Henry Ford once famously said that his customers could have any color car they liked as long as it was black. When playing with the various emerging "mass customized" versions of already popular products, I am often reminded of his statement. Thus far I've tried custom Nike running shoes, a custom bag from Timbuk2, and just this weekend, Lego's "Factory" product. In all three cases, I've been disappointed.

The reason is simple: for a mass customized/personalized product (by which I mean on that can be manufactured according to the enduser's design specifications) to justify its premium, it has to be measurably better than its off-the-shelf cousin. And note that when I say premium I don't just mean price (in fact one of the mistakes I think companies experimenting in this space are making is not charging a price premium), but also the cost in time as anything made-to-order is bound to take longer.

The case of Lego Factory is a good example of how a company can almost get it right. The idea is simple: after downloading a Legoized CAD program, users can design their own Lego set by picking from a catalog of bricks. After this design step Lego will custom build your un-assembled set and ship it directly to you from overseas in about 2 weeks. You can even customize the cover of the box with an image of your model set against a number of canned backgrounds. Overall the price is fairly decent (I paid $25 shipped for a set with 110 pieces) and the basic experience delivers on the promise of "making your own Lego set." However far too many edges are too rough for this to ever become a mass market experience.

The configurator: In the photo creative world this is what we call the online service/software that lets you assemble the photo book/card/product. In Lego's case, the downloadable CAD Lego Factory program is B-grade software at best, and that is for a truly committed user. There are plenty of bugs that riddle the experience, the catalog is tough to navigate, and the physics of snapping the virtual pieces together doesn't work consistently. But most importantly, the company forgets to provide enough partial starting points towards building a compelling model and offers little in the way of guided help to get people to feel creative. True you can start from a car or a plane, but these are finished products at different scales that don't even manage to combine well with each other. Partially finished consistently scaled models or better yet, the ability to "draw" out shapes and have the tool fill in with bricks would go a long way toward fixing most of the heinous issues.

The out-of-the-box experience: It turns out that after messaging that you could "customize the box" what actually comes with the set is a cheap generic box with an inkjet printed version of your model stuck behind a cellophane cover. I doubt that Lego would actually ship a product to a store with such a poor box. Similarly instead of the choke-proof sealed plastic bags, the bricks come jumbled together in what looks like the ziploc bag grandma used to give you cookies to take home with.

And worst of all, there is no manual! Instead there is a lame letter stating that the Factory kits don't come with manuals suggesting instead that you use the software to step through an arbitrary assembly process while the CAD program reverse-explodes the model you've built one step at a time in what seems like a truly random process.

Until my experience with Factory, I had no idea how much human editing went into making Lego build instructions (I had naively thought that most of it was automated by computer). And as it turns out, these highly edited guides are an integral part of the experience, especially if you are a kid who has been trained in the 3 rules of Lego (1. Never lose pieces, 2. The manual is always right, 3. Don't ever force it).

As with Nike and Timbuk2, I think Lego deserves praise for the Factory experiment. But unless they remember that mass customized goods need to be that much better and not actually worse than their factory-churned relatives, it's going to remain an interesting side project.

If I were in charge of that team, I'd work hard on the software— both during the design phase, and in the creation of a compelling and usable guide (two software problems which appear to me immanently more solvable than some of the others the team has already solved in the end-to-end experience)— in order to move this experience into the mainstream.