Question for PG/HN: Where do you think we stand on your list of "ideas"? | Hacker News Reader

Question for PG/HN: Where do you think we stand on your list of "ideas"?

172 pointsjmspring103 comments

About three years ago, Paul penned his essay "Startup Ideas We'd Like to Fund". Three years along, how do you think this list stands up today? Are any solved (in part or in whole)? Anything new you would add to the list?

I ask, because I remember reading this years ago and getting inspired to work on a few things (that didn't pan out). Given the popularity of (and success of YC), it would be interesting to see what Paul and HN see as the state of the world for this list.

View on HN

Link to the essay -- http://ycombinator.com/ideas.html

When I go down the list I can only think of a few YC startups that have done well by following these ideas (dropbox, heroku). The leaders in the categories listed are not YC companies, and the big hits that YC did have were in categories not listed (eg. Airbnb).

Airbnb = Craigslist competitor (#25)

More like Craigslist spammer.

I'd put Airbnb as a competitor to hostels, they're not really competing with Craigslist directly.

Funny, because they seem to get quite a few referrals from craigslist.

Wikipedia gets quite a few referrals from Google. Are they direct competitors?

That depends on your view of Knol.

It's easy to look at the problem with Craigslist and think that to challenge it you should build something which does everything that Craigslist does.

AirBNB is a competitor to Craigslist in that it is helping to take it apart piece by piece. Hopefully other YC companies will take on other categories / pieces

They're competing with a specific part of Craigslist, namely short-term accommodations.

Other companies are competing with other parts of Craigslist (e.g. StubHub)

We built something alot like: 22. A web-based Excel/database hybrid.

The question is though, what problem does it solve? It was very hard to explain to people why they would need it or care about it.

Spent a year developing it and spent alot of money on it. For nothing in the end. A very cool technology and I still like it alot.

Lesson: build something people want, or that solves a real problem. Don't build a solution looking for a problem.

From the 1-line description it sounds like Wufoo took care of that one

Wufoo doesn't look like a database/spreadsheet hybrid to me.

Wufoo solved the "mail a template spreadsheet, receive filled in spreadsheets, manually consolidate results" problem, which is huge in any company with more than a hundred or so employees.

Link?

The software demo starts at about 1:30

http://www.youtube.com/watch?v=yuSDU0JiI2c

Stupid name - NeoGopher.

The basic idea was that with 10 to 15 lines of Javascript you could bring back results from any XML or JSON web based API and display it as a list. You could then link any row or cell to any other list.

In a way it was like a reinvention of the old "gopher" protocol which allowed the user to navigate from one data list to another.

The site is gone now.

i like it by the way. thanks for sharing it here. i can kind of see what you are thinking.

Are you kidding? 80% of 'enterprise' systems involve a spreadsheet that is (conceptually) a database. Some business users use the words spreadsheet and database interchangeably. Almost every enterprise system I've build involved Excel as one of it's data sources. Users often give you requirements in terms of excel formulate etc, or ask for screens that are 'just like excel'. Microsoft has even recognized this and now offers 'server-side excel' as part of office 2007. http://en.wikipedia.org/wiki/Excel_Services

I'd suggest that if this product wasn't a success you either didn't understand how end-users currently use spreadsheets as a 'poor man's database' or the marketing let you down.

No doubt we did a crappy job marketing. Also ran out of money before being able to implement the full vision/feature set.

We built it a a desktop application too which I think was a mistake. Getting people to download and install things is hard.

Or competing with Microsoft in this domain is still extremely difficult. As you point out, Excel still has very strong mindshare.

Realistically it was more like editable gopher than excel - or that is what it was meant to be anyway, we didn't get to implement the editing functionality.

I still very much like the idea of being able to build lists from web based API's with only a handful of lines of Javascript.

Wait.. what?

You built something that was supposed to be an Excel clone, but you didn't implement the editing functionality?

Isn't that pretty much 99% of the point of Excel? I mean yes, people do use Excel to share data as well, but most scenarios involving shared Excel spreadsheets involve multiple people editing the same spreadsheet, and then generating reports out of it.

That's where the pain is, and that is what I had assumed you were aiming at. If not, then what did your system do exactly?

(Also, build lists from web based API's with only a handful of lines of Javascript isn't Excel, either)

Edit: I see further down you noted that your system was desktop based software. A web-based Excel/database hybrid isn't a big requirement list, but I'm not convinced you are covering a whole lot of it..

My comment yesterday seems relevant here: http://news.ycombinator.com/item?id=2714515

You might like http://hypernumbers.com.

Um, what about DabbleDB? I'm not sure there was anything bad about it. Except that one day it just went away.

http://blog.dabbledb.com/

It might be that the lesson pertains here, too.

I think if it was built by a small, low cost team it might still be around today. Maybe?

It was built by a small, low cost team. At the middle of last year it was run by four people, but then the company got bought by Twitter: http://blog.dabbledb.com/2010/06/140character-dabbling.html

Pretty sure DabbleBD shut down because the company got acquired mainly to get the talent and not for the service. Cannot recall who acquired them.

I only ever played with it, but it seemed like a great piece of work, and was all coded in Smalltalk (Seaside).

People do need an alternative to Excel, but it must address the issues Excel has, not compete on the same ground. One example: data consolidation. Try consolidate 30 excel files from 30 different people across the globe and you will be asking why nobody is providing a better solution.

As an analyst/controller I work with Excel every day, let me know if you want to know about issues with Excel, there are tons of problems to solve.

Would love to hear about them. Mind emailing me?

No problem, when I have time I send you an email. I would like to discuss this myself to see if there are cheances to build something in this direction.

I would love to hear about these problems, too. I'm currently planning (as in not having started programming yet) a web app that allows people to manage all sorts of data. Which of course would also include financial data. I'm always looking for new use-cases.

Interesting topic as I recently spent three months on customer discovery for a CS researcher/wannabe entrepreneur who developed a spreadsheet consolidation (combining data from disparate sources) solution. My discovery work was primarily focused on enterprise. The primary finding was that everyone in enterprise claimed that they have the problem but the pain is not high enough for them to consider buying a solution that only consolidates data. They were looking for end-to-end solution.

Have you looked at google fusion table and google refine? These are powerful tools in same space for non-enterprise applications.

Do people really need an alternative to Excel? I would suggest that as soon as someone has gotten to the limits of Excel, then they would start looking at a programming language as an alternative, not at a "more better Excel".

Andrew, have you ever worked in a business environment? People get scared for some humble sql, how the hell are you going to convince them to learn programming (also vba is a problem)? Plus, what about cost/opportunity? Would you prefer to pay 10/20$ per month per person to improve their excel efficiency or either paying custom code or coding training?

Point is, there must be a market out there for some functionalities that Excel does not have, it also should not be that difficult to build such a solution (call it a hunch, but I think with nosql that should be really easy to do). But the whole thing requires stopping thinking logically (of course excel makes no sense) and start thinking business(ly) (how do I get people to understand and use my solution).

You're both right. Spreadsheets are already a (poor) programming paradigm called Functional Reactive. I say it's poor but it does something amazing: it doesn't require structure, planning, and forethought; and it can be used by anyone. The problem is that these constraints pull at each other: make it more powerful and you make it less useful.

I think it's only when you make a complete rewrite from the bottom up in another paradigm that you can extend the power while retaining the usefulness.

'I think it's only when you make a complete rewrite from the bottom up in another paradigm that you can extend the power while retaining the usefulness.'

True, but again, look at the users. People with very poor IT knowledge and very scared of change. Any new solution pointing at this market will have to:1) be as easy to use as excel; 2) much better than excel (why change otherwise?).

I think that being better than excel overall is impossible, but looking into specific businesses and uses it should be.

I think that being better than excel overall is impossible

We disagree.

David, if you have something better than excel in mind, I'll be glad to hear that.

Personally, I have been giving it a lot of thought and I can't come up with a complete better solution. But other (Palo, Hyperion) seems to make money on solutions that do not change the excel paradigma.

I would be happy to show you what I'm working on. Send me an email: david927 at gmail.

The key question is what are the users trying to accomplish with Excel? There may be ways for them to continue to use Excel and gain the benefits of a more complex reporting environment.

The base of users with knowledge of Excel is very large. Most organizations have a built in base of knowledge of Excel. Why swim upstream and go against that. Try to make Excel work in that environment.

Take a look at www.excelcube.com. It may be helpful.

We have written a new spreadsheet and spreadsheet consolidation is one of the core cases.

http://hypernumbers.com

Our core business case is to address the problems of distributed business processes implemented in spreadsheets. We want to continue to offer the flexibility of the spreadsheet paradigm whilst addressing the most weaknesses of the current paradigm.

This short video (when its finished) will be the centerpiece of the new release: http://vimeo.com/25937882

That's funny because I'm also building a web-based Excel/database hybrid, and I'm also a bit concerned about demand. But your story doesn't deter me at all, for two reasons: one, because I'm crazy; two because of La Femme Nikita. Let me explain.

People talk about ideas as being worthless, etc. No, they're not. You can't get an oak tree with a mustard seed. But there's a big difference between even oak trees. Idea and execution are the same thing, just two ends of a continuous line.

The films La Femme Nikita and Point of No Return have nearly the exact same plot. The former is one of the greatest action films ever made; the latter is one of the worst.

That whole "build something people want" is only for people who want to make money. And honestly, if what you want is to make money, be a dentist.

People waant Facebook and Twitter and Farmville. But they need core technological innovation. They need better ways to work with data. You did the right thing, Andrew. It wasn't a waste of time. And I won't stop either, because the world has far too much of what it wants and far too little of what it needs.

Mail me when its ready to show, I'm interested andrew.stuart at supercoders.com.au

I think I understand your point. AltaVista is nowhere and google is.....

I am also a fan of the original (the french version) Nikita - great shootouts.

Thanks, I'll do that, and my email is david927 at gmail.

Actually, Facebook, Twitter and Farmville are great examples of products nobody said they wanted and didn't solve any perceived problems, but are succesful any way.

I've compulsively weaned myself off the "solve a problem" mentality because it gets in the way of building truly innovative products, IMO.

You're right. You don't have to solve a problem. But you do have to create something people want, or something that appeals to them. Technically, there aren't many "problems" in life besides survival. Everything else though has to appeal to something in the Maslov hierarchy

Couldn't the problem be that people just have way to much time on their hands and need help with that? Haha, seriously!

And honestly, if what you want is to make money, be a dentist.

And if all you want is to build something just for the fun of it, don't start a company. You'll lose money, time and nerves. There really are only to options when building something. Either build something people want (and if somebody built it before you did, build a better version - see: Altavista, Yahoo and others vs. Google) or build something, and then make people want it (see: iPad). But don't start a company to build something just so you can satisfy your professional curiosity. That's what side projects are for.

The web was a side project. It changed the world. Linux was a side project. It also changed everything. Angry Birds was a side pro... wait, no it was a way to make money... and that's fine. I'm not trying to say, "Don't make money." I'm saying, making what people want gets you Angry Birds. Making what people need gets you the web.

So your point is that you should build something that you believe people will need even though this need/want doesn't yet exist. Ok, I can agree with that. But if you're going to run a company, you better build something people want now, or be able to make people want it, otherwise you'll go bankrupt.

And I can agree with that.

I am currently building/have built a web-based spreadsheet with a really unique approach - something that Excel does not and can not do.

I will be presenting this at the Seedcamp in London later in the year. I am really exciting about the project, so I wont be giving anything away here! Unless PG wants to get in touch directly...

A Python company just outside the city, whose name escapes me right now, was doing very well with a web based excel.

Most/many hedge funds manage billions of dollars on the back of very complicated excel spreadsheets, mostly because the "talent" understand excel and can bend it to their needs.

But auditing these things is an utter nightmare, which version was used to lose a million here or a million there. Is it still on bobs laptop? How can I reuse code?

By providing a excel like front end and a centralised, versionable, auditable back end, they got a lot of interest going.

The space is still very wide afaik, but I have been out of the City for some years now.

(In fact you could take a similar approach to any CFO's office)

Jane Street is very unusual in having such an advanced technology management set up (IIRR OCaml, directors review the code before production etc etc)

I think you mean Resolver Systems. The web based spreadsheet is not their "core" product, though: they started with a desktop spreadsheet app.

DabbleDB was pretty impressive but also appears to be shutting down.

I've designed such a system a couple of years ago, too. When users complain about Excel, now that sounds to me:

They're really saying, "I love Excel".

Don't forget it's only a tip of iceberg. Users will complain more about missing features after you built something users have cared about.

Not just Enterprise software 2.0. but Localized Enterprise software. The idea to adapt, say, SAP for Russian or Indian or Chinese laws, regulations and actual document flow is totally broken. Hosted services also limited to US market.

The problem is - it is almost impossible to get funding for such projects in domestic markets.

So, it is a very simple and obvious idea to build Enterprise software using the same technology that powers modern global-scale online services, targeting emerging markets, but it is quite long and very expensive journey.

There are huge demand for hosted niched services for small and medium business, such as tour operators (schedules, synchronizations), goods carriers (simplified logistics), B2B systems (customer - supplier) and so on.

Something I've been pondering lately is whether there's a viable niche in internationalization as a middleware library & service: it takes a lot of effort to get right, it's tedious stuff (as I discovered when trying to document postal address formats for most of Africa and Europe), and it involves a lot of domain knowledge that's completely irrelevant to the actual core business of most SaaS vendors.

Yes, that is why it should be done by domestic startups and teams 'right on the site'. So, YC's model doesn't work here - they fund only US-based startups (because it is quite difficult to fetch a profit from aboard =) but this one must be an international one. (same as that network of Google offices around the globe).

What I mean is whether there's a niche for a single entity whose job it is to be knowledgeable about all the sorts of issues you can run into in trying to internationalize software and services, whether they be local tax issues, best practices for translation layers in a software stack, cost estimations for freighting products.

Like, i.e. an all-in-one service: build your SaaS platform against our framework, and we'll handle all i18n issues on this given list, for this given set of countries and languages, including payment handling and shipping, and take a cut. A successful startup often doesn't have time to deal with its own internationalization before local competitor clones get the jump on it (e.g, Ebay, Groupon). So what I'm envisioning is a company that acts at a higher level than the usual outsource-to-the-translator-agency routine, yet has product-ized its offerings enough that it's not locked into the consultancy business model.

There is definitely a market for this. International payment gateways like CyberSource (www.cybersource.com) are extending their services to include address validation (and correction), fraud detection, etc.

It would be very hard for a startup to compete in this space though - perhaps a better opportunity for a partnership between a payment gateway and a major distributer.

YC has funded non-US based startups

My opinion:

- Simplified browsing has happened ala tablets, mobiles etc, especially purpose-specific apps.

- Photo/Video sharing services .... they went mobile, big time

- CRM ... I think twitter, facebook and internet outrage are changing that pretty rapidly

- Web office apps ... I (still) don't get the sense there's any major urge to move to web office apps, but obviously the 800 pound gorilla in that space is (went?) web

- Online learning ... Kahn Academy

- New payment methods ... Square, NFC et al

- The web os ... kind of became redundant with the explosion of mobile and tablet, but Chrome OS might end up mattering

- Application and data hosting happened in a huge way lol

re: - Simplified browsing has happened ala tablets, mobiles etc, especially purpose-specific apps.

Mobile and apps certainly have helped. But, there is still a problem of knowing what urls (in the case of a browser) or sites to search or which apps to get to help out.

It seems like there should be a better way of just "getting information" based on location/interests/curiousities...

You mean google?

Google may be an answer, but it is still oriented towards typing in search terms, maps, etc. Is there a better way to package this up as a ui/service?

Seems like you are looking for an App Store for SaaS apps, like the Chrome Web Store, but more general purpose and less Chrome specific.

Edit: as much as I despise the term, Social App Store. You would need algorithms like Pandora uses, that tracks how often and for how long a user uses certain kinds of apps, and through categorization, tailor a "Recommended" page to popular apps in the same category, as well as ones that "friends" use on a regular basis.

Semantic web. Hunch is trying, other startups as well. You need so much "opt in, we will find cool stuff (after 200 questions)."

The big boys of online data preferences who Could do these things would be Google, perhaps FaceBook (which still brings up ridiculous unrelated ads to my interests, however, I don't "click" any apps or "like" many pages out of privacy). More on FaceBook, even my "prolific" friends (checkins, 200-400 friends, likes and friends with 100 or so movies, bands, apps, etc.) still don't click ads.

NetFlix seems, after 4 years, of finally getting a semblance of "interests" down.

That is quite a bit of data, time length, and me, the user actually rating the movies I watch in a semi-diligent manner.

I would, in fact love to log onto Google, or wherever, see my "bored" category of movies I may like to watch online, at the theater, local events going on that evening, something coming up tomorrow and all somewhat related to what I would enjoy, not just "pushing" or scraping a bunch of different calenders / pooling app info together.

Very difficult problem.

How about Apple? They have access or the potential to access a whole bunch of info through that little touch screen life portal everyone clings to on a daily basis. Add in iCloud and their Nuance partnership and you have something pretty interesting.

"Hey iPhone, is there anything fun to do after work?"

It grabs your location, sees you have a late meeting blocked off from 6-7, sees you were watching Transformers 3 trailers a few days ago, sees your buddy Mike tweeted "Transformers 3 looks pretty rad, I definitely want to see that".

It now has an approximate time and location, a pretty good idea of what you're interested in (both long term and short term) based on app data, and can see if any of this aligns with your social graph...

Would you be interested in seeing Transformers 3 at 7:30 with Mike?

"Yes, send him an invitation." iMessage sent with all the details, when you get a yes from him, Fandango launches and you buy your ticket, it gets added to your calendar so you get a reminder after your meeting, maps out directions, etc.

It's a stretch but if anyone can make harvesting that much of your personal info look appealing, it's definitely Apple.

Stopped reading on "- Simplified browsing has happened ala tablets, mobiles etc, especially purpose-specific apps."

How is the answer "mobile" the answer to anything? it's like hearing microsoft saying "cloud"

About purpose-specific apps, how this even get close to simplified browsing? It's the opposite! What to read website X on mobile? wait 7 months until they launch an app. And if they already have an app, you will probably have to go to the desktop to do something trivial (like adding a video to a list on youtube under android... you will have to favorite it so you can find, go to the desktop, move the video to a list)

My mobile browser for sure does not make anything simpler! Took me several minutes to import a bookmarklet to change the background to black and text to white on the current page. It involved setting up a webserver on another computer because it could't open pages on the local memory.

Purpose-specific apps give them what they want without the hassle of the web - browser incompatibilities, plugins, codecs, address bars that might as well not exist, malware, scams.

Remember that time a blog accidentally outranked Facebook for 'facebook login' on Google and people couldn't find Facebook anymore? The web is hard for a lot of people.

And instead now they have 200 unofficial craigslist apps that do god-knows to your phone.

We are back to the BBS days, everyone's downloading binary files!

I realize you are posting your opinion, but you posted examples of past execution, NOT future innovation. I would argue that each of your "opinions" are ripe for disruption.

Disruption on the scale of displacement happens all the time. And mark my words, it will happen to each of your opinions listed above.

Of course .... not accusing anyone of having solved those problems forever, not to mention tomorrow's answers and tomorrow's problems may not be anything like today's (or in this case, 2008s).

I'm tackling the web OS...

And so I don't think it's going to be Chrome OS, also because it has to work on any kind of device to be the web OS not just a Chromebook and too many partnership deals would need to be signed for Chrome OS to spread as quickly as the web OS needs to spread.

I'll take the opportunity to post about my recently started (2 weeks now) project on "design dependent search" (idea 16).

The idea I'm trying to leverage is that typing long search queries on mobile devices is super awkward, but the touch screens make clicking much easier and more fun. My site displays similar search results geographically near each other on the screen, allowing you to make your searches more focused just by clicking on certain areas. Right now, it only works on blogs with RSS feeds due to my limited internet crawling resources. I hope to add media (picture,movie, and maybe music) search by the end of the summer. I don't see this necessarily working as a Google replacement though.

You can check out the prototype at http://blogumbus.com and visit the developer blog at http://dev.blogumbus.com.

Interesting. I had a play with it and I think there is some potential there. Its somewhat similar to http://www.chunews.com/ which I wrote some time ago in that you can follow the links between similar articles and paged. I like your implementation though.

How are you working out whats related? Links or some deeper textual analysis?

Thanks for the heads up on chunews. I was having some trouble finding similar projects to judge where I should focus development. How many users did you get?

As for linkedness, I'm going to do a detailed blog post about it sometime, but here it is in short:

Use the HITS algorithm to get hub and authority scores for each domain. Pages are related only if websites with higher scores link to both of them. This makes spam much less of a problem (I think!), and will make for some interesting search algorithms that I haven't quite implemented yet.

I was more my attempt at recreating TailRank when it closed down (since I assumed that it couldn't be that difficult to write). It just seems to overlap what you have there quite a bit.

Not many users. It tends to get a lot of SEO traffic though (usually breaking news stories funnel through it pretty quickly and you see a large spike) with anywhere from 50 to 500 uniques a day.

Ah interesting. Chunews does statistical word ranking on each article it pulls down to group similar stories together. From memory it uses links as well but in reality since it checks exact links that never affects this.

One of these days I may iterate on it but for the moment it shows me whats happening in the world at any point which is fine for me.

I find it quite hard to quickly visually scan the results. And I'm not getting very good results for the queries I tried. Are there any specific type of queries this is good for?

Considering it's a search engine, you should probably not hide the search box behind an extra click.

I called it a search engine, but it is not supposed to be a google replacement and do general queries. It only finds blogs similar to ones you're already interested in. At this point it's more of a fancy RSS reader with a twist.

I've been wondering whether or not it's best to hide the search box. I decided to hide it because I don't really want people to use it. As you've pointed out, it's not really that good. I'd rather people find blogs similar to ones already on the screen by clicking. That's where the strength of my display comes in, I think.

I wanted to experiment in new search interfaces, but I thought that taking on google directly would simply be too hard. That's why I didn't jump right in to general purpose search.

Dating? Anyone think the chicken & egg problem has been solved or is being solved? Any ideas how to solve this?

build it on top of facebook or twitter

When I look at the history there is a clear pattern:

- Match.com solved it by starting in 1995 and working really hard.

- eHarmony solved it by attracting the christian crowd with endorsements by pat roberson and great offline marketing.

- plentyoffish solved it with a great brand name, and using the "free forever" look, feel, and marketing.

- OKcupid solved it by creating tons of questionnaires that were fun and interesting.

- AshleyMadison solved it by targeting the infidelity niche.

- AdultFriendfinder solved it by marketing to suckers on gaming and porn sites.

The pattern is clear. There is no pattern.

Don't forget apps like Grindr. It's hyper-local, which is a whole new niche in this space.

There is one pattern: a lot of marketing.

Tastebuds are still getting started, but have an interesting approach of matching based on people's music taste and Last.fm profile.

I thought the pattern was fake profiles.

OKCupid did something great by making the questionnaires entertaining to fill out. (As opposed to eHarmony's, which are a slog) I know some people that are just on the site because they have fun answering the questions and poking around.

Makes using it a drag, because lots of people never reply to any of their messages, but make it fun still seems like a good place to start.

smartphone/location-based ... e.g. "these 3 single women are at o'malley's pub. want to meet one?"

solve chicken and egg by giving the women free drinks or food when they get to a place. this helps the dating network AND the bar by attracting more people to both.