>> Virgil: If you tag everything the same, you're tagging nothing. I mean, that is a reality and we see that all the time. It's a lot of internal organizations when they're looking at their own 365 repositories, their file shares or whatever like that, where they just tag everything and it's like, well, they come up with a tag X, and that tag is applicable to 95% of the stuff they want to tag.

>> Intro Speaker: You're listening to Discussing Stupid, a podcast sponsored by High Monkey. Join your host Virgil Carroll and others as our podcast helps you transform bad digital experiences into teachable moments you can use.

>> Virgil: Today's episode is just, gosh, one of my most favorite topics to talk about. And you'll probably notice when you listen to the podcast that I am very passionate about this. And today we're going to be talking about metadata. And what do I mean by metadata? But the things that help people find stuff through, whether it's your internal site search engine, or maybe it's, public search, but the information that helps people not only find but also understand it. And in all the years I've been doing this, one of the things that I've always noted is that this is one of those areas that everybody wants a silver bullet. They want that absolute answer. And I'm here to tell you it doesn't exist. And anybody that tells you that it does is wrong. many years when I was giving speeches around everywhere, I used to always say that information architecture, you know, kind of metadata and all that kind of stuff was really 20% science and 80% art. And that really is key to understanding that you have to be able to work at this. And it's something that's going to be a continuous process. And today we're going to talk about it big time from an aspect, of talking about, you know, kind of how much tagging do you really do? What is too much, what is too little? And that's just so important to understanding that there is no right or wrong answer. But there are strategies that you can use to make this successful.

>> Cole: Yeah. So, I mean, maybe we should start by, you know, opening up the discussion, explaining to anyone who doesn't know what metadata is, what metadata is. Because if you know you don't, you probably should. And even if you think you do, you probably don't know, you know enough about metadata.

>> Virgil: Yeah, well, that in and of itself is a dicey question, because what is metadata? Well, I mean, honestly, metadata means kind of a lot of different things to a lot of people. for our purposes, it's probably just easiest to look at it as keywords, but it's not really just keywords. It's keywords, it's key phrases, it's basically descriptors. I mean, you know, in, in essence it's something that describes the content and you know, in the context of what we use it. When we start talking about things like SEO, and you know, different topics like that, we're really looking at things that help better define and describe content and, and that really what it is. And that could be an individual word, but that could be a phrase. You know, honestly, it could be a full description. I've kind of seen it all, all kinds of different ways, but overall it's really about something enabling people to better classify and clarify what content is about and how to find it.

>> Cole: Yeah, I think as I've, you know, gotten to know metadata more, I think I initially approached it in kind of like a very boxed perspective. like I thought, okay, I need to understand this. Yeah, I have to like, understand perfectly how it works and because it works the same across all websites, right? Well, I mean, in some ways it kind of does work very similar across a lot of websites, but it's never going to be the same, you know, purpose. It's never going to be implemented in the same way.

>> Virgil: Right, right. I mean, you know, what you say is very true. I mean, the reality is there, there is no right or wrong way. I mean, the one thing I've always seen about metadata, when you start talking about and kind of search engine optimization, everything like that, is everybody looks for that silver bullet like that. You do it this way and it'll be perfect. but that doesn't exist. It has never existed and it never will. Even with the inception of AI, where it's supposed to be smarter, it still doesn't exist in that the reality is there are so many things that go into content. Not only the content, but the context of the content, the context of the person looking at the content and what they're trying to get out of it. I mean, those are all things that affect it. And so when you start looking at metadata, you're really trying to help make it easier to find things. But you know, I mean, you're right, there's metadata and different things. Like on an E Commerce site, a lot of times metadata is really considered like, you know, an attribute. So now I don't just want a shirt, I want a white shirt. I don't want, just want a white shirt. I want it in a large size or, you know, a particular cut or a particular style or something like that. That's all metadata. And I could use that not only to kind of narrow down something like when I'm trying to order, but also find something. You know, I mean, you look at like the, the searches on like Amazon and you see like the 500,000 refiners that are on the side where you have to, you can select everything. That's really what it's about. you're right, it's not in a box. I've been doing this for a long time and I can tell you there's patterns and practices that are very much the same. But overall, the way we implement metadata at one customer versus another tends, to be very different.

>> Cole: Yeah, it's, you know, becoming pretty clear that it's not like a perfect science. There's not like a one all, you know, do all. Is that what they say, One all, do all way. Ah. Of doing it. but, you know, I think one thing that I've really gathered in my recent experiences of doing metadata is, there's, there's a big problem with kind of two sides of it. One, under tagging a site, like the pages across the site, and two, over tagging. And you know, I bet you could speak to both sides of that issue pretty easily. But I, like, you know, we'll get to this later in the, in the episode. But I think there, there needs to be kind of like what, what I have personally coined as like a Goldilocks zone per, per site and like, how much tagging you do per page. You know, find the right balance of it. But let's speak to first under tagging, like, what's, what's the issue with under tagging, a site?

>> Virgil: Well, it's context. I mean, it really is context. If you don't tag anything and you are relying on the talents of the search engine that you use on your site or the public search engine that's crawling your site, you're relying on it to kind of convey the meaning of your content and what it is and that. And where a lot of times, in today's modern search, which I'm going to put AI in that, and I realize there's a lot more complexities to artificial intelligence, but overall, in general, it's looking at your content and it's defining something from that. Well, that's fundamentally what a search engine does, is looking at it and it's trying to understand it. Where you lose the context is it's like a lot of times it's like, well, people have very simple searches that they do, and people that do those very simple searches, you, know, tend to find things, in a very large scope in there. All right, we're gonna have to cut out this part. Sorry.

>> Cole: No, all good. Or we could just keep it.

>> Virgil: Or we can just keep it in and show you how real we are.

>> Cole: We are real people!

>> Virgil: Yeah, yeah. But I mean, you know, you take it like, you know, and I, I've talked about this in many presentations. how many times I've like done usability tests and usability tests. You have somebody search something on a site and you'll say, like, find XYZ form on this site or something. And somebody will go into the search box, they'll type the word form in and they'll hit enter. but that's just the simplicity of how people search. And so if you looked at a typical site and you know, and you know, like a banking site or something like that, how many times have the word form been used? So you're just relying on the usage of the, the word and the proximity of that use and all the kind of just general search algorithms. So if you don't tag anything and you don't give it any additional metadata to really understand what the content is and what is more important, then you're basically just relying on the technology and you're not really helping anything.

>> Cole: Yeah, that makes sense. Like you can't just rely on the word form to get you through your, you know, banking site because, you know, how many forms do you have across the whole thing? And it's like, yeah, but when it comes to, I mean, over tagging, I mean, it seems like, the same logic almost kind of applies because you can have like, you know, a bunch of different types of tags that you're using across. But if you are using the same, you know, like, on an even bigger banking site, you know, if you're using form a million times, it's just the kind of the same issue.

>> Virgil: You're right. Yeah, I mean, and I kind of say it. If you tag everything the same, you're tagging nothing. I mean, that, that is a reality and we see that all the time. You know, I have done this a lot, especially in internal organizations, probably even more so than public websites. A lot of internal organizations, when they're looking at their own 365 repositories, their file Shares or whatever like that, where they just tag everything and it's like, well, they come up with a tag X and that tag is applicable to 95% of the stuff they want to tag. So they tag 95% of that with there. Well, if that's now a refiner on a search page and I click that as a refiner, I have just narrowed 10,000 search results down to 9,500 search results. Why? Because I tagged it. So over tagging is a huge problem. And so you're right, what you kind of said before, you really have to find that Goldilocks zone, which is like, what it is. And so it's really two things. It's one, how much tagging you do, but even more importantly the tags you use. Like, if I was dealing with a form site, you know, that had a bunch of forms on it, I wouldn't use a tag called form, because everything implicitly is a form on that site. And so that's a lot of times what other things that happen when people create metadata is they create really simple metadata structures and they, they come back and they're like, well, but this technically does with that and that does with that, and that does with that, and that does with that. Again, metadata is both helping define, but also differentiate content from one another so that people can better solidify and narrow down what it is not only just from natural search processes, understand. you know, because I mean, metadata is also, you know, alternate spellings, alternate words, acronyms, you know, synonyms, all those kind of things that, that fall in there, but also from the standpoint of, refiners and helping people refine down processes and understand things better.

>> Cole: Yeah, you mentioned refining. And that makes me kind of think, I know there's probably some people listening that are thinking, okay, well, I mean, there's probably a concern about getting the metadata, like, perfect so that it will, you know, give them the best, give people the best search experiences on their, like, website, whatever platform it is.

>> Virgil: Yeah, well, you just create a committee, and that committee takes six to nine months to come up with the initial metadata set. And then they need to take it out to other people and they take another six to nine months to come up there. And that's honestly what happens. I cannot tell you how many times I've been brought into situations where that's the exact scenario where they've been working on this stuff for a year or more because nobody can decide. And there's both because they want everything to be in there or they want nothing to be in there and, or they want it to be too broad or they go crazy and go too narrow. And so it's exactly along the lines of what you said. There's a Goldilocks zone that you have to figure it out. And it's different for every organization, but overall, you gotta go through that process. And this is honestly where, and I'm not saying hire a consulting company, but this is where bringing in an external kind of input is very helpful. Whether it's, you know, an external consultancy or maybe just somebody internally that is kind of away from the process, away, from the site that can help kind of objectively look at things and understands kind of content science and everything goes around there. There is no silver bullet. It's going to be different and it's something that you're going to do and then you're going to adjust and you're going to adjust and you're going to adjust. And it's something that takes a lot of time to really get well. And so you can't just look at it in a box. As a matter of fact, you know, one of the things I think would be a great episode down the road is talking about, okay, once you've gotten metadata in place or refiners in that, how do you really look at the data that you're getting back from that and actually decide what to do next? from there. But, but there is no key there, there is no do this. You're set, you're, you're good to go. It's really a process and experimenting, and you got to find that balance between too little and too much. And most likely up front, you're going to not going to do it well in that, and you're going to have to refine it. But if you do something and you get data from that that you can learn from, how is that ever a bad thing?

>> Cole: Yeah. And regardless, you're probably improving the odds of people finding the content in the first place if you do something about it. And then on top of that, if you keep refining the metadata over time based on looking at the analytics and whatnot, I think that's really on the right track.

>> Virgil: Right. And to kind of wrap up here, if I was to give one piece of advice that people can say is, I'm not going to say it's a silver bullet, but it's one of the big answers is a lot of times what people do is they spend 80% of the time trying to fix 20% of the issues. And I really look at this as, like, a lot of other things. If we can get 80% of the traffic that's very simple, that does very predictable things to be more successful, then we can spend time on other things, trying to make that other 20%. That is so difficult to nail down. But one of the things that really jam up this thing is that people are, like, worried about, like, obsessive, obscure information or obscure, you know, concepts in that versus there. So that 80, 20 rule is probably the thing. As a consultant, I preach the most with all my customers is. Is kind of, you know, let's worry about this over here. If we can make this better, then that gives us more time to worry about this. But if we have to worry about this 80% all the time, too, we're. We're really failing.

>> Cole: Hey, I know it can get, you know, really hard not trying not to get tied up in obscure concepts. So just want to pass on a little, sympathy for those doing that right now, but you don't need to.

>> Virgil: Yeah. So speaking as a metadata, one of the things we'd love is, you know, we're continuing to try to grow our podcast and, you know, get people's help with that. So, you know, if you enjoyed this episode, please hit the like button, and subscribe and also share it with your colleagues, because that's kind of how we get it and that's how our metadata spreads around.

>> Cole: Indeed.

>> Virgil: We do that as well.

>> Cole: Indeed. All right, thanks, everyone, for tuning in.

>> Virgil: Just a reminder, we'll be dropping new episodes every two weeks. If you enjoyed the discussion today, we would appreciate it if you hit the like button and leave us a review or comment below. And to listen to past episodes or be notified when future episodes are released, visit our website at www.discussingstupid.com and sign up for our email updates. Not only will we share when each new episode drops, but also we'll be including a ton of good content to help you in discussing stupid in your own organization. Of course, you can also follow us on YouTube, Apple Podcasts, Spotify or SoundCloud, or really any of the other favorite podcast platforms you might use. Thanks again for joining, and we'll see you next time.