← The Missing A’s
Missing A Awareness

Six months to do a day’s work

Six figures to rescue data from a Flash application everyone had forgotten, because the people who understood it had moved on.

Randall Wheeler ·

Everyone had known Flash was dying since Steve Jobs published “Thoughts on Flash” in 2010. Nine years later, I was being paid to get the data out of a Flash application that nobody at the company knew how to build any more.

It was 2019, at a large automaker. The application was written in ActionScript, and as far as anyone could tell it hadn’t been changed in about a decade. Nobody in the company had any experience with ActionScript, and nobody was familiar with the code. They didn’t know how to build it, or whether it would build at all. What they did know was that there was data inside it they needed.

So they hired my company, and my company sent me.

Screen by screen

I couldn’t rebuild the application, so I wrote another one to scrape the data out of its interface, screen by screen. It took about six months. Between evaluating the application and extracting the data, I’d guess the company paid well over $100,000.

Someone who had actually built that system, who had committed code to it, could probably have done the same job in a day.

The cost came from everyone who understood the application having moved on, without anyone noticing when that knowledge started to matter.

Who is your map?

Every company I’ve worked with is trying to get new things done. There are strategic objectives, and everyone wants to work on them, because they carry the promise of growth. And there is tech debt and maintenance, competing for the same people and the same budget. Planning is the work of balancing the two.

To balance them, you need to know what you have: which systems support which parts of the business, what they run on, who looks after them, and where the risk is. That’s what a capability map is for.

A lot of companies rely on middle management to be the map, long after the company has outgrown their ability to give an accurate picture of the enterprise.

It makes sense. Middle managers know the landscape better than anyone. But they’re also some of the busiest people in the company. Their expertise runs deep in their own silo and thin outside it. And what they know best is whatever has been urgent lately.

A system that quietly works never becomes urgent. It doesn’t come up in planning, because nobody in the room is thinking about it. It doesn’t get a migration, because migrations go to the things that are loud. It just keeps running, year after year, while the people who built it move to other teams, other companies, retirement.

Then one day it becomes urgent. By then, it’s a six-month job.

What being aware of it would have changed

Old systems are often fine. The Flash application became an emergency because nobody was looking at it.

If someone had been able to see, at any point in that decade, that this application:

  • supported something the business still needed,
  • ran on a platform with a public expiry date,
  • and had nobody left who knew it,

then moving the data out would have been a small, boring project, done while the people who understood it were still down the hall. Instead it became an emergency, and emergencies are priced accordingly.

None of that information was secret. It just wasn’t written down anywhere anyone would look. It lived in people’s heads, and people move on.

The missing A: Awareness

Tech debt waits until the knowledge needed to deal with it cheaply has gone, and then it sends the bill.

The organizations that avoid that bill can see what they have: what each system supports, what it runs on, who owns it, and when it last changed. They see it all the time, not just through the few people who happen to remember.

That’s the awareness a lot of organizations are missing. It’s what we’re building cpble to give them.