The Dude is feeling somewhat overwhelmed lately. And if you read the title of this post, you might divine that the Dude has been shoulder deep in EOL (End of Life) land that he is becoming maudlin. And you would be correct.
The setup is that about 7 years ago, the Dude joined a large company, working in their group that builds and delivers technical training where he has increased his responsibility to the point where he owns all the internally developed content, and some adjacent sources.
Not going to lie, this is a huge job. When the Dude started in summer of 2016, there were 9 product managers, each managing a part of our total portfolio, and multiple rounds of RIFs later at the beginning of last year we were down to three, with me owning the lion’s share, and in the latest reduction, the Dude is the Last Man Standing. It is a major block of work, overwhelming even.
But the Dude can handle it.
What this has brought to the front burner is that we have had virtually no discipline in managing the lifecycle of the product (content) that we produced.
That is a long way to get to the fact that since – well – Forever we have neglected to properly retire old, out of date, and frankly embarrassingly ancient content. There was no “next” step after releasing to production, and prior peers just moved on to the next thing.
Sure, if we do major updates, we would retire the older version as part of the workstream. But when a technology shifted, when a physical product we built training for is end of sale, end of life, and past the end of support window, we still have content sitting on our catalog, clogging the system, and leading to customer satisfaction issues when someone buys it.
And, when we had 9 people to manage the threads, we were no better. No, this is not caused by lack of resources, but mainly due to laziness.
And frankly, the Dude can sympathize. As he is clearing away the cruft, the dead wood, and the huge backlog of truly embarrassing offers in our catalog, he is hating the processes to EOL. And he has gained a lot of insight into how we got here. Some of this may resonate:
Wrangling the approvers
In most rational PLC’s there is an explicit step after mainstream production that handles when the demand declines, other products replace it, and when it becomes impossible to support. That is sunsetting the product, and it often has a lot more involved in it than just shutting down sales.
If you build hardware, you archive drawings, specs, procurement notes etc, in case someone decides they will pay a dump truck full of money for just one (or a few) more units.
In software, you often archive a development environment (virtual machines make this a LOT easier than it used to be) escrow the source code (because if you have very large customers like big banks, or the US military, they will insist upon it), and then wind down.
But for us, it is simpler. There isn’t a huge cost, we just mark it as unavailable, and sometimes move it to cold storage (save $$$ on our AWS bill)
In each of these cases, the activities are not the problem. It is getting the buy in from the various stakeholders.
And frankly, this suck.
There is always going to be at least one functional lead who will not want to retire the product. Usually, it is sales. They argue that it takes something away that they could sell, and that having more items in the bag is better, even if some of the items are badly outdated.
Regardless of who inserts the roadblock, it is a roadblock, and it is very tiring on a product manager to navigate and get to agreement.
Many product managers choose to not go down this path, and thus you get into a situation where you have training that highlights how to perform some task on say Windows XP, in a training you are selling for $3K in 2023.
Fucking madness. Yet, the Dude got MAJOR push back from the sales lead.
You can’t make this shit up.
If you can’t initiate the end-of-life protocols, then you have products that are in limbo.
You do a lot of limited run one-offs
This is a polite way to say that you do a lot of customization, and that will pollute your catalog with offers that are micro segmented to a specific customer or industry.
A customer wants something nonstandard, you build it, you deliver it, and in the event that the customer may want to buy more of it, you just “add it to the catalog” and it becomes orderable.
This is just fucking laziness, but it does simplify some parts of the process.
But the downside is huge. You have this small team doing bespoke work for limited use, turning out huge numbers of catalog items, and virtually no PLC besides XYZ customer wants something, they gave us a bunch of money to do it, and now that we have it, let’s see if anyone else is interested in buying it.
And this is the hell that the Dude finds himself in. There are literally hundreds of catalog offerings that are active “just in case” someone wants one like it.
He is literally cranking out 3-4 EOL packages a week to shut these down. And yes, Sales is squealing like a stuck pig about it decimating their catalog.
There are some bright rays of hope. When the Dude socializes the next tranche of EOL paperwork, some of the stakeholders cheer him on.
But at the end of the day, the Dude is taking Ol’ Yeller behind the shed and euthanizing him most days. It has grown to be about 30% of the Dude’s working time, and there is still a mountain of cringe-worthy candidates to “retire”.