This episode explores common time sinks in technical writing and practical solutions to streamline documentation workflows, with a focus on automation and AI tools.
Transcript
We’ve had a month off, and the Cherryleaf podcast is back up and running.
So welcome. And in this episode, we’re going to be looking at time sinks in technical writing and solutions. My name is Ellis Pratt. I am one of the directors at Cherryleaf, which is a technical writing and services company.
I want you to think about your last writing project for a second. Was it a smooth, streamlined process, or did it feel more like you were drowning? Drowning in scattered feedback, endless drafts, or just version control nightmares?
If that chaos sounds a little too familiar, then this episode might be of interest to you.
Now this idea of talking about this has been inspired or prompted by a couple of things that are coming up. So before we dive into the podcast itself, let me do a little advert.
One of the things that inspired this is I will be talking at the TCUK conference on the topic of
Automate All the Things, the Present and Future of Automation in Technical Writing. And that conference is going to be held at the Radisson Blu Hotel at East Midlands Airport which is in Nottingham in the UK. And the conference is going to be run on the twenty fourth and twenty fifth of November this year.
And the other activity that we’ve been involved with is that we have launched a new course, and the course is managing and mastering documentation projects with AI.
We’ve got a date for the pilot course, which is the third and fourth of November.
If you’re interested in that, then email info at cherryleaf dot com. And after the pilot, we will be scheduling more courses. And that is an instructor led course delivered over Teams on two consecutive mornings, but more about that later.
Every job has its time sinks. Those activities that take longer than they should. Those moments when you’re sat waiting for input from other people or feedback once you’ve completed your task
And technical writing, technical communication, is no exception. There are those situations where you wish you could just push a button, and it happens, so you can concentrate on the more valuable or more interesting parts of your job.
And that’s true at Cherryleaf as it is anywhere else. For this episode what I thought I would do is ask other technical writers, other technical authors what are the time sinks that they have to deal with to see if there are some common issues that everybody shares.
I posted a query to the write the docs forum, which is an a forum aimed at people primarily doing developer orientated docs, also onto the technical writing forum that’s on Reddit, and also onto LinkedIn to get a wide perspective.
What I’ve done with those responses is I have put them through some text to audio software so that we can hear what they’re saying with a bit of variety of voices rather than just me reading them out from the list. So let me play some of the responses that we got:
I think chasing reviewers is probably one of the biggest, but I’d also add changing priorities and expectations of instant responses to quick UI message wording, checking on language use, etcetera, questions so you are constantly switching focus.
As a manager, attending many unproductive meetings where my contribution is minimal and my time could be better spent elsewhere, or dealing with nonsense demands from senior managers that add no value, all time wasters*
How about the risk of documenting too early? All looks well, so you get ahead of the game for once, and then there’s changes during testing, bug fixing that require a rewrite. Conversions to and from markdown.
If your platform needs a MD source but you draft and collaborate in a g doc, you’re bound to waste time moving content back and forth even with scripts that get you part of the way there.
My biggest time sink is futzing with tables in InDesign.
Table styles can only do so much. And of course my favorite, converting docs from my colleague who still works in Word. Sometimes it’s hard for my reviewers to read the actual documents for review, and it takes me a bit of time to get them on the preview server. My team is working on that.
Follow ups is a big one.
Tracking my requests is sometimes also not great.
Time zone issues for scheduling meetings, chasing reviewers and SMEs, tracking issues through a spreadsheet.
The largest time sink for developers and writers is trying to produce products with poor or no requirements. Features that don’t work as expected. Spent two hours troubleshooting an issue yesterday.
I could have finished what needed to be done in that time. Why didn’t “thing” make it into the docs? Because you did not check the “add to docs” box in Jira. Rinse and repeat.
Not having set communication standards or soft deadline type approach to reviews is such a time sink in my experience*
We got many, many more. We might post them on to our blog. Otherwise, you can see them on LinkedIn and Reddit and Write the Docs, see the comments there. We’re not clearly the first people to talk about inefficiencies in technical writing and how they can be improved.
And in fact, there’s a buzzword for looking at this and that is content operations. The concept of content operations has emerged as a way of improving the way content is produced in technical writing, marketing content, other forms of content.
So what is content operations?
Content operations looks at the people, the processes and the technology involved in publishing content. And it’s typically a strategic approach that aims, amongst other things, to get people in different departments to work collaboratively. It can also be about getting different systems to work collaboratively as it were.
If you’ve ever been involved in a change management project, you’ll probably know that implementing and managing change at a strategic cross departmental level can be challenging.
And this was highlighted by a sticker that I saw on a laptop of Laura Hiles, who until recently has been Head of Content at the Government Digital Service. And the laptop sticker says, why have difficult conversations when you can talk about tools instead? So it’s a little bit ironic.
And if we look at that list of problems, those time sinks that people talked about, the solution would often be a need for other people to change the way that they work. For example, reviews, not being told of changes.
Some of the issues also relate to changing technology, like getting content from Word into another platform.
We didn’t hear many comments about the inefficiencies of people themselves
And there’s a temptation with something like that to focus on the inefficiencies of others
Time sinks may also be things that you do but you’re not aware of. And they may be in your control to change without having to involve others.
Let’s look at the solutions to fixing these time sinks.
Can you take the person out of the process so that they’re not a bottleneck?
Can you make it easier for them to do what you want them to do? So there’s less friction for them. It’s an easier task for them.
Can you spread the workload, delegate the work to somebody who has the capacity to do it?
Can you automate the manual tasks that you do? For example, moving data from one system to another, or improving content.
In the past, a lot of the focus has been on tools because that’s been a way to automate processes and procedures.?
But what do we mean by automation?
You can think of automation like a tireless intern. You give it, the automation system, a rigid set of rules, such as save every attachment from this sender to a specified folder and it does it. And it does it consistently and regularly every single time.
So automation is good if it’s possible for you to apply a rigid set of rules, if things are predictable
The problem has been that that’s not always the case. There are often many exceptions to the rules. There are edge cases, those situations often rely on somebody with expertise to know what to do
However, with AI, things are starting to change because we now have things called intelligent agents
and they are more like a junior project manager than an intern.
You give it a goal, such as deal with incoming support emails, and it thinks, it figures out the best way to achieve that. It adapts. It makes decisions.
So let me explain. While every intelligent agent uses automation, not all automations are intelligent agents.
And intelligent agents as a buzzword, can also mean no code tools. And this is where you can prompt an AI agent to create a software application for you.
And that application itself can be an automation with no intelligence built into it, no thinking built into it, or it can have its own intelligence be an agent in its own right.
In most cases, they’re automations. For example, you can get an intelligent agent to build an app that replaces the data in a screenshot. In the fields there might be personal identifiable information, get it to replace that information with alternative information. So instead of it saying, Mrs Brown at 6 Acacia Avenue, it says, Mr Jones at 123 High Street which can save you time having to blank out text or go in and manually make those changes yourself.
And this is where there is a new opportunity for technical writers to reduce the time sinks. You have the capability now to streamline or automate activities that only involve you or your colleagues in a writing department, things that don’t involve reviewers or other departments.
You could also use an intelligent agent to create an app that makes it easier for people in other departments to do what you want them to do. Reduce some or all of the friction that’s been an issue for them.
And what we found is that when people know that’s possible, it’s surprising how many solutions they can come up with. For our new course, managing and mastering documentation projects with AI. We’ve developed nearly forty applications and we could probably easily create another ten more.
When it comes to automation, you don’t have to use AI. In the conference presentation at TCUK 25, we’ll be looking at a wide range of ways to automate, and many of those don’t involve AI at all
So if you want to do this, where do you start?
Well, obviously, go to our conference presentation, take our course. In addition to that, let me give you some suggestions Number one, focus on the stuff that would create greatest value, the high impact stuff. Think about your most time consuming, repetitive tasks. Things like generating first drafts, checking content for style consistency or putting together change logs or release notes.
These are typically good candidates because they solve significant pain points and can give you a quick return on your investment, your time.
An AI system, an automation might be responsible for doing the initial first task, like creating a draft. But a human technical writer is always accountable. They have the final say. They own the quality.
So this isn’t replacing a technical writer, but augmenting them, supplementing them. The AI or the automation provides the raw material, but the human expert is always in charge of strategy and accuracy.
So it’s not just about feeling more efficient, it’s being able to prove it.
And that leads on to recommendation number two, track the key metrics so you can measure the real world impact.
Metrics means you can do more than just tell people. You can show people what the impact is of what you’re doing.
Third is do things in stages. Have a phased roadmap.
And then what that means is you can lay the foundations for better improvements, get buy in from other people who maybe want to get involved, and that can help with those difficult conversations that Laura Hiles talked about. It can make it less of a challenge for others to change the way that they do things.
Fourth tip is to remember automation, AI, they’re just tools. As powerful as these tools are, that’s what they are, a successful rollout depends entirely on human oversight, on strategic direction, and having a strong framework in place.
If you automate anything, there are risks, But they tend to be manageable. One risk is you have a system that generates inaccurate content, which is why having human review as a fundamental mandatory task.
It needs to be non-negotiable
You’ll still need to get people involved from the beginning to get buy in.
And to avoid over-reliance on AI systems, you have to make sure that humans always always own the final content strategy.
So our philosophy is not about replacing skills, but refocusing them. Getting the mundane tasks that technical authors don’t necessarily like, delegated to an automation, to an AI system, so it can focus people can focus on the more valuable, more interesting stuff.
The goal is to augment, not to replace. Let the AI handle the work that’s repetitive, tedious, and time consuming. And that frees up people, your writers, to focus on the crucial parts that require deep critical thinking, creativity, strategy.
So I want to leave you with a question. Think about your own workflow.
What is that one tedious time sucking task that, if you could just automate it would free up you to do your absolute best work. Well, there, that’s the place for you to start.
You’ll find more details on Cherryleaf at cherryleaf dot com. You can contact me to agree, disagree, comment via email info at cherryleaf dot com. And you can contact me, see our activities on platforms like LinkedIn if you search for my name, or search for Cherryleaf.
So thank you for listening.
Leave a Reply