Thursday, December 02, 2010

Why Products Really Suck, Part 1

Image from:
I recently read this rather glib and uninformative article about "Why Products Suck". It got me thinking. I've made and shipped a number of products. Some won awards. Some were terrible.

I've definitely learned a few things, and I can definitely provide more insight than the above-mentioned article's key points, which were:
1. It only takes one person to make your product suck.
2. Nobody ever got fired for sucking.
3. It’s easier to suck more than suck less.
4. There are more ways to suck than to not suck.
5. Customers demand sucky products.
So I'll be doing a series of articles on making good products. First, let's look at what's wrong with this article.

First point. The author claims "It only takes one person to make your product suck." There are a number of interesting and correct directions he could take this statement. But his actual point is that you need consensus among your "team" that your primary goal should be "not sucking".

It is true that it only takes one person to wreck a product. There's a New Yorker cartoon I used to have taped to my wall that showed a guy at a desk writing on something his subordinate handed him, saying "I just need to make one change that will ruin all the work you've done."

Frequently, that person is not who you expect them to be, and they will be the subject of a future article.

I will also discuss the pros and cons of team consensus.

Second point. "Nobody ever got fired for sucking". This is sadly true, however the point as presented in the article is rather weak. The author is basically saying "you should fire bad people" without clarifying the differences between:
  • appropriate and inappropriate risk
  • initiative and running wild
  • genuine accidents/mistakes and negligence
Good teams and good managers not only get enough rope, they sometimes get tangled up in it. Not every risk pays out, and some risks lose big. That's why they're risky. Start-ups take bigger risks than established companies, and new products take bigger risks than established ones. They have to in order to survive.

I'll discuss risk and failure as well.

Third point. "It's easier to suck more than suck less". His somewhat dippy phrasing covers the best point he makes in the article: Serve the right customer. Don't be so desperate for customers you build features that aren't core. You'll attract the wrong customers and then be beholden to them with even greater costs for getting back on track.

There will be a nice long article about discipline in product design.

Fourth point. "There are more ways to suck than not to suck." His point here is "you need to be careful with your choices and define your product clearly before you start development". This is true, but many of the examples he cites are not relevant to that point at all. There are tactical mistakes one can make. Choosing the wrong user interface widget isn't one. Good UI designers don't really have choices with too many questions if the product is well defined - in those cases, the best (or at least better) tactical choices are fairly obvious.

Strategic choices are much more difficult and much more important, and frequently glossed over in both theory and practice. I firmly believe that making great products is not particularly difficult if the environment is good and the team is competent. I'll cover that as well.

Fifth point. "Customers demand sucky products." This is a rather provocative statement, given the "customer is always right" attitude that dominates. The author's main point is , but he follows it up with a poor, or at least insufficiently specific suggestion:
Trust your instincts and the tiny set of users who use you, and resist advice from the billions of people who don’t.
Good product designers do have instincts, but those instincts are informed by experience, best practices, and lots of research.

Your current users are frequently the last people you want to be talking to about how to evolve your product. In the early stages of a product, your users are likely to be unrepresentative of the larger user demographic, unavailable in sufficient numbers to float your whole business model on, and highly demanding.

For more mature products, your users are again likely to be unrepresentative of the next tranche of users you hope to attract, and will be used to how your product works. They are likely to demand minor iterations or evolutions of features.

The short answer is "talk to people who are like the users you want to have", and the long answer will be coming in a new post soon.

Up Next: Who's wrecking your product? (Answer: Steve Jobs. Seriously).

No comments: