The great thing about being a T-Mobile customer is that you can get a new phone without ever having to wait for an upgrade. And their rates are low. And I think the rates are so low because only three people in the country are able to make a call at the same time.

So, I’m driving to pick my daughter up at her friend’s house, a friend who lives nine miles away. I fire up the GPS on my aforementioned T-Mobile; estimated drive time of seventeen minutes.

The route is mostly back roads, and as soon as I reach the backest of the back roads I am informed, “GPS signal lost.” After having my way with the recording, I continue to drive. And drive. And you know the rest. I head to places with higher elevations, find a signal, and am directed to roads that even I know have nothing to do with where I am headed. My route is again transferred to another, and so forth and so on.

I am talking to myself, and yelling at drivers who look like they know where they are going. Squirrels on the side of the road point at me, and double over in laughter—I hope they choke on their acorns. After an hour and two minutes I arrive at the house, upon which my phone chirps, “You have arrived.” I started screaming epitaphs in front of the squirrels.

An hour and two minutes is four minutes longer than I spent the other day trying to schedule an appointment with a large east-coast health system. The person with whom I spoke the longest seemed to have a pulse equal to that of a hibernating bear, and had he been a household pet he would have been put down purely for aesthetic reasons.

I find it helpful to trace everything back to a seminal point like just prior to when the random swirls of gases in the chartless universe got together and formed the earth; or not, depending on which side of the Darwin bed you sleep.

From the side on which I sleep, the seminal point for patient access can be traced to the phone. If the people manning your health system’s phones cannot meet the basic needs and solve the problems of its callers, very little else matters because the callers will call somebody else.

Hammers, like phones, are very evolved tools. They haven’t changed for years. And why would they change? Nails haven’t changed. Nails have been the same forever. Therefore a hammer’s necessary features were worked out long ago. A heavy metal head, and a handle. All you need, and nothing you don’t. Phones are also very evolved tools. The primary difference between a hammer and a phone is that if you do not have able people on both ends of the call, the phone is useless.

Transferred calls, dropped calls, callers placed on hold, being told my call would be recorded for quality purposes—an illusory promise of something unseen. When I heard the recorded voice inform me of that tidy bit of news I yelled back, “Wouldn’t I actually have to speak to someone to make the recording worthwhile?” The entire scheduling process—I am overstating the concept as there was no process, just a series of random interactions—reminded me of a series of cattle pens at a slaughtering house, and the only thing that saved me was my refusal to be herded. By the time the process had concluded I had developed nictating membranes over both eyes and recessed to the lizard part of my brain.

Maybe the process was designed to wear people out, to get them not to call back. Sort of like the customer experience models used by payers.

Just to be clear, there are no catastrophic patient access failures. The failures people suffer through are made up of a series of small things that become big things, much the same way that individual snowflakes become an avalanche.

Health systems should bear in mind that even though the people who work there may know the number to call to get an appointment, nobody else does. In most cities only six people have a phone book or the Yellow Pages. Chances are good that even if they could find the listing for the health system they would not know which of the dozens of numbers to call.

Here is how real people—patient people and consumer people—figure out how to call a given health system. They go online; they see a number, cross their fingers, and dial. They do this process again and again until they either give up, or until they declare themselves cured and no longer need to purchase healthcare.

The chief marketing officer of one of my former clients, a large academic medical center in Philadelphia told me her health system had more than 1,200 URLs. That probably gives the average consumer more than 1,000 different numbers to call to have their needs met.

The most functional feature of any health system’s website, if viewed from the perspective of someone wishing to schedule an appointment would be a big, flashy display of the single phone number they need.

Your website’s home page has become the digital phone book for everyone wishing to interact with your health system. I estimate that more than fifty percent of the people who visit your system’s website do so to figure out which number they should call to have their need met. My experience suggests that almost one hundred percent of the people who are trying to call your health system have to call more than one number to speak with someone who can help them.

If one thousand people go through this exercise each day, the health system receives an extra thousand calls simply due to the fact that everyone has to call more than once. If each call’s fully weighted cost is thirty dollars that means thirty thousand dollars a day goes up in smoke. That means the health system spends more than ten million dollars each year because people must call again and again.