I had a conversation last week with one of our long-time clients in the biotech industry. He had just rejoined a team we’ve been working with for years, and what he told me illustrates just how much AI tools depend upon solid foundations to be as useful as we expect.
This particular team has experienced quite a few changes in recent years. Mergers, acquisitions, reorganizations, product consolidations, staff turnover, and more have been part of the mix. Our old friend reached out because he was hoping our archives still had source files for a set of manuals in need of an update.
Thankfully, we did have most of the source files, but there were some we hadn’t worked on. Some other vendor (now unknown) might have them in their archives, but given the deadlines involved, there isn’t time to find them. So we’re recreating them from the PDFs they still have.
Going forward, we’ve offered to help this team establish and maintain a reliable content management and asset tracking system.
Companies need to know where their documentation source files live and who is responsible for their ongoing maintenance. And that system needs to function in our tumultuous world.
While I’m always surprised by these requests, they’re more common than you might expect. As busy as everyone who works in tech is these days, and how much change people deal with, perhaps it shouldn’t be surprising that documentation source files sometimes get lost.
However, a good system helps keep track of who owns what, all of the digital assets (text, templates, images, source files, etc.), an audit trail of changes and updates, who reviewed and approved, and who is responsible now. Of course, that system needs to be used, and available to the team of people ultimately responsible for the documentation.
Getting all of this set up can be done with a file system, a database, a repo, or any of a number of capable digital asset management systems. We’ve helped to create and maintain these kinds of systems for many clients over many decades. We’ve seen them work well, kinda not work, and not even exist (like the latest example).
Now, not having a system imposes an even higher cost: Without a well-structured repository, there’s little chance AI-powered systems can ever reliably access the knowledge they contain. That means the AI systems your employees and partners and customers use won’t benefit from your unique source of truth, best practices, experience, and perspective. And that’s a significant lost opportunity in our competitive modern world.
Worse, the disconnected AI tools those people use can still find information somewhere, or invent it. And then they’ll use it, right or wrong, to attempt to do whatever they are asked. The results they produce (which might be advice, code, instructions, or who knows what) might sound right, but might not be right. Regardless, the odds of errors, re-work, wasted time, and frustration are much higher.
So, do you know where your docs are? Do all of the key people in your organization know who’s responsible for what and the process for getting updates done? Do you have a system in place that collects and organizes your crucial documentation? If so, does that system make it easy for an AI-powered tool to find and use what it needs to be successful?
If you need help answering those questions, or putting the right kinds of systems in place, let’s talk.
Paul Gustafson is CEO at Expert Support. A long-time Silicon Valley veteran, Paul has helped hundreds of clients optimize their technical communication strategies.
Expert Support helps technology companies explain complex technical knowledge for the people and AI systems that depend on it.