Dave: Hey everyone, Dave here with another episode of the Philly Tech Connect podcast. Today, I’m speaking with Lisa Delaport. Lisa is a career tech writer and documentarian who has worked as a consultant and in eight different startups. Lisa, how are you doing?

Lisa: Hi, I’m well, Dave. Thank you for having me on the show.

Dave: I appreciate that. You’re the first of your kind that I’ve spoken with. I was on LinkedIn, and I noticed one of your roles is called a Knowledge Management Specialist, which is the first time I had encountered that. Tell us a little bit about what a Knowledge Management Specialist is or can be.

Lisa: So, knowledge at a company encompasses a lot of different things. It’s everything from technical bits about your product and the things that you’re building to how do I sign up for benefits to where do we get those chips that we really like, right? Knowledge can be so many different things that keep your company running and allow people to be comfortable. Knowledge Management is the idea of taking on information, distilling it so that people can use it and be empowered to do things properly, and then also just making it really easy to digest and find as your company grows.

Dave: So, I think a lot of companies probably, when they get started, there’s this assumption that somebody new comes on board, they’ll have a chat with someone who’s been here longer, and knowledge will be transferred, and the job is done. Does that work at a certain size, or tell me what’s wrong about that?

Lisa: It feels good when you’re three, four people in a room, and the fifth person comes in, and you all get to talk together, right? That doesn’t work when you start to grow and start to scale. When you have a company that is 500 people, it’s not tenable to do that for every new person who comes on, right? And if you’re a remote company especially, you’re not in the room together, you’re in the virtual world, so it makes things a little more complicated. I think a lot of times folks don’t really think about how they can keep doing that on a scale that grows with the company just because it worked in the beginning, so it must continue working, right?

Dave: How does that problem manifest itself? I’m not aware of some Holy Grail KPI that measures the amount of knowledge saturation that each employee has, so it must manifest itself in other ways. How do we know if we have that problem? How do we identify it?

Lisa: One of the common things I see in startups when they have a knowledge problem is a very stark and very clear difference between the satisfaction and success of early employees and the satisfaction and success of later employees. People who’ve been with the company for five years, who built the code base with you, they know everything, but they haven’t written anything down or they haven’t really documented it in a way that allows other people to read it. So, if you talk to them, they’ll say, “I know everything, everything’s fine.” But then you talk to someone who was hired six months ago, and they will feel sometimes a sense of panic even. “I don’t know what I’m doing,” or “I don’t know where to go to find answers,” or “Half the time I have to just sit around and wait until this person is free to tell me how to fix this thing.” So, it manifests its way in productivity for sure. Are there newer people who you feel like just aren’t pulling their weight, and is it actually that they’re not pulling their weight, or is it that you haven’t set them up for success? And also in employee satisfaction. If you talk to those new people, a lot of times the picture that they paint in terms of how well they’re able to do their work will be very different from people who’ve been here for years since the beginning.

Dave: Yeah, there’s like a massive risk here, right? What if one of those ancient knowledge keepers leaves and all that knowledge goes with them? Do you ever see anything like that, and then what? The institutional knowledge scenario?

Lisa: It’s extremely common, and oftentimes I found it is the triggering event in a company thinking, “We need to start working on Knowledge Management,” right? If developer number two leaves and all of that knowledge goes with them and you weren’t prepared, now what? Sometimes it’s a really kind scenario where they say, “I’m not happy, I’m going to leave in six months,” and you have a runway to take that information out of their head, put it somewhere else, train other people. Sometimes it’s not that graceful, and it’s, “Hi, I’m out on Friday.” So, my advice and the way that I kind of try to instill Knowledge Management in companies is to prepare yourself as if anyone in the company could get hit by a bus tomorrow, and if it is, all of that information would still be within the company, and you’d be free to kind of continue onward without the institutional knowledge that is just contained in that one person’s head.

Dave: So, what does that solution look like? I’ve run smallish companies of like 20 before, so maybe we were kind of decentralized with the way knowledge was kept. There was Slack, there were Google Docs, there were spreadsheets, there was just the pure brain. How do we get started? What does the solution look like here?

Lisa: I think it’s important to, from the top down, from the leadership level down, instill a culture of knowledge, a culture of “this is a thing that’s important, so we’re going to notate it somewhere.” In the beginning, it’s more important to have it notated somewhere than it is to make sure you’re notating it in the right place, right? If you’re in a company of 20 people, all of you were using the tools, it’s a small enough group of people that you have time to find your feet before you find a really streamlined process for Knowledge Management and dissemination. So, it could be, “Oh, I don’t want to forget this, I’m going to make sure that we pin it in this Slack channel so that it’s always relevant for everybody,” or “This is a really good thing, I want it to be in this particular folder in Google Drive, and then that’s just the folder that everybody gets on their first day,” right? So, it can be something as simple as that. Then you start to get into the question of knowledge bases, so these special software and tools and systems that are made to catalog the knowledge and allow people to find it, and that’s everything from, sometimes people will use Google Drive, Confluence, which makes some people scared, or Guru, which is a Philly company—what up, Guru? You know, there’s a bunch of different solutions that people can use, and the one that you pick is totally dependent on your company’s situation, the product, the type of people that are going to use it, the size of your team. But as you start to kind of move through scalability and growth, picking a knowledge base and then setting it up and just putting all of that knowledge in there so that it can be searched for is the end goal.

Dave: The trouble I run into with this wonderful idea of basically managing my knowledge and getting it kind of more centralized is the keeping it up-to-date part, because it just feels like it so quickly becomes outdated. So my question is, this is a two-part question: at a small company, whose job is it to keep this knowledge up to date, and in what cadence should that be done?

Lisa: Good, very complicated question, but let’s break it down a little bit, I guess. So, in terms of knowledge getting out of date, this is the issue with knowledge, right? And I think sometimes because of that scenario, a lot of companies or a lot of people will just say, “We’re moving too quickly, so we’re not going to write anything down because it’s going to be out of date next week,” right? I think that—I understand that, but it’s a fatal flaw in how you view transference of knowledge. Because even if the knowledge is out of date, if you’re properly cataloging it, right, if at the top of the document it just says, “This was last updated October 2023,” the person reading it who’s coming onto it after you can still see when this was last updated and look at that and make an assumption of, “Okay, this isn’t right. 80% of this is probably going to be right; the other 20 is going to be out of date.” It still gives them somewhere to start, it still gives them the jumping-off point to then go on and do their own knowledge exploration, which as a new person is a whole part of your onboarding process and getting up to speed and kind of jumping in with the team. Even out-of-date knowledge, when it’s cataloged in a way that is cohesive and coherent, it’s still good knowledge, right? Someone can use it as step one. In terms of the updating, this is the chronic struggle that a lot of organizations have, especially if they don’t have a knowledge manager. Because a knowledge manager would be the person who would update, or your technical writer, depending on how you kind of peel the onion. I think some of this can be solved with the tools that are coming out now. There are a lot of different solutions where you will get pings: “Hey, this documentation hasn’t been updated in six months.” It’s a reminder that you have to look at it and mark it out of date at the very least, or update it. So, it’s not always up to a human brain to remember to do all of these things. There are other things that are happening in the world of AI too, where we’re having plugins that can be installed in certain Knowledge Management tools that are going to start doing some of that heavy lifting for us. It’s very much still an in-progress practice for Knowledge Management, but a lot of the things that are happening kind of make me really excited for where we’re going to be in five years.

Dave: Cool. Any common mistakes or dos and don’ts you see companies that may be listening, they’re out there right now, and they can go, “Oh, we’re doing that,” that might sort of trigger them a little bit?

Lisa: I would say as much as possible, do not treat Slack as your knowledge base, and a lot of early companies do this. The base Slack is our knowledge base, all our knowledge is in Slack. Slack search is terrible. It’s going to pull absolutely everything, it’s going to just overwhelm a new person because they’re going to get every instance of that particular word, even if it’s not related. The Slack AI plugins also, sometimes folks will say, “Well, this can summarize everything.” In my experience, they’re not there yet. They’re not great at parsing together multiple scenarios, right? So, if you have a feature that has changed names multiple times or changed shape, which totally happens in development processes, Slack will not pick up on that. AI will only look at the thing that you query, so feature X that used to be feature ABC will only be shown in the results. The whole history of it and the context in the beginning will not be shown. Use Slack as a chatter tool, a discussion tool, but any clear insights that happen from Slack that you want people to find later should be pulled out and put somewhere, even if it’s a Google Doc, even if it’s your intranet if you’re a larger company. The place that you’re putting it should be static, and that way it’s information that can be more easily found and understood by the new people who are joining later on.

Dave: Do you find most documentation in this area is written or videos or images? Are those things kind of used as well? Is that useful, or is it unnecessary?

Lisa: Depends on the product, for sure. I’ve worked at companies where the product involves video in itself, so you have demonstrations of the product that use the product, right? That’s one of the best ways to dogfood your own product and document at the same time. I think there’s a misconception that everything has to be written down, but also a misconception that there’s no value in words. So, think about searchability. Most search algorithms are going to pull from the text that is in that particular article or that particular document. If you have all of this information in a video and you have no metadata for search to pull from, it’s only going to pull from the title to tell you what’s in that video. So, I always advocate for a hybrid. If you’re going to do videos or images, which can be really great and illustrative, especially if you’re somebody who’s like building an app and you need screenshots or a really complete picture of your product journey, make sure that the text that accompanies it describes what’s happening in that video or that image. And that way, when someone is looking for that particular thing, even if you have the best video in the world, the person who needs to see it will actually be able to find it.

Dave: Very cool. Lisa, for people that want to probe you more about this area of Knowledge Management and get your insights, how should they get in touch with you?

Lisa: I do have consulting services if folks are interested in that. I have a portfolio on my website, which is delaport.xyz. And I also encourage folks to just start looking at how other small companies are doing this. It’s one of the things that, to me, has been really wonderful. It’s a common problem, and it’s completely acceptable to reach out to another company and be like, “Hey, what tool are you using? What’s your documentation stack?” And looking at what everybody else does can help paint a picture for you as to where you want to go, what your budget is, and where you would see your knowledge direction going in the next three to five years.

Dave: Awesome. Thanks so much for sharing some knowledge with us, Lisa. Couldn’t help with the pun here, and we appreciate you being a member of the PT community.

Lisa: Thank you so much.

Dave: Thanks, Lisa. You too.