Startup Ideas

Category Idea Issues
MMORPG Character Management As massively multiplayer online games become more popular, it's likely users will have several characters to manage.  A web service that tracks how all your characters are doing will become necessary -- ranking, inventory, last login, etc. (03/01/2006)  
Location-aware services
  • General ideas related to widespread GPS availability...
  • Clear vertical apps to do while the market is growing:
    • Automatic time sheet service for lawyers and consultants and contractors.  Keeps track of what clients you've visiting during your workday and builds a time sheet automatically.  You edit/correct and it learns and gets better.  Prints out nice invoices on your company logo.
    • Network-based exercise/cardio/training service that tells you how many miles you've run, calories burned, etc. just by looking at your GPS track.  Your phone becomes your fitness training device (you'll have it with you anyway!)  Offer a free/ad-based service equivalent to the clunky hardware products early adopters are now spending $99+ to buy: see Garmin GPS Watch.
  • Broader consumer apps:
    • You would have the equivalent of a giant map of the world with (someday) millions of customers moving around.  You could mine that data, comparing where people are and have been, to understand what they like and want to do next.  And where their friends are right now.  And where their kids are right now.  And when you need to leave work to get home in time to see your daughter's school play.
    • Notice traffic patterns and alert people to restaurants/shops/etc. that are popular.  e.g. "After dinner at La Folie, 80% of people get a drink at Harry's Bar", or "People who visit these 5 shops that you like also visit these 3 others that you've not been to." or "Tonight, all the people like you are over at Sweetwater."
    • People never bother to enter the data that tells you their real preferences (i.e. where they shop, where they spend their time, when they visit the doctor, when they hang out at Starbucks, when they're at work), but this thing would know that and would therefore know more than any other consumer-tracking service now available.
    • Track medical visits for health monitoring.  It could notice when you visited the dentist, doctor, hospital, gym, McDonald's, etc.
    • Note when customers visit the doctor, dentist, hair salon, etc. and alert them when they haven't been in a while.
    • Build a match-maker app using location (e.g. "a woman in the restaurant you're now visiting has the same interests, click here to introduce yourself").  Protect anonymity until both parties agree to meet.
    • Major privacy/monitoring concerns, but face them head-on.  Let people "mask" where they have been over any period by, for example, telling the service to "pretend they were at Starbucks" for that particular 90 minutes.  Better than just leaving a blank track in the record -- that looks suspicious.  This is not to imply, of course, that I have any need for such a feature!
    • Work with credit card companies to avoid credit check failures (i.e. Visa will be alerted that you are in Europe so that charges that appear there don't seem fraudulent)
    • Advertisers could be alerted that you are a regular customer of a competitor (e.g. Jiffy Lube, McDonalds, Peets, etc.) based on your GPS patterns.  They could then make aggressive offers to get you to switch (e.g. "next tune-up free at Midas")
  • Opportunity
    • Recurring revenue service
    • Building a valuable database of real-world user behavior


    • Technology risk:
      • GPS power consumption and accuracy may not be adequate for 2-3 years
    • Market risk:
      • GPS adoption
      • Selling PDA apps tough


    My Health
    • Website that aggregates all of my medical records. 
    • Get 2nd opinions from India/etc. 
    • Chart progress on weight, cholesterol, etc.  Insurance info too. 
    • A "yodlee" for health info.
    • Tie in to calendar (appointment reminders, etc.)
    • My personal blood test data (I had to hand-enter it!)
    • WebMD market cap = $1.7B (though not 100% website-derived!)
    • Demonstrated consumer interest in online health management (e.g. Dieting)


    • Privacy issues
    • Regulatory hurdles


    The "Jet Blue" of Health Insurance
    • Trend: aging baby boomers realize they need to take charge of their own health care.  The current employer-managed tiered system does not work.  It separates customers from their own health management.
    • Build a simple, low-cost, customer-friendly health insurance firm
    • Just like Jet Blue or Southwest Air: no frills, but quality and low cost
    • Sell only catastrophic coverage (i.e. high deductible) -- per Milton Friedman.
    • Focus on self-employed COBRA recipients who have no real firm "for them"
    • Include online health tracking services (see below) to ensure customers do take charge of their health care
    • Huge market
    • Help redefine health care


    • Lack of domain expertise!
    • Insurance is a lousy business


    Super PIM
    • Everything needed to build a great "super-PDA app" is accessible inside Pocket Outlook. This means you could write an app that...
      • ... could check to see what meeting the person is supposed to be in NOW and use that data to create filenames for photos taken ("2005 CFO Conference.jpg") or to influence SMS/email recipient lists (e.g. since it knows who you're seeing later today and should be calling).
      • ... could mine your IM, phone call, and email "sent" lists to figure out who you're staying in touch with and who you should check in with.
      • ... compare current GPS location vs. next meeting location and fire off appropriate "I'll be 20 minutes late" messages automatically.
      • ... look up traffic patterns ahead of time and suggest "time to leave" or alternate routes (problem: GPS will not always be "on" in this next gen -- still a power drain).
      • If your friends have this software, you can vote on things to do (a quick IM-based decision tool), tell each other where you are, etc. You can see their calendars' free-busy time. You can share pictures phone-to-phone.
      • Combine GPS info with Google Earth-type data to build a realtime map that tracks you as you walk (
    • A Windows Mobile 5.0, GPS-enabled phone can do a ton of interesting stuff. Enough for 10 new companies. See details on Windows Mobile 5.0 (e.g. GPS, camera support) here.
    • Market Size?
      • Smartphones -- projected 2008 unit sales: 200 million (source)
      • This report says Microsoft doubled its market share in Q1'05 (to 4.5%).  Assume they get to 10-60% by 2008. That's 20-120 MM new units sold in 2008. Get your app onto 10% of that market: 2-12 MM people.  If they pay $30-$100 a year each, that's $60 MM - $1.2 B.
      • This is without porting to anything else (if they're is an "anything else" in 2008).
    • Own next-gen PIM services category


    • Hard to sell PDA apps (distribution = carrier deals)
    Cellphone-as-remote control
    • Enable Windows Mobile/Smartphones to act as super-smart home remotes.  Requires either that phone has decent IR or build a bluetooth-to-IR "bridge" that sits near the stereo equipment.  Download configurations for your equipment ( like Harmony remote).
    WiFi-enabled toys
    • Talking stuffed animals that download new phrases from the internet every day/hour.
    • e.g. Spongbob Squarepants doll talks about tomorrow's episode, then talks about it while it's on.
    • Buy a Squidward and they talk with each other.  Dialog's different every day.
    • Power Rangers that argue during fight scenes.
    • Get the TV shows to promote it.
    • New HP WiFi micro-PDAs coming out
    • Would probably need same power as an iPod.  An iPod with speakers.  Ideally, use Candela-style inductive chargers as seen on modern night-lights to avoid plugs/shocks/etc.
    Online dollhouse world
    • Lets kids describe a fantasy world ("a land with an evil dragon, populated by hobbits and talking squirrels, trees have huge flowers, clouds are cotton candy", etc.), choose the key buildings/castles/homes (using templates), then click "Go".
    • 3D world is built, complete with AI-driven non-player-characters.  Kids can wander around the world they created and interact with the characters.  Meet the mermaids, watch the burrow-elves sing.
    • A live fairy tale.  Like walking around inside your own "Shrek".
    • Use Valve's Half Life 2 or similar high-res 3D engine.
    Online game tied to TV show
    • An online multiplayer game tied in to a national TV show.  Kill the dragon, end up on next week's episode (your character in the show, your name in the credits).
    • It would be a massive, Survivor-like phenomenon, where lots of people can now discuss this one unfolding "show"
    • Quality and speed of PC-based videogames getting incredible: World of Warcraft, Everquest, BF2.  Networks, cable providers, etc. will see the size of these audiences and get interested.  Pay-for-play model works well.
    • Recreate historic towns (e.g. Mill Valley in 1928, New York in 1776, London in 1500) in a high-quality game engine (e.g. Half Life 2 Source, BF2, etc.).  Host worlds just as game servers are hosted.
    • Include driveable cars, ride-able trains, costumed non-player characters, etc.   See Half Life 2 for the potential.
    • Users can visit these towns (log in), walk around, leave comments on the buildings, trees, cars, trains, etc (e.g. "that building was more blue and had a small oak tree in front of the mailbox", "the Mt. Tam train was actually a 1915 Harville Steamer").  Geeks can log in later and fix any errors using mod tools.
    • Make money from trains and cars and building objects that are sold to collectors who work on improving the world.  It's like a collaborative train set "wiki"
    • Enable users to move through time (i.e. visit New York in 1776, 1800, 1965, etc.) by visiting different servers.
    Home Video Server
    • Problem:  People are building up massive home video libraries that they'll never watch since they're unedited, unindexed, un-labeled.  This will only get worse.  I have 70 hours of DV tapes I've shot.  Informal survey suggests most 40-something camcorder owners have 30-40 hours already.
    • A 1 terabyte hard drive could hold 75 hours of captured DV video at full quality.
    • Hard drives are getting cheaper fast.  An external terabyte drive is now under $1000 (i.e. $1/GB).
    • Could build a device and/or develop software to enable future high-end PCs to be home "video servers".  Sell software to Dell?
    • An ideal home video server device would have the following specs:
      • 1 Terabyte hard drive
      • MiniDV tape drive
      • DVD-R drive
      • Ethernet connection to network
    • Customers could pop in tapes when they're done filming.  The device would extract all of the "clips" from the tape (and auto-detect start/stop just as Final Cut Pro does), date the clips (from the data encoded there), and store to hard drive.
    • UI on PC would let you review a timeline showing all of your available clips.  Select a date and play all the clips from that date. 
    • Select a range of clips and the software automatically burns a DVD "movie" of those clips.  See "" for software that does the work here: adds titles, cross-fade transitions between clips, and music automatically.  No editing required.
    Instant Messaging device
    • Per Steve Glenn: Low-cost, wireless (SMS?) device that supports AIM or MSN messaging
    • Per Javier Rojas: Include a GPS to see friends, allow parents to track kids?
    • Danger article.  Danger online demo
    • Consumer-friendly, scalable videoconferencing add-on to AIM/Yahoo IM/other
    Wireless gaming
    • Wireless gaming article
    • Sorrent - wireless gaming startup, JamDat - mobile games
    • Keith Kirby wants to do a super-cheap wireless bingo game.  Play anytime.
    Home Movie editing service
    • Problem: People are building up massive home video libraries that they'll never watch since they're unedited, unindexed, un-labeled.  This will only get worse.  (see home video server above)
    • Upload (or mail) videotapes to an offshore editing service that delivers a 90 minute film for less than the cost of a portrait or a pricey funeral.  Under $30K?
    • Requires lots of cheap bandwidth (4GB/hour of footage).
    Real-time face recognition
    • Wearable camera uploads images of everyone in the room to server.  Server (or home PC) runs face-recognition against LinkedIn or Google.   Readout of who in the room you know, and what you know about them ("2 kids, names...").  Great for memory loss.
    Online band jam
    • Enable band members to play from separate locations (i.e. I can practice with my brothers in Seattle).  Requires very low latency connections.

    Other Links


    All contents 2005-2006 Ted Barnett