Thought Experiment: The ATM Machine.. An 'Epic Story' Fail

The story of how the Automated Teller Machine (ATM) evolved over the last 70 years is actually pretty interesting. It's also a great case study on agile and how keeping an open mind is critical when writing stories/epics.

You may not be aware, but an interesting part of ATM history occurred in 1983 when IBM released the 4730 Personal Banking Machine that, among other features, could dispense an exact amount of change. For example you could request to get $37.05 if you wanted.

Obviously, this approach ended up not being viable, but IBM did manufacture the product and ultimately cancelled it a short time later. For a moment, let's turn back the clock 35 years and pretend agile methodologies existed, and pretend you are in the room with the 5-7 people as they are writing the epics that described this product. I can almost see it... someone starts off by saying,

"We need a machine that can completely replace a bank branch"

which ends up turning into the epic story

As a bank member I need the ability to withdraw any amount of money, up to my daily limit, so that I can get the amount of money I need

Maybe, just maybe, someone pushes back and says something like

"Wait... do they really need to get the exact amount? What if we limited it to just paper currency?"

and someone in the room replies

"No it needs to be exact, that is what the users want"

I'm sure as some of you are reading this you can relate to this dialog. This is a critical decision that impacts the hardware design, the software needed to run the machine, and the operations to keep the machine running. The lynch pin of the decision came down to someone saying "that is what the users want". Frequently, people make blanket statements without supporting data because they assume all users behave like they do, which is why data driven story/epic validation is so important. For example, if the group had a set of data in front of them which showed that the percent of people who go to the bank to withdraw an exact amount occurs less less than .1% of the time, then the above conversation most likely would have gone differently. The group would have realized that in order to achieve a Minimum Viable Product (MVP) the epic should have changed to something like...

As a bank member I need the ability to withdraw money so that I can pay for goods

With the epic written this way tens of thousands of work hours would have been saved and a more appropriate product would have been produced. But all of that was thrown away because of 5-7 words decided by 5-7 people. This is why truly understanding the business need and continuing to challenge the methods and approaches to meet that business need is so critical in agile.

I find that this ATM story resonates with folks because it's tangible and it helps those who have a hard time understanding why they can't always just "get what they want". It shows that keeping an open mind and realizing that your initial vision may need to change can be the most important quality you bring to the team.