A story to share: My wife worked for a local Land Rover dealership for a short time a couple years back. I visited one day and saw the computer system they used and was shocked at how ancient and horrible the interface and hardware were. For a dealership selling 80K Range Rovers I thought it was downright crazy that they were using software that looked like it was from the 80's.

I realized however after talking to her boss that I couldn't just waltz in there and tell them I'd solve all their problems with my new shiny software app. I needed a connection, several steps up the corporate ladder to have any chance of landing that gig. Which in turn would require years of working within the corporation or at least experience in a related field to give me credibility (and designing/building web 2.0 apps doesn't count)

Something so simple to me was in fact tied up by a very traditional model where "who you know" is more important than what you know or what you can deliver.

Some uber geek walking in to your industrial fishing plant and offering to re design your software (for a fee of course) is not going to be welcomed with open arms. Pre-existing relationships are very important and still play a major role in where money gets spent (software industry or not).

My wife has the same problem. She's a public school teacher, and their software makes my eyes want to bleed. They still use a text-based GUI for their attendance system.

And they've got a piece of software to input your lesson plans, but you still need to print it out at the end of the year.

However, the school and the district have leadership problems, so software should be a little lower on their priority list.

I've thought about hacking up an easy classroom/school management app that would be similar to WordPress. Install and go.

There is actually some (relatively) good software out there for public school teachers though. Some are even web-based, SaaS models.

The startup I work for :)

http://schoolrack.com

We're extremely (this week) close to launching student and parent account types, so students and parents will be able to easily collaborate with teachers via the already existing features we've got, plus we're going to add a gradebook, discussion board, and quite a few other features.

I just submitted something related: http://news.ycombinator.com/item?id=205295

That's basically what I'm doing right now: walking in and making my pitch, and it doesn't go well. What worked was we made a software app for a company in one industry and got a call from another company asking about it. So next month I'll probably just sell this one app and make free money of it. I wonder if that's why it's so rare to find a true custom software company...

This is what business partners are for.

I've considered entering the medical billing industry on several occasions...because my sister owns a medical billing firm with several small to mid-sized clients in Atlanta (large clients have in-house billing departments). And the software in the field is atrocious and expensive. But I wouldn't consider it, if I didn't know just the right person to partner up with that would allow the software to be used immediately for real work with real clients.

Such a partnership doesn't need to be 50/50. You can certainly find someone who is experiencing pains in the industry you're interested in. Invite that person to dinner, bring a note pad, and figure out the smallest functional piece of their infrastructure you could replace that would solve a real problem better than existing tools. Once you've done that, get commitment from that person that when you build it they will use it (not buy it, just use it, for real).

From there, keep working to replace pieces of the existing solution with better ones, while marketing directly to the people who need what you're building. Every new customer will increase the word of mouth network effect within that particular industry, and when they love your first widget, you can then start campaigning for them to buy your second widget. Lather, rinse, repeat, until you've replaced every component in their existing system with a better product.

Another way in is to build ancillary tools that make aspects of the existing solution nicer. Think Lotus evolving from the charts plugin that Mitch Kapor wrote for Visicalc.

"Can you learn an industry without working in it"

Why would you want to?

Maybe I'm in the minority here, but I write software to enable me to do other things, not just for the sake of writing software (although that's pretty cool, too).

I've worked in many different industries and the software needs are astounding. Between what you will witness and what people tell you, you will never lack for ideas or opportunities. I can't imagine sitting in some ivory tower (or cubicle farm) trying to come up with ideas and building solutions. I have to get out there and see what's happening.

Make no mistake about it, I do enjoy sitting and hacking, but only as a means to fulfill much greater ambitions. I can't wait to get out there and implement. For me, that's way more fun than writing the code.

If you wrote code and no one ever used it, would you still be happy? For me, that's a resounding "no". Probably because I've spent so much time in "industry" that I realize there's still so much more to do.

Ask a different question: Where can I find a place in an industry that would give me a better chance to blossom?

No, unless you can find some information independently (via research or people). That info is hard earned and you will be a lot less convicted (and it will be harder to convince people that you're right) if you have no direct experience.

If I was a hacker, I'd be looking for smart/irreverent/idealistic people in industries with big problems to partner with. That sounds a lot more fun.

See Fred Wilson's post from today.

I think this very point is why there are soooooooo many "productivity apps" and "music reccomendation" apps. All software people listen to music and want to be more productive.

What about fashion? What about auto mechanics? What about schools in India? What about transportation in rural China?

I strongly recommend Umair Haque's article on "Hacking the Industrial Economy":

http://discussionleader.hbsp.com/haque/2008/05/last_week_i_a...

Great question!!!

I think one of the best ways to break the barriers software developers/entrepreneurs face in understanding industries outside of the normal niches is to sit down with people in the industry and have a therapy session of sorts. Don't sell anything, but instead listen and learn carefully to what their concerns and problems are.

People in industrial fishing may not think that they have a need for new software, but there are undoubtedly needs and issues within the fishing industry that are completely opaque to those outside of it. And where there are pain points there are opportunities to innovate.

Yes & No.

Yes, because, you don't come with the pre-conceived notions, assumptions that hamper that particular industry. you come with a fresh set of eyes, you ask why a lot and come away with a lot of interesting observations.

No, because, you don't have enough data points; meaning what you observe could be representative of one company or the entire industry.

To the how part of your answer. Work at a job that gives you the data points. You could speak to the founders of all the companies in the hospital health records and HIPPA compliance industry for eg.

Yes, you can. It may be easier if you do work in it, of course, but even that doesn't always help, because you may be mistaking your particular problem with a more general need.

The trick (if you can call it that) is to speak to a number of people in different roles in the chosen industry, and try to elicit different problems that need solving. Then, run the results by the others, and listen carefully. Read the trade journals, and see what solutions are out there, and you'll see what the gaps are.

I have to agree with this point. No, you do not have to be part of an industry to do work in it. It will certainly make the proposition very difficult however.

One of my YC ideas was a product aimed at the HR industry. I personally have no affiliations with it but my co founder was a part of it. While my ideas were very interesting from a freshness perspective, I was often was off from what our client base wanted the most.

The benefits of being part of the industry is having connections, knowing where to look, and what to look for. If you are not part of it, this can be time consuming and costly. It's extremely beneficial to have a co founder, or just an adviser with industry experience. Like Michael's comment, their leads give you many heads to run ideas by, and the resources necessary to understand your clients.

The ideal situation in many ways is to partner with somebody from the industry. As you say, this can be a great source of domain knowledge, and contacts. A domain expert with business skills makes a great co-founder; alternately, a trio of techie, domain expert, and business manager.

I would say yes, but it really depends on the industry and situation.

We entered the powersports vertical because we saw that the best of breed ecommerce sites really were not cutting it. We planned to differentiate ourselves on technology, infrastructure (due to our technology) and user experience. We are about a year in and are exceeding. We came in with really no experience in this market or industry. See my past comments to see what startup I am referring to.

We did however, spend a lot of time researching our market as we went a long, so their certainly has been a learning curve. However it was not the most difficult obstacle we have faced.

Quick iterations on our product / agileness in our software and listening to customer feedback has been key.

Feel free to drop me a line if you have any specific questions.

The answer to this is that it depends what their problem is. There is a lot of horizontal stuff that is applicable to any industry. You could get a lot of experience building (say) workflow software in one industry and reuse that knowledge in any industry. This is the classic systems analysis approach.

But if that industry uses software to solve problems directly, then you will need domain knowledge. You could be the best C programmer in the world, but without the domain knowledge all you could offer (say) a biotech company would be help in optimizing their code or writing generic functions such as the user interface or the data access. How could you write code to simulate protein folding without knowing it yourself?

I think you are asking a very important question for any entrepreneur which is "how do I go about finding a substantial business opportunity upon which I could build a viable enterprise?"

Recently I wrote about "entre-sumer" which I believe is an important cornerstone for bootstrapping startups. Basically I am arguing that we need to refine the "Build Something People Want" concept further by narrowing it to "Build Something I want".

That is, to maximize our chance for success, we need to be the "entre"preneur who creates the solution and at the same time, be the con"sumer" who has a burning need for such solution.

So in that sense, the Founders of Google, YouTube, Facebook and Twitter, etc., were doing just that. They were the representatives of their generation and they were the "alpha" dogs. By eating their own dog food, their success is assured since by definition there are plenty of "beta" dogs who would follow.

But rather than saying that these are "software-related" opportunities, I would say that these are E2C opportunities, building an "E"nterprise to sell to "C"onsumers.

My own opinion is that while these are interesting and highly visible opportunities, they have the fundamental problem that more often than not, you will be competing with "free". It is difficult if not impossible to get consumers to pay, however much they might live to depend on your products and services. So typically the viable business model is advertising, i.e., monetizing eyeballs.

The other side of the coin, and I believe this is the gut of your question, is E2E, building an "E"nterprise to sell to other "E"nterprises.

Then the answer to your question is YES. Someone in your team would need to come from that industry who can be the "surrogate" customer.

My experience with selling to enterprises is that customers are actually not the place you start when you build a company from scratch. They know how to criticize an imperfect product but they have no idea what to do with a perfect PowerPoint. You need to have someone inside your team who has the insights to guide your vision. If that person is you, it is the best.

Hope this helps.

I think that not working in an industry often gives you a unique perspective that might as well make you more likely to spot a particular gap its regulars are taking for granted.

Of course you will need some insider knowledge, but oftentimes not all that much - a lot of industries develop similar approaches to solving similar or analogous problems, but with varying access to the underlying knowledge/skill needed to making it effective.

You will need to be a bit of an anthropologist and a good communicator to get the relevant information from people whose working days you plan to optimize/improve with software.

I'd say observation with an unspoiled eye can work wonders :).

In my case I didn't work in the industry that I'm trying to serve now, but instead realized that a solution from an industry I did have experience in could be applied to another one.

The problem is that you're guaranteed to get it wrong at first, simply because you don't fully understand what people in that industry need or want.

What we did was build something, show it to some people, get feedback, improve, repeat the process. We're in our third iteration now, and I believe that we're closing in on a real solution. Of course, if it was a solution for the software industry, we could have been there much earlier...

Of course. I have a hobby that is also practiced professionally by a lot of businesses and I have been thinking recently of building a software package for the smaller providers since I have a lot of insight into what they need. Because it's also my hobby, I'd be eating my own dogfood and can see what works & what doesn't.

Can you learn an industry without working in it? My boy, that, or rather the converse: that you can work in and manage an industry without learning it, is the great promise that MBA programs make to their students. The idea that there are abstract principles of management, of dressing down subordinates, of bringing in new business and cutting costs (that's all that CEOs do, essentially) that free the MBA from ever having to dirty his or her hands with the business--their primary, all-consuming interest, they protest--is for them a feature, not a bug. It's business processes all the way down...

If there is an industry that interests you, go to a few trade shows in that industry space, you can meet the key companies, see demos of all the latest things, learn the lingo and so on.

Most professional trade shows aren't cheap (expect to pay 500-3000 dollars) but they are probably the most time efficent way to get to know a new industry.

This also works great as a creativity boost, learning about different problems and different solutions can generate a ton of new ideas for your domain problems.

You don't have to "learn an industry", you only have to be made aware of a problem in enough detail for you to solve it.

That can be done by learning an industry, but that's the expensive way. Cheaper is to have broad first-hand experience, or a large network of people who will tell you about problems that need to be solved in their industry.