Seven Do’s And Three Don’ts For Creating New Web Products

Yesterday I wrote about how everyone else seems to be talking about how lame the new product ideas that people are developing are. I had at least one commenter over on Hacker News who said, in effect, that it was useless to comment on such stuff because being creative is *hard*.

My first reaction was, duh!

But my second reaction was that perhaps I could provide a better service by helping people actually think about the process of idea generation. First, I wrote a rather detailed article in January called how to create great ideas. In conjunction with this article, if you are looking to make a new truly creative product concept I suggest you read it.

With regard to my do’s and don’t please be aware they are guidelines. There are no hard and fast rules because there are always exceptions. But the rules, combined with the logic behind them should be helpful in thinking about how to actually create something useful. As discussed in my how to create ideas piece, the key element here is iteration. Your first idea may be interesting, but it’s rarely in and of itself good. Good idea development requires work and iteration usually with feedback from others in the loop somewhere.

Also note that some of the rules are mutually exclusive so clearly use common sense when interpreting the list.

The Don’ts

1. Don’t build a communications application.
Communications applications are deceptive. They have an incredible ability to get big fast because they are, done right, inherently viral. Having large numbers of users is always appealing to investors because there is often some fat cat corporation that feels they can monetize your communications app successfully. The tide on this I think is changing.

The problem is that communications apps never make money. You can’t charge people because then it won’t go viral. You can’t run ads because people don’t want ads in their communications application. Of course if you can do something easily (like twitter) and go viral I guess I would go for it, but if you do, you should be prepared to head for the escape hatch as soon as possible. Just be aware that as far as I can tell, for the most part these are not real businesses.

2. Don’t build a social application.
Most of the rules regarding communications applications apply here. Social apps can go viral, but they have all the same problems as communications apps except that they *can* be monetized if you get very, very big. The CPMs are very low and so running ads is very difficult until you get to a massive scale. Moreover, today the market is so saturated it is hard to imagine a social application succeeding if it does not hit one of the below “do’s” out of the park. Indeed of you can marry something social to something where there is a real path to revenue then you have indeed done something very hard and very good. One significant exception to this rule appears to be sex and dating related sites, although the market is very saturated right now so a really significant twist or a really narrow focus within that space is required.

3. Don’t copy someone else’s wildly successful application and think you can add one little twist. It has got to be a really big "oh my God" twist.

This is perhaps the biggest crime of all. Its not even the me too apps. It’s the me three, or me fifty apps. It’s insane. Yes we all will have competition but why walk directly up to the shotgun and demand they pull the trigger.

The DOs

1. Do become a domain expert in some domain.
If you are a computer science/programming oriented entrepreneur, there is a great temptation to try to solve general broad problems. This is because we tend to have no other skills. If you study programming that *is* a skill, but you need to marry that with some understanding of some real world issue. I strongly suggest finding an activity that has a problem and learning inside and out about the people that do that activity and what real issues they have that you may be able to solve.

2. Do solve a real world problem that requires solving an interesting computer science problem.
This may seem like a bit of a contradiction with Do #1. Really they can of course go hand in hand and should. But if you have an interesting solution to a general problem you may be able to do it in a broad general domain instead of a narrower one.

3. Do solve a real world problem that requires some unpleasant engineering.
This is where most real software companies make their bread and butter. Its doing things that are just ugly. This really breaks down into two categories. One would be something like BuzzWord, the word processor that was recently acquired by Adobe that is written in Flash. It’s a really general problem and they are not solving any deep mysteries. But creating something that has never been done that is clearly hard on a new platform provides some level of competitive defense.

The second category is where the engineering is tedious. Plaxo is a good example of this. They have figured out how to sync with a whole bunch of contact managers. It means staying up on data type changes in the products they connect to and generally doing stuff that no self-respecting software engineer enjoys doing. That kind of stuff is very defensible and often very valuable. Plaxo may be a poor example of making something very valuable, but there are rumors of acquisition talks. I mainly chose them as an example because everybody knows them. But there are lots of software companies making real money doing this.

4. Do build a tool that helps people create things.
Tools that help people create stuff are worth money to people that make money creating things. If you can substantially increase someone’s productivity it is worth real money to them and therefore to you.

5. Do build a tool that helps people manage valuable resources.
This is really an extension of #4. Helping people manage time or money for example is very useful. That does not mean people will pay for a calendar program, but they may pay for a project manager. Or they may pay for something like WebEx, which is an online meeting tool that helps people avoid travel. It is often deployed for helping companies sell without having to meet in person. Yes it’s a communications app but it solves a distinct real world problem that helps people save time and make money. And it is not dependent on being viral.

6. Do build a tool that helps people sell something.
If you are selling things, particularly real world physical stuff, you can make money. It is not the most glamorous activity, but it does work. In order to get to real scale though, it is helpful to not be in the actual business of picking, packing and shipping. Etsy is a great example of this, as is PayPal. Needless to say this is hard because lots of people understand that if you want to make money, standing someplace where people are actually likely to give you money is a very good thing. But taking money from people (legally) will never go out of style.

7. Do Get lucky.
No matter what, obviously some luck is involved. But the key here is to get yourself in the ballpark. Doing some me too mashup widget is not likely to get you there.


In summary, if you start out internalizing the principles I have presented above and you start iterating and refining ideas as I suggest in the how to generate ideas article, I believe you can indeed come up with something that is unique and valuable. You may even end up with some idea that will work despite breaking one of the rules. But its like my old english teacher used to say, it is OK to break the rules, but the key is you have to know the rules before you can break them.