It’s a conundrum.  To be a successful company, you need customers.  To get customers, you must build something great.  Most people say, “Go to market quick, then learn and iterate.”  I think people are missing a key factor about what it means to have customers.  I saw this first hand at several different companies.

Many people in the organization will think you can innovate and build just as quickly before customers as after.  One executive I knew said, “We lost our mojo!  Why can’t we build like we used to?”  What he didn’t understand was that we introduced customers into the mix and that changed everything.  One big cause is Bugs.  When you are building, you introduce lots of unseen bugs that only get surfaced after you have customers doing things live.  All of the sudden, the quality of the product seems to drop, when in actuality, the bugs were always there.  Thus, the workload of the team is shifted dramatically to supporting the initial offering.  Illustration:

The next problem is “change management”.  Before you have customers, you can radically change your product offering architecture.  You can change the names of the primary objects and move the IA all over the place.  However, the minute you have customers, you can’t make those shifts anymore.  I once “fixed” an email editor and it was like I unlocked the gates of hell itself.  Complaints that were so over-the-top, you would have thought I poisoned their milk.  New customers who hadn’t experienced the old way were perfectly happy.  Illustration:

All of this has a significant impact on the feature plans.  Product managers often make roadmaps that look like a steady stream of features, when in reality, they will have end up with a fraction. Illustration:

You can’t keep building features once you have customers requiring support and when they will resist change.  I wish people would recognize this so they can be smarter about going to market.  It’s important to ship code, but ship too soon and you will can fail that way as well.  Of course, this is a judgement call, but we warned:  The first customer is very expensive.