First, keep it simple.

Start small. Talk to people you know. Talk to employees, sales people, colleagues in other companies, your family-whoever is the best representative of your customer. Find out what's up in their world; ask about their problems, their needs, their goals. Drill in on their product experience and their issues with it.

Find a group of individual customers you can talk to. Get more out of your existing sales efforts by making sure you use them as a chance to learn. If you aren't already, go on a few sales calls or man the phones. It's often best to do this with the most demanding customers or prospects. As Bill Gates once said, "Your most unhappy customers are your greatest source of learning." They will tell you the stuff you need to know. Ask tough questions; be prepared for tough answers. If they seem to be telling you only the good stuff that's nice to hear, don't be satisfied; ask again.

One company we work with targeted teens and found out that talking to them was a sure eye-opener. Adolescents aren't afraid to tell you what they think is cool and what isn't. But how about insurance agents? These guys were used to making sure they said the right thing-they didn't want to let down their listener and chance losing a sale. Trained to be polite and agreeable, they just kept smiling. The company folks really had to push to get them to open up about the problems they had with the company's products.

Second, keep it digestible.

As we've seen, your target customer base can be widely diverse. It really helps if you can narrow or break this discovery process into digestible chunks. There are many, many different ways to do this segmentation: competitive, demographic, psychographic, etc. But two ways seem to be most useful: segmenting by the customer's roles in the purchase process, and by their attitudes toward your category's products or technologies.

The Avogadro and American Express examples illustrate the first steps of segmentation by role in the purchase process. Taking this further, you want to know who is most interested, who holds the pocket book, whose approval is required, and so on. You're looking for the fulcrum or lever-the person whose approval is most important, whose voice on your topic carries the most weight. At Microsoft we broke this typically into four segments or targets -

• The general business user, or GBU, who, although belonging to the majority of users, tended to be a follower and had limited influence or none at all;
• The non-technical business decision maker, or BDM, who owned the budget, and whose approval was important;
• The technical decision maker, or TDM, who generally did not initiate interest in a new product but who often held a veto based on his/her judgment of whether or not it could be implemented;
• Finally, the influential end user, or IEU-the enthusiast or resident expert to whom the GBUs and BDMs turned and whose enthusiasm could often overcome or circumvent the objections of the TDM. When we were at Microsoft, we generally found that it was the influential end users who were the fulcrum.

In terms of segmentation by attitudes, one of the most useful approaches involves understanding how quickly and regularly different people or groups adopt new products-where they fit on the spectrum between bleeding edge and laggards. At Microsoft we called the segments regulars, seekers, doubters, and sleepers.

• Regulars were those who always bought.
• Seekers were those who needed sufficient information first to make a considered choice.
• Doubters were those who needed to see others adopt first before they even considered.
• And sleepers were those who only came along last, kicking and screaming all the way.

Each of these types of segmentation determines how you refine and even narrow the customer gap you are trying to fill, and when you should attempt to fill it. Those who spend more money become a higher priority. Those who are most receptive offer the easiest place to start.

Third, keep your customer analysis direct.

Doing market research and analysis is fine. But to really absorb the conclusion, you have to get your hands dirty. It's not just business, it has to get personal.

Of course, you need to meet customers up close and look them in the eye. But you can and should go beyond that. You should be one of them. You should make sure you are your own customer. It will keep you honest and it will teach you things you might not otherwise have found out.

For years we tried to understand and influence the purchasing of Office, Word, and Excel by small-business owners. We tried using the same approach with them as we did with large corporations. We tried including all kinds of fancy new products based on focus group and primary research. None of it worked. It wasn't until we really experienced for ourselves what the small business customers were experiencing that we realized where we were going wrong and changed our strategy.

While we were running all of our focus groups and surveys, a terrific summer intern, Tony Liano, who didn't know any better, went out and had business cards printed, and got a few of us to pretend to be small business people. We did our own "Secret Shopper" research. We'd go into a store, tell the salesperson we were setting ourselves up in business, and ask for his advice. And we'd make it clear that we were very budget-and time-constrained.

The salesperson would start by selling us a bunch of other things like office furniture, phones, faxes, etc. And even when he got around to selling computer technology, it would be all about the newest computer. When it came to software, he'd usually recommend a bookkeeping program and something like a simple contact manager. Office wouldn't even be mentioned. And when we asked, the sales rep would usually try to sell us something from a Microsoft competitor. Why? Because he made a better margin on them.

The same kinds of issues and priorities were confirmed in our direct discussions with other small business people-who now would talk to us as colleagues. They didn't think of themselves as running small businesses. They saw themselves foremost as contractors, or florists, or accountants.

What an eye opener. All along we had been doing "small business" campaigns and "small business products," when our target actually did not think of themselves this way. It took some hard turning but our response was to adapt to their characteristics. We made sure to target vertical industry-specific partners who provided software solutions to small businesses. And we targeted our packaging and presence to complement new PCs and grant incentives to the channels where small businesses bought their overall computing supplies.

How do your own customers think of themselves? Do they refer to themselves with the same kind of category labels you've been pasting on them?