To the best of my knowledge, it was investor and entrepreneur Chris Dixon who popularized the term Normals (Caterina Fake suggested “muggles”) to signify your everyday person. His theory (paraphrased) is that your business will never be a huge success unless your userbase includes a vast majority of Normals. Early adopters are good for initial traction and launch buzz, but until you attract Normals, you'll never get past that first reaction. Early adopters are also fickle and will quickly jump ship when something "hotter" comes along.

What exactly is a Normal?
A Normal is maybe not an everyday person in every way, but has limited Internet knowledge. They certainly don't read TechCrunch, they haven't heard of RSS feeds, they probably don't have a smart phone or at least don't have many apps installed, and although they surf the Web a lot, they have little clue what a web browser really is. Another telltale sign is that instead of going directly to web pages, they use the search bar. You know these people if you're reading this blog.
What is an early adopter or techie?
If you read this blog, you're probably a techie, with the exception of my mom and dad (they're Normals). I'd define techies and early adopters as people with smartphones, readers of tech blogs, anyone on Hacker News, people in the startup scene, users of Foursquare, Gowalla, HotPotato, Plancast, and many other "hot" startups that haven't yet reached the Normals.
Why is it more important to reach Normals?
The simplest reason is that Normals make up far more than 99% of Internet users. If you fail to reach the masses then you'll simply fail. You can be the hottest startup on the block with 100,000 active early adopters, but I'd trade every one of those users for Normals in all cases.
Another reason, often overlooked, is that Normals stick around a lot longer and are far more loyal than early adopters. When they've settled on a service choice — and it's usually one of the first they come upon — they're less likely to shop around as long as everything is going okay. This isn't the case for early adopters, who by definition are always looking for the best big thing.
Besides positioning yourself to attract a larger part of the market by going after Normals, who are loyal in themselves, their friends are Normals too, and that propagates the cycle of adoption. While early adopters are great at getting the word out to their friends, those friends are other early adopters in most cases, so the circle remains closed.
Focus on Normals!
At Carbonmade, we've focused all of our efforts on attracting everyday people who may not be the most adept Internet users. That's why we don't support HTML editing or too many advanced features — our product is designed for easy on-boarding — easy in that we boil all of our features down to simple-to-use components.
But beyond keeping your product easy to use, your startup won't be capable of attracting Normals if it's not in a space that you can imagine they or their circle would be interested in. A great litmus test for this is asking a Normal if they'd use your product and getting a "no" answer followed by "but my niece, nephew, friend, etc." would. You don't have to be able to attract every soul on earth, but if everyone you talk to knows at least one person who would use your service, then you're on your way to reaching Normals.
This has proven true with Carbonmade. Everyone I've talked to about Carbonmade knows at least one person who would use it themselves: everyone has creative friends or family members eager to use an online portfolio. A good exercise is to put your startup to this test and see how you fare.
Now there are certainly plenty of cases of reaching early adopters first before the Normals follow in their path, e.g., Twitter, but I'd argue that this is a more difficult and time-consuming approach. It's also more likely to require outside capital. When you have to force your way into the Normal scene, this often requires marketing dollars that lean startups can't afford to spend.

Jonathan Wegener about 5 months ago
YES YES YES! Love it!
Ever since you got asked 'Why would anyone pay for this product? Why wouldn't they setup something themselves?' at the Dogpatch Labs event last week I've been thinking about this exact point. The guy asking the question (and much of the audience) didn't get it -- Carbonmade isn't about introducing groundbreaking technology to the world. It's about creating a solution for the normals -- the 99% of people who aren't going to be able (or want) to figure out how to setup a custom wordpress install or setup simpleviewer.
I started using carbonmade a year ago to help a 'normal' manage their portfolio. Since then I've been most impressed by the lack of features that Carbonmade has! It's a great testament to your strong product vision and market focus. And it's clearly working.
Nice to see these thoughts clearly laid out. I would love to see your last paragraph expanded into another blog post.
-Jonathan
P.S. Should we techies ever forget that normals make up the majority of internet users, just consider the RWW/Facebook incident: http://www.readwriteweb.com/archives/how_google_failed_internet_meme.php