In Good Ideas Take Time or How to Brainstorm - Part 1 I wrote about how I come up with ideas for blogging. Today I'd like to take that a step further and talk about how I generate ideas for business.

The first step is to have ideas, to find ways to get into that creative mode that let's you generate ideas. For me it's often a cup of coffee and a window to look out of (preferably a Picard window, but anything above first floor is good) when I'm thinking alone, or good iced tea over a 2 hour lunch when I have someone to discuss ideas with. A lot of this is just be willing to take things one step beyond possible/doable/likely. Be willing to 'waste' a few minutes because sometimes it will open unexpected doors, sometimes it helps you implement part 1 with a greater level of confidence because you have a idea of what step 2/3 might look like if things work out as hoped.

Next, realize that many of your ideas will be bad. Bad because they are unworkable, too expensive in some way, not enough time/energy, or just bad. Perhaps 1 of out 10 ideas will be good, maybe less. That's not cause for concern. Generating ideas is the goal, the absolute worst case is you come up with 20 ideas and they all get tagged as unusable. So what, you spent an hour working on it, and you're learning to be creative.

Write down your ideas as you go. I've tried using a voice recorder here in the office and it's sort of useful, but just writing down ideas makes it a lot easier to work with post thinking, compared to listening to an hour recording to make the list later. It's definitely a challenge to keep the idea flow going and to write at the same time, especially when you have two of you thinking and going off in different directions. Do what you can, occasionally stop to clarify. It's hard work, you're trying to listen, think (deeply) about what they are saying while at the same time your brain is also trying to pursue a separate avenue. Just keep going.

Lot's of ideas sound good when you come up with them. In general don't go implementing them the same day. I like to let them sit for a week or more, and then go back over them. Which ones sound like they would really add value? Often you need to run the ideas by someone who wasn't part of the brainstorming to get a sanity check. The trick to this is not take much ownership of ideas, because that will lead you to do things because they were your ideas. Remember "I love it when a plan comes together"? Easy to get excited about an idea, doesn't mean it's a good idea.

I think most good ideas are incremental, and often you find out that you've invented something that someone else already thought of - but how would you have known that? Here are a couple exams from my own past:

• Back in 2002 or so I was looking for ways to increase the number of authors for SQLServerCentral.com. Ultimately that led me to think about how could I make it more rewarding for them to write for SSC? The first idea was to publish a book with their names on it, and eventually that became the 'Best of SSC' which was a repackaging of already written articles. Not a huge commercial success, but definitely was a way to reward and celebrate authors. If you look deeper, you'll see that we were essentially reselling content people had already seen (for free). Most magazines do exactly the same thing, publish special editions that are made up of selected articles from the past year. Easy to see in hindsight, but then we were trying to solve a problem and create a product. Long, slow process to implementation.

• Yesterday we were looking at our numbers from JumpstartTV and one thing that interested us is that we still have a decent number of people that sign up, then never click the confirmation link (crucial to prevent spam). Takes a lot of luck/work to get someone to register, heart breaking to then have them leave without ever watching a video. We had taken a number of small steps already to reduce the number, but exploring more yesterday we saw that we had a decent number of typo's in email addresses. In many cases we could track back to see that they came back to register again with the right address. How to fix? Regular expressions only go so far, so we implemented a process that connects to their mail server to verify the address at registration. Not foolproof, but should help a lot. And what happens if it fails, say due to a mail server being down? We warn them, give them the option to change addresses, but they can continue registering with that address if they wish. Definitely incremental, and in this case applied the same day we decided to pursue it.

I find most people don't have the patience for brainstorming. It's a combination of expecting immediate results (unlikely) and not wanting to voice any idea that isn't perceived as brilliant. It's a very analog process and it takes some time to learn that all that time (and money) for one good idea is a great day, and netting no good ideas is no reason to cry.