all know the advantages: Everyone owns one. We all know how to use them. Women love them and they fit nicely into a pocket. Did I mention they do wireless?
Cell phones seem like the obvious platform for the next generation of billion-dollar startups. But they’re not.
Hundreds of web entrepreneurs have gone into mobile. How many successes can you name?
That’s what I thought.
What follows is an explanation of why creating a successful mobile-wireless software startup is not just improbable, but impossible. Specifically, why
- The underlying technology is broken
- The business case is a proven recipe for failure
- The social aspects are more awkward than a middle school dance
The underlying technology is broken
1) There are hundreds of different phone models. Your software needs to run on all of them. How hard is this? Nokia makes a competitor to Loopt called Nokia Sensor. In the last five years, Nokia has only been able to get its software working on ten of its forty-three currently sold phones. And Nokia doesn’t even have the challenge of porting its software across the operating systems and architectures of multiple manufacturers.
Modifying the software for each phone’s display is a matter of brute-force labor. There’s no intellectual way around it. Yahoo! is one of the few companies that’s been able to pull this off, but only because they have an army of Ph.D. hackers working for them.
You won’t have an army of hackers like Yahoo!, nor will you understand the hardware better than Nokia.
2) The carriers partially disable Bluetooth functionality to prevent customers from downloading their own ringtones. This also means all those good features you came up with in the last brainstorming session aren’t going to work.
3) In order to load software, you need to buy the optional cable. No one owns the optional cable. Even if you gave your customers the optional cable for free, it only works with windows. Your early adopters use Macs.
4) You don’t know how to install software on your own phone, so why would you expect your customers to know how to do it?
5) Any software that pings the cell tower will quickly drain the battery. Pinging the tower every five minutes completely drains the batter in two hours. So much for making calls.
The business case is a proven recipe for failure
6) Cell phone carriers will never partner with you. At least not on terms that allow you to make a profit.
7) Even if one carrier partners with you, the rest won’t.
The next generation of WiFi will make your product obsolete in two years anyway.
The social aspects are more awkward than a middle school dance
9) Let’s say that against all odds you get a few early adopters. To everyone else it will look like they are just sending text messages. Unlike the iPod, your software is invisible. Invisible software isn’t viral.
10) You also can’t flaunt what you can’t see. So much for your idea of your product being a status symbol.
11) Cell phones don’t fit into girl’s pants. Remember how the women you asked said they would only use your software if it had a vibrate mode? Oops.
The canonical formula for business success is luck, pluck, and virtue. Success in mobile wireless is mostly luck. Maybe some prayer. It pains me to see some of the smartest people I know falling for the mobile trap. I’ve observed dozens of entrepreneurs go into mobile wireless. All have failed.
That’s not to say there will never be a day when it’s feasible for startups to venture into mobile. How will you know when the time is right? Ask yourself this question: Could I make money as a distributor of mobile software? If the answer is no (because there is no software to distribute) then find something else to do and check back again in a year. Mobile is still the future and it isn’t going anywhere. But in the meantime, better to let others get stuck in the billion-dollar sand trap.
Yes, the opportunities for a small start-up selling to the public are daunting. Some VC’s, e.g. Ann Windblad, blanch when you suggest Mobile as an area. It may, indeed, be only for the big players like Nokia, Yahoo, and Google to play in.
“The carriers partially disable Bluetooth functionality to prevent customers from downloading their own ringtones. This also means all those good features you came up with in the last brainstorming session aren’t going to work.”
Is this why the GPS/Bluetooth receiver won’t talk to my BlackBerry Pearl?
“Cell phone carriers will never partner with you. At least not on terms that allow you to make a profit.”
But maybe Steve Jobs will give you an entry.
“The next generation of WiFi will make your product obsolete in two years anyway.”
So do a WiFi app.
“Let’s say that against all odds you get a few early adopters. To everyone else it will look like they are just sending text messages. Unlike the iPod, your software is invisible. Invisible software isn’t viral.”
ALL software is invisible. So viralness must come from communication externalities,e.g. you have to upgrade MSWord to read the document some $%#!! early adopter sent you.
[reply this comment]
I agree almost completely. Of course, the situation outside
the US is much better than in the US. If you create mobile
games, then you can make money in Korea. If you have
the best site for European football, esp. in UK, Germany,
and Italy, you can be successful.
The business models for mobile data are different in other
countries. If people use your web-based app from a mobile
device, the network operator can charge based on the
volume of the download. That way, there can be revenue
sharing for the vendor. (Of course, most of the money goes
to the network operator, at least for now.) The problem is
much more difficult with installed apps, unless you are one
of the rare types who is able to get the app preinstalled on
the device by a large number of network operators. Even
then, the financial terms have to be worked out with each
operator. Some of the handset manufacturers have created
business units to help get applications onto their phones and
to use their relationships with network operators to simplify
the task of creating these business relationships. But it’s
painful and slow, without a great chance of success.
There’s a product called Celsius that has a database for more
than 750 devices tracked by 648 characteristics (sort of
like a mobile answer to curses). It’s produced by a French
company Mobile Distillery ( http://www.mobile-distillery.com), which
claims to substantially reduce time to market, testing, etc.
They’re going to show their solution for Android at the
Mobile World Congress (formerly 3GSM) in Barcelona in
mid-February. This approach helps in creating the UI for
mobile devices.
As with any other computing device, someone is likely to come up
with a killer application for which people are willing to spend money.
Unfortunately, it’s likely to be porn or gambling, a la Zivity
(home page OK, rest of site NSFW).
In summary, I agree that installed apps using the phone network
are a good way to lose money, but WiFi access to SaaS apps
from a mobile device has a better chance of succeeding.
[reply this comment]
Thanks for this. These are all good points and most of them valid. But before we all give up and try something easy like robotics, or try to bootstrap a new form of energy to save our planet let’s draft a counter argument.
Does Ann W really blanch? Is that before or after she invests in open source.
#1 - In 1997, my firm invest $26M over 3 rounds in a company called UTStarcom. founded by a UC Berkeley graduate he founded
a developed a cell phone and wireless base station using old “cordless phone technology”. He started a company selling this technology
and cheap phone and services to Chinese villages. Went public in 2000. NASDAQ: UTSI. My firm made a $200M profit on our investment.
#2 - In the early ninties the smart boys at Kleiner Perkins invested in Palm Computing. Who missed that one?
#3 - in the mid-ninties some crazy Canadians started a company called RIMM, a few lucky angels got in early but not much VC. They are now the
richest men in Canada.
#4 - In 2000 my firm invested in Danger a company founded by Andy Rubin. Company filed its S1 last quarter. Fingers crossed whether it will ever go public.
Okay you’re tired of hearing about the money made in the “handset” market.
#5 - Who invested in Phone.com (now called Openwave). Did Ann get this software company?
The company started in 1996 as Libris, Inc. and focused on developing mobile client software for “pull” services while the general mobile market was rapidly growing “push” services based on SMS. In 1998, it changed its name to Unwired Planet, Inc. and launched its proprietary end-to-end mobile network solution for Internet access and web browsing, known as up.link (browser and network server/gateway).
In 1999, with the introduction of WAP standards, it acquired Apion, Ltd. of Belfast (formerly Aldiscon Northern Ireland, Ltd.), changed its name to Phone.com and went public on the NASDAQ. In 2000, amidst huge growth in revenue and stock price, Phone.com acquired several high tech startup companies with niche products to integrate across its end-to-end solution. In 2001, it acquired Software.com and changed the company name again to Openwave Systems, Inc. With Software.com’s large installed base of email servers at ISPs, Openwave expanded by providing its mobile operator customers with software infrastructure for mobile email applications and other multimedia messaging (MMS) applications.
In January 2006, Openwave closed the $120 million acquisition of Musiwave, a French music application services provider for mobile phones.
#6 -
In a landmark transaction, it was announced today that Paragon Software Ltd, a company backed in 1998 by Kennet Capital, the venture capital fund focused on European technology companies, was sold to Phone.com (Nasdaq:PHCM) for approximately $500 million.
Paragon Software is a pioneer of synchronization technology allowing PC-based personal information to be easily transferred to mobile devices. Paragon Software’s product, FoneSyncTM, enables users of PC and Internet-based personal organizers such as Microsoft OutlookTM, Lotus NotesTM, Excite@Home, and Yahoo to download and synchronize contact information to over 240 digital mobile phones from 20 major manufacturers including Alcatel, Ericsson, Nokia, Panasonic, Philips, Samsung, Sanyo, Siemens, and Sony
#7 - Number two handset maker Motorola said on Thursday that it agreed to buy mobile phone technology developer TTP Communications for £103 million ($193 million).
The list is long. It is NOT easy for an entrepreneur to make money in mobile. Is it easy for an entrepreneur to make money anywhere?
But the market is BROKEN, and in transition. Apple and Google are trying to shake it up. to change the paradigm. There are opportunities.
Big (skype like), and small (sell to google, motorola, nokia, Rimm etc.) but they aren’t easy.
Always invest before everyone else sees the opportunity, not after.
Good night,
Scott
[reply this comment]
jhm reply on February 25, 2008:
[from Martin Griss]
I agree with Scott.
The new generation of phones, the impact of the opening of handsets, software, services and spectrum are the beginning of a disruption (I was at HP when people said you couldnt make money on teh Web, even when Joel Birnbaum was trying to get people to see the web as disruptive).
Things like Yahoo’s fire-eagle, and google’s open social will provide new platforms for mobile social applications of various kinds.
When handsets acquired GPS, 3G and cameras, all of a sudden new kinds of applications became possible. Someone is going to do other hardware additions for some platforms that will catch on - in security, displays, biometrics, RFID, NFC, motion sensoring, other sensors, etc. and these are going to enable a new raft of capabilities.
- Show quoted text -
It seems to me that that these lines of argument get more at the opportunity: (1) the notion that the cell phone is expanding as a platform for more general mobil computing. So it seems to be eating up part of the market space of cameras, PDAs, hand-held GPSs, MPG players, hand-held gaming, maybe even credit cards. It adds portability and contextual (eg., locational) awareness over PC platforms. (2) the notion that the current market structure can’t be maintained, but is going to be disruptively restructured by Apple, Google, and others and that toolkits will be available to build on in research projects and classes, and (3) the notion that new phones will just be other kinds front ends to database-centric, server-based applications. Other phone like devices are also possible in this space, like Amazon’s Kindle book reader that uses the phone network to buy and download books, but isn’t really a phone.
These lines of argument might indicate that the block on technology and business described by the “mobile computing trap” article are surmountable in future disruptive developments and that starting a center in this area might be able to catch the wave of this change.
It would be interesting to structure this space of possibilities somewhat more in terms of technology, use, and research needs. For example, on the interface end, there are problems with delivering high functionality on such a small device (all those impossible hierarchical menus), there are dialog, display, and input problems. Some of these can be solved by “input on output”, where contextually determined output on the screen then limits how much input is required for command or argument specification. There is the searching of the possible application space to be done. This space might benefit from a new set of conceptual inventions. For example, should there be a mini-me notion of electronic identity that resides in my phone, such that if I am carrying the phone I can auto log into my PC without a password, I can open the door of my house without another key, and I can keylessly start my car? Can I put my phone onto “autotwitter” so that my kids can tell where I am on their Google map? Can we make a barcode suitable for magazine ads or the sides of a bus that will let my camera phone and browser use it to call up information from physical objects in the environment? Etc.
[reply this comment]