Content Strategy with Dave Davies

In today’s webinar, Dave Davies is joining us. Dave is Lead SEO at Weights & Biases and he’s also the CEO of Beanstalk Internet Marketing in Victoria, British Columbia.

We’ll be talking about content strategy – a topic Dave could write a book about. Watch the video to hear his ideas on how to build and implement a comprehensive content strategy, from know your audience, all the way through their journey.

As always, there’s a transcript below, and of course, we always appreciate the Likes and Subscribes! 😉

Transcript

Steve [00:03]: Welcome to Deep SEO conference, a webinar series. Today we are speaking with Dave Davies, content strategist, all-around brilliant guy. Currently working with Weights and Biases. Today we are going to have a conversation about content strategy and how it can help you with Google and your website, customers and all that fun stuff. I’m Steve Gerenscer. My partner up there is Doc Sheldon. Let’s go ahead and just jump right into it. Welcome to the show, Dave.

Dave [00:37]: Thank you very much for having me.

Steve [00:41]: I’ve been a fan of yours for quite a while. I watch a lot of your stuff. I pay attention from the bottom of the barrel, so to speak. I am not a content guy. That’s not what I do. I can, but I haven’t. My own blog for my own business has been updated once in the last three years. It’s just not an area I’m comfortable in. So maybe we could start with some basics and why this is important and what people should be thinking about when they’re thinking content.

Dave [01:14]: Sure. The simplest sort of 30,000-foot view of content is, that’s what Google’s trying to put in front of people. That is the only thing that matters at the end of the day. It’s the only thing that we can make different arguments about different levels. I don’t think they’re lying when they say you can rank without links, without big pushes and stuff. So that part is not as necessary, but you’re not going to rank without content. It just is the only thing. 

[01:52]: I think a lot of people, myself included, when I think content, you go, what’s your content strategy? I naturally defer to written. I go, oh, okay, I wrote an article on X, Y, Z. And for the most part, that’s a lot of the content we as SEOs produce, is written content. It’s predictable, counting code and schema. That’s what we naturally think about. I think for a lot of viewers and listeners. It is important to also consider going outside of that. This right now, we’re producing content. It isn’t written but we’re producing content and that content has a space in the search results. So even when we’re thinking SEO, not only is there video and audio setups and results, but on just homepages, we see video results. On homepages we see all sorts of different questions being answered in different formats. 

[02:51]: So I think the concept of content can sometimes get a little messy and I’m the first one to make this mistake. I go, oh, okay, I’m going to do my keyword research. Do I do that on YouTube? Nope. Could I? Yeah, but I don’t defer to that as my starting point. So I think that’s sort of how top-level, what we need to be thinking about when we’re thinking of content is the broad brush stroke that is content. 

Doc [03:18]: I agree Dave. I think an awful lot of us are often guilty of defaulting to written content as what it’s supposed to be. Even though it’s not on your own website, things like TikTok, little reels on Facebook, YouTube videos, that’s all content. Basically, anything that could be consumed by users qualifies as content. If you can make it deliver your message. Get the right message to the right people in the right way at the right place, then you’ve accomplished precisely what you wanted to with your content.

Dave [03:57]: Indeed. I write some ads for Google, that’s content. All of this is content. If we’re thinking about top-level content strategy. That’s not my primary focus, obviously. My dominant skillset is in SEO. But I dabble a little bit over there. It’s again, another type of content that has a different format. Paid search on that side, from an SEO standpoint, can be highly valuable to go, okay, let’s refine, let’s see which keywords actually convert before we go and invest thousands, or maybe tens of thousands of dollars trying to chase this down organically. Let’s do a little testing over on the paid search side, see which titles and descriptions are working well for that content type. Or you can just go, I’m just going to give you a few hundred bucks and you’re going to tell me the answer. It’s just how that’s going to work.

Steve [04:48]: That’s always been my favorite approach to SEO and content in general – it’s always been paid traffic because it’s immediate and you know right away. Yeah, that key phrase got you 10,000 clicks, but you only sold one of whatever it was you were trying to sell. Maybe that’s not a path you need to be going down. I’ve worked with a company before that they rank fabulous. They drive staggering amounts of traffic to their website, but it’s people who have already bought what they sell, not people looking to buy it. And it’s been a two year struggle to get them to refocus and point in a different direction.

Dave [05:32]: Right, I think a lot of companies focus on a type, that’s talking about overarching content strategies, that is important content. And I’m sure you’re not saying get rid of that because that’s your loyalty, that’s your community. That’s why they’ll be back because you are that person. You don’t just want to sit with what you have, you want to grow. We all do. What content does it take to push it forward? And then what content does it take to keep people happy and engaged with you and coming back over and over and over again? Using you as a resource. And of course there’s also the content type and my favorite is the resource type that can also be our link bait. That stuff that just naturally gets links. It’s a versatile piece of content. 

[06:22] Let’s look at that. I’m in the machine learning space. So how much faster at a specific task is the 3090 over the 1080 TI video cards? That sort of thing where people will go, is it worth investing $3,000 in this thing because the Bitcoin market’s driven up the price, maybe it is? And it will provide that data. Those sorts of examples. But is that going to make people go, oh, okay, we need to all of a sudden start using your service. Not in our case. It’s the right people, but it’s not actually the content that would drive them to do that because it’s not task oriented to what they’d be looking for. 

[07:01] But a lot of people link to things like versus content, because it’s quick and easy and good. Can you do that to help you rank on the other stuff? There’s all sorts of different, little fun. That’s why we’re having a chat about content. We all sort of get excited about different areas of it.

Doc [07:21]: Don’t you love it when you’re talking to a new client and you ask them, okay, who is your ideal user? “Everybody!” Okay, no, I’m sorry that’s not how this works. So finding the audience and determining who is your audience is critical to me. To me that’s always the first step.

Dave [07:46] It definitely is and that’s another case. You can do a bunch of tests in content, but might be another case, as we were talking about earlier, for paid search, because Google gives you a lot more demographic data when you give them money for it. When you’re going, I’m buying clicks, they’ll go, here’s the clicks you bought. Not as transparently as they used to, but they do still. So yeah, go ahead and drop depending on your budget. So I’ll just throw around $10,000 for many companies to just go, okay, just drop that in there and run a few different campaign types and content types. Run some remarketing campaigns to find out who’s already on your site. Pull that demographic data. You can probably get pretty decent data out of your analytics. But also like where are those ads showing up later? What else are they doing when they’re out in the world? 

[08:38] And then draw your content strategy from that and go, oh, okay, these are the things that my users are actually doing. This is how they’re actually engaging with the site. And here’s how they’re engaging with the outside world. Maybe engage in a LinkedIn campaign to figure out things such as what companies are they from? Things like that, that can inspire you to get a little bit more understanding of who your people are. Because you bring up a spot on point, which is, if you don’t understand them, how are you going to write for them or produce video content or produce audio? How are you going to produce content for them? And then the different types of content you’d want to produce. 

[09:18] When I’m on TikTok. And I love TikTok, but when I’m on there, I’m on a very different mission. When I’m on Facebook, I’m on a very different mission than when I’m over on Reddit or when I’m over on Twitter. When I’m over on Twitter, I’m probably just there to fighting with somebody. I’m just there to just rant at somebody or maybe find out what Barry Schwartz is doing. Those are the two tasks I’m probably trying when I’m on Twitter, maybe John Mueller. Each time I’m in a different spot doing a different thing. So I’ve been thinking a lot about what content you’re putting in front of people. If I’m on TikTok, I’m really easy to distract, but it better be entertaining because I’m there to be entertained. That’s my only ask when I’m on TikTok, is make me laugh, but you know, I have free time. I’m here doing nothing, I’m here just swiping and I’ve got a three second attention span, because my habit is to go, not laughing, swipe, not laughing, swipe. That’s the mode I’m in. 

[10:11] Whereas if I’m over on LinkedIn, I’ve clearly got a bit extra time unless I’m there to do specific research, but I’ve clearly got a bit extra time, but I’m in work mode. So you could probably pitch me on things a little more related to SaaS products or something in my field. It’ll be different for everybody, but you could probably hit me up with stuff related to work tasks, but you’re still going to have to speak to me very different than if I’m in Google search, where I’m looking for a solution. I’ve got a specific thing I want and I’m there to do a specific thing. So when you rank that content, just fulfill my need, that’s it. Just like Dave’s looking up hyperparameter tuning. I mean, I wouldn’t look that up cause I’m not a machine learning engineer, but they would. Okay, now I need to just land them at a page that answers that query in its variety of different meanings.

Doc [10:56]: That’s a great point Dave, because it demonstrates how even the same user needs to be approached in a different fashion, depending upon where you’re finding them. Are they on their mobile? Are they at home asking about pizza? Are they on the street asking about pizza? Are they look for clothes while they’re out at a mall or are they looking for clothes while they’re at home on Amazon or whatever? So, the approach has to change even with the same user.

Dave [11:32]: Very much. To me, it depends how you sort of approach it and how the brain works. To me, this is the part that I really, really enjoy. Both of you I think know my wife and she does sort of UX stuff. So I have a really good and easy phone a friend on that one because I’m not creative in that sort of way. But I am data driven. So I like the data and that’s exactly the sort of scenarios that I really, really enjoy, is that ability to go, okay, it’s the same person. How do I distract them, which is all we’re trying to do? Unless we’re talking just pure Google search, which is how do I fulfill their need? We can get into that strategy, not just their need, but every need from everybody who could possibly be searching for that query, which is different – these are two different things.

[12:23] But when we’re on TikTok or I’ll call those secondary environments. So TikTok is devastating to Facebook at the time of this recording, but at least they pay people. They get big value because of TikTok. At the end of the day just understanding that part is really fascinating and it does require a lot of empathy. And I don’t mean empathy, like I really care for you. I mean that’s fine too. But that’s, a different thing with a different place. But the ability to go, okay, if I’m you and I’m on TikTok and I have my demographic data on who you are roughly, what can I possibly put in front of you to distract you and in a good way? Where you feel happy to be distracted and you want to engage with that distraction. Even if on TikTok. I saw this company while I was swiping, they made me laugh. And that’s all – they moved on. You may never have any idea how they contributed to a conversion later or engagement with your community later that ended up leading to them being a sign up to a newsletter or a webinar event or whatever. These chains might be down the road. 

[13:33] Whereas on Google or going back to that one, looking at content strategy and going, if I’m trying to rank on a page level, you can look at any given query and go, who am I trying to satisfy? One of my favorite queries just because the layout of the SERP results and it’s changed over time, but it’s always my favorite query is American civil war. You just look at the variety of what goes on on that page. It’s got book recommendations. I haven’t looked it up in probably a few months, but I can almost guarantee books are in there. As I just consistently see, there is going to be some videos with walkthroughs of it. It’s got all sorts of components. 

[14:10] But a page that would rank, you’ve gotta wonder if I wanted to rank there, how do I do that? Well, we know it should probably have video because video is on there. And even if it doesn’t rank, hopefully your video will. People might casually be looking it up like me. I’m not even American, but I have looked that up not just as an example for a search result, but just like, oh, okay, what actually happened? So you’ve got somebody who’s just wanting to get a general gist of something. You might also be having a university student who’s writing a deep dive into some things who needs some resource points. So you’ve got that going on. You’ve got just a whole array.

[14:46]  A professor who might be looking for great places to send people to or a reporter who’s trying to look for a resource to add to their piece. So you’ve got this whole array of different people and Google’s going, I need to satisfy them all, right? The best piece hits them all. And also in the format that all of them individually are going to want. So how do you do that? And that’s a really, really fun challenge for me as well. Or do you write an individual piece for each one? And then how does Google decide which one to rank for the university student versus the high school student versus the professor versus, like, Dave, who’s just trying to figure out, too-long-didn’t-read version. 

Doc [15:26]: Another wrench in the works on a topic like that is there is so many hundreds of millions of words already in existence on that topic. So now it’s become extremely competitive to get somebody’s attention, to distract them and make them want to look at your page.

Dave [15:43]: Oh yeah, that gets into just a whole different nightmare, doesn’t it? How do you differentiate when you’re going after something that is… it’s funny, as a search competition, that’s probably for exactly the reason you’ve illustrated. Probably when I’d go, let’s pick about a hundred different things before we get there or we can just go bankrupt waiting to fight that battle. We can just focus really hard and we’ll go bankrupt in the process or let’s start with these smaller battles and stuff and work our way up to that ranking. That’s what’s basically responsible for you to go that route. It is interesting, and then of course that gets into a different content strategy selection, is understand where you are, run some tests on what you can rank for and figure out what you could actually do to pay the bills. 

Doc [16:36]: What’s realistic. 

Dave [16:38]: I’ve seen so many companies, it was alluded to earlier, like what do you want to rank for, everything? Who’s your customer, everythone? That’s great. And in six months you could go from these low-level hanging fruits. Well, that’s not great. No, it’s not. But it’ll keep the lights on while we go for this one. 

Doc [16:56]: There’s a reason we’re not all trying to rank for SEO.

Dave [17:02]: Because there’s going to be 10 results on the first page of which three or four of them are going to get clicks. So four companies-ish are actually going to make enough traffic from that to stay in business. So you got to figure out something else to do. I mean, I’m not saying everything below the fold is kind of irrelevant, but most of the stuff below.

Doc [17:22]: Chose your battle.

Dave [17:25]: So you better have something else going on there than just that one. And of course, then you also get into other content strategies. I know very, very successful firms – we all probably do. And I won’t label them because I don’t want to go and say, these ones, I haven’t seen them rank for stuff. But I know a ton of really successful SEO firms that I don’t see in the rankings. Well, that because they have other content strategies. Heck, I would even consider going to conferences as a content strategy. It’s a link strategy and it gets your content in front of other people. And then you hopefully get some writeups from some other journalists or bloggers who are there and then they’re promoting you.

Steve [18:09]:  I know in the past we used to do a lot of just pure image search work, especially with jewelry stores and places like that, where it was so difficult for customers to figure out what they wanted. They would go to Google image, search and type in engagement ring or skeleton ring or whatever. And just start scrolling through pictures. And that was wonderful for as long as it lasted. And then we had sites like Pinterest come along. A lot of Instagram has gone that route. And trying to help clients understand that yes, it’s social media, but it’s also still a content strategy. And it’s still important because we’ve kind of lost this channel that you were relying so heavily on. You need to pivot and go to this channel or this to try to get that traffic back or that potential customer back. And it’s a challenge just to convince them that these are still the same thing. At the end of the day it’s content. And we need that for your customers.

Dave [19:18]: Well, yeah, a hundred percent. And I think it is a challenge. I think probably every marketer and every genre of marketing has had some form of that same problem. And I never blame the client or end user for that. It’s that we’re equipped with the information, they’re not. But explaining to them that, what is it you’re actually trying to accomplish? You’re trying to get in front of people when they’re looking for images? You’re not actually trying to rank in Google image search because the users aren’t there anymore. They’re over here now. I mean, there’s still a lot of people using Google image search. You need to diversify now. 

[20:00] One of the things and this is my favorite statement in SEO – diversity is security. And I told everybody who listened, don’t go for the primary phrase, rank for a bunch of phrases – it’s better. And you’ll survive more updates from it. Try and rank in a few places. You list a great example. Be on Pinterest. Be strong on Pinterest because people are there. Be strong on Google’s image search because people are still there, but be on Instagram. It’s going to cost a little more but you have to be in more places. But now there are more people in more places and heck there’s the same person in multiple places, who’s now going to see you multiple times. It’s an older study by I think it was 365 had done it and I’m sure that the data has changed. But had found that if people and we’re going back as far as there used to be a right-hand bar on paid search.

[20:52] We all remember about right hand, it’s going back to then. But they found that if you rank on paid search, it improved your click-through rate on organic by 15%, even if they didn’t click it just by brand recognition. So you can extrapolate, it just makes sense. We’re all human beings. I don’t know what the figures are, but I know if I saw you on Instagram and then I saw you on Google image search, there’s going to be a trigger of a reminder. There’s going to be a, okay I can naturally trust this person a little bit more because every time we see something we trust it a little bit more, unless it’s just rubbish, in which case we just are reminded how horrible something is. We can all think of companies like that where we’re like, ooh, yeah, every time I see them, it’s just bad. But they’ve just got reputation management problems. 

Steve [21:35]: That’s where the familiarity with them becomes a little less good. 

Dave [21:41]: And you should just try and avoid that. You’ve probably got some other problems you need to solve other than your content strategy at that point.

Steve [21:47]: And they never want to hear that. 

Dave [21:50]: No, I know, it’s like, if you’re coming to me with a reputation management problem, there’s probably a problem before that. I have seen companies though, where it’s like, oh, it was like this person five years ago and it’s still haunting you today. They were let go five years ago and you’re still trying to get out from under this, I sympathize. Most of them are just like, but you’re still the same person. You’re still doing the same things. So we could probably solve this one and I won’t feel good about it, but you’re probably going to find yourself in the same spot again. So maybe I just have like this rotating wheel of you’ll be a client of mine.

Steve [22:27]: That’s recurring income.

Dave [22:30]: I won’t feel good about it, but somebody’s going to take the money. 

Steve [22:37]: Is a dollar a month high enough? Sure. Why not? Maybe a question for me, because again this is not where I sell at. When you’re looking at a random new client, what kind of process do you go through just to kind of get a feel for where they are currently?

Dave [23:01]: What I tend to do. I mean, there’s the typical tools that we would all use. Like going to SemRush, ahref, whatever. We all have our tools to go, okay what are you ranking for now? I never trust a client really to tell me what they’re ranking for, because they tend not to know. And again, I don’t blame them. That’s not their thing. They might have worked with an SEO in the past or have done old ranking data and stuff like that. But if they have it, they’ll have an instinct just like they’ll have an instinct for what they want to rank for. But not understanding anything to do with competition. 

[23:34] If I didn’t know what house prices were, I would know exactly what house I wanted to live in, but I do know what house prices are, so I’m a little more responsible and reasonable. My expectations, what that house is going to be. And that’s sort of the same spot. Of course, everybody wants the moon, but some of you hopefully will hit the clouds and then maybe eventually you’ll start a SpaceX and then maybe eventually, you can get up there to, the moon or something like that. The first thing I usually do, and it’s a little different with every client depending on what they have. But the first thing I usually do with a sort of stock scenario would be, ask the client to send me what they think they want. And I’ll always word it as what you think you want. But then independently I’ll review their sites. I’ll pull some data on their competitors and then I’ll put together my own keyword research, then I’ll access theirs. 

[24:24]: I just sort of go without your bias being sort of influencing me, what do I see from your site? And then I just send it to the client with just a big list. I’ll sort of reduce it down to what I feel and go, rank each one of these on a scale of one to five. You rank each one, not caring about what the keyword sort of volume is, but just how likely would it be to produce a conversion. Just rank all of these keywords on how valuable do you feel that would be to you? And then from there just going into, do we run a paid search campaign or do we just dive right in? Maybe they’ve got some paid search campaign data, which I would obviously love to see, but then sort of refining down from there and then going, okay, now, based on our competition level, what are we going to need to do as far as like links, supplemental content? 

[25:13]: If you wanted to rank for American civil war. You’re definitely not just going to rank with just that content. We’re talking about a section of your site now and probably a big site, unless you’re Wikipedia, we’re going to need to build something out here. Let’s look at all the different battles that took place. Let’s look at all the different scenarios and people involved and come up with those sort of strategies for interlinking these large sorts of sets of data to make sure that Google understands. Okay, if they’re actually not just interested in the American civil war, they’re interested in these people that were involved and they’re interested in these specific battles, oh, okay, we’ve got something for everyone. 

[25:53] It may not be on that page, but that page references it and then goes to a page that has all that information on it. And we’re seeing all the entities on that page that we would expect to see and stuff like that in the concepts. The reading level is appropriate for the target audience. Hey, they’ve even thrown in a couple of videos, in case people like a video. I think a lot of people for content like that would like a video. I tend to lean to textual content because I can skim it. But that’s why I go to textual. But most of my searches personally, like a lot of them just involve, I’m looking up how to do a thing. So I just want to like right to that part, that’s like, here’s where the steps start. I already know all the filler data on this thing. This is a new registrar. How do I set up a secure certificate on it? 

[26:40] Something like that, where just like, I know all the surrounding stuff. I don’t need an introduction into secure certificates. I just want to jump to that part that just tells me, where is it in your interface because your interface is stupid? And non-intuitive, and it doesn’t just have an SSL button on or something like that. But even that’s sort of relevant. And you can gain some insight, just speaking of content if you put jump links. Not only are those nice little anchors for search engines to go, ah, here’s the part where SSL jumps. But if you set up, I like to set up in tag manager, just tracking on those jump links and go which ones? What content are people clicking on these jump links because if they’re not, if you look at your scroll depth, it might show 90%, if you set up scroll depth tracking or something. 

[27:27]: But if that was just like, oh, they clicked on this jump link. It jumped them all the way to the bottom of the page because that’s where the thing they’re most interested in is, that might tell you something about, oh, it’s not such an interesting piece. They’re going all the way to the bottom. This is actually what they want, it should be up at the top. They actually just want to know how to use our interface. They don’t want to know the history of secure certificates or something like that.

Doc [27:52]: Well, yeah. A good case in point we both know somebody who deals an awful lot with recipe site clients. And one of the things that I see constantly people complaining about is why when I go to a recipe site, I’ve got to scroll three fourths the way down the page before I could actually get to the ingredients, instead of hearing about how Anne Edna decided to do this the first time back after her wedding? Because these bloggers have been reading so much about storytelling being the key to success. So they preface it with 3000 words of story before you ever get to the recipe. I didn’t come to a story site. I came to a recipe site. Jump links. These are very nice. They’re your friend.

Dave [28:38]: Yeah. I think recipe sites are just kind of fun. I’ve picked on them for that exactly. They’re the Hallmark of that sort of angle of things. I’ve seen some e-commerce sites do that with product descriptions, they’re just like, I think Amazon people, it only needs to be that long with like a show more. You do not need to give me the history of version blue widget in size, triple XL or whatever. 

Steve [29:08]: I used to do a lot of work in the fitness and supplement niches. [cracking sound]. 50,000 words and they go on forever. And before you even get to the part where you find out how much it costs and whether you can buy it or not. I convinced one of them to just put an orange bar at the top that says, I don’t care, I want to buy this. And he was shocked at how many people would just click that button because they don’t care about the long spiel, unless it’s really good. Why?

Dave [29:41]: I don’t even know the number and Amazon does a good job, but Amazon – part of their benefit right now and I use them a lot, when I’m thinking of things. I don’t know if there’s a company on earth that has spent more on conversion optimization than Amazon, because like a 0.01% increase in conversion is worth millions of dollars. It’s like Expedia travel. I know they’ve put in their time. So you can gain some insights. Amazon now does have the benefit of we’ve all been there. We’ve all been there a hundred times. So I know right where to scroll to. It’s just right there, but their layout is very friendly like that.

[30:20] I can go, I just care about the reviews. The first thing I’m going to jump to is the review. I’m going to take a look at that. I’m going to first make sure are these reviews for the same item or is somebody doing a switch? A bait and switch here on me and like moving products just to carry over some good reviews from something else. And then, okay let’s jump to that. Once I know that it’s a decent product, let’s actually find out what this thing does. Why do I want these earbuds versus a different set of earbuds or whatnot? I’m just picking the last thing I purchased, which I happen to be wearing right now, on Amazon. Again reviews, you can count that as content too. 

[30:57] But going back to the recipe site, I do agree, I think somebody read too many SEO things or just didn’t let go. I understand starting with the little block we all do. Maybe not that we all do, but I wouldn’t just go, boom, here’s your recipe without any content preceding it. I would totally go, okay, here’s 50 to 100 words. You can probably see where the recipe starts before you hit the fold. So you can see the end is near and then go down into the rest of it.

Doc [31:27] Or at the very least, as you were saying earlier. And at the very least use some jump links. 

Dave [31:36]: And use some jump links. Just get me to the recipe. But one of the things I find really funny on the recipe ones, cause I looked at them from that same standard, is it’s amazing how many of them tell these great stories, but they don’t actually use their keywords very well. They’re clearly doing this for SEO, but you’re not actually putting in the words that I’d be looking for to find out how to make this flan or whatever it is that I’m looking up. So you’ve only used flan in there, like twice. I’m not saying keyword density matters, but it’s 3000 words and you’ve used flan twice. That’s not a way to reinforce relevancy inside of it. But I do know a lot about whatever the beaches of Puerto Vallarta, whatever now, from this great write-up that you’ve done on them. The thing that I didn’t want to know.

Doc [32:32] Well, there are worse things to know a lot about than the beaches in Puerto Vallarta though.

Dave [32:36] It is and I have become a little bit of an expert, but from recipe sites, that’s not an expert on Puerto Vallarta just the beach and the friendly people that will bring you a Corona while you’re on. Something just to make sure we chat well. Something I think does get overlooked by content producers and not the person that we know talking about recipes. Because this actually doesn’t relate to that at all. And I just want to be clear, that person, if they’re listening this is not you that I’m talking about here, because I know you got your stuff together on tech SEO. But I think a lot of people when they’re thinking of content, do think of the words and that should be the first thing you think about is, can I write this?

[33:22] If I can’t write this, who can write this? Let’s find a good person to write this, and you’ve done all your research on it. One of the things I always like to make sure to think about, I think of schema as a content format. In fact, often when I’m getting writers to do stuff, I’m like, well, this is a how to, I’m going to need this, this and this. While you’re writing this piece, make sure I have short versions of this and this and this, especially I’ve worked in machine learning now, so you can’t just drop code in because code messes up schema because schema is code.

[33:52] You need different descriptions and things like that for that. But thinking about schema as part of your code and going okay, let’s keep just a little short list somewhere until you have it in your head going, what are all the schema types that I might apply to content? Oh, this is a how-to? Because what it’ll do is dramatically change what the layout of your search result looks like. A how-to all of a sudden goes, okay, I wrote one on Yolo V5 and object detection to get the little boxes around people. And it now has the 1, 2, 3, and estimated times to complete. The cost is free to do this. You can download all the stuff that’s open source. So it just changes. And my clickthrough rate is higher than the position would indicate because they’ve just got these extra features around it that make you look different and clearly identify, I’m about to show you how to do this. I’m about to show you how to do object detection in windows. 

[34:49] My webcam will show boxes around me as I’m moving. You can do that sort of stuff. And it’s actually quite easy, because I’m not a machine learning engineer. I just happen to follow a video that one of the people I work for or work with had created just to go, can Dave do this? Turns out he can. It’s at that level, but just going, okay, this is easy and it’s going to take you under 30 minutes. Because it took Dave under 30 minutes and this isn’t what he does. This is how this is done. And so considering schema as well. And there’s a bunch of great scheme generators. And I think a lot of people tend with all types of content to go, but I can’t produce the best. It won’t be perfect. And I’ve gone painfully through the schema or gone through all the level. What’s everything I can add that’s relevant in here? But some is better than none.

[35:40] It’s like when you’re doing like AB testing on a website or you’re like deploying it, sort of a new site. Well, we can’t quite make it perfect because we don’t have like the dev skills. Is it better? Is it better than what you had? Then just do it and fix it later. If you can only renovate the kitchen and not the bathroom, renovate the kitchen, don’t wait until you’ve saved up enough for both. Just get this fixed so that your life is better. And then do the other part later.

Steve [36:07] That brings me to a statement, a thing I’ve been doing for the last year-ish, we’ve taken this saying, the things worth doing do it well, and we’ve flipped that to, if it’s worth doing it’s worth doing poorly because at least you’re doing some of it.

Dave [36:24]: And then you run your AB tests, your refiner or pull your analytics. Okay. Here’s what that got me. Let’s refine a little bit. I’ve worked through a lot of designs with people. I’m sure you both have. I’m sure a lot of the listeners have as well, where it’s like, as soon as it hits good enough, it’s good enough. As long as it’s better than what you had, then what are you waiting for? I can’t count a full disclosure on Dave’s failings at following his own advice. I can’t count the number of times I’ve started a redesign on the Beanstalk site. I’d probably got eight half-built versions around and then sort of gone, oh, I got delayed a little bit. Couldn’t make it perfect. Didn’t have time to completely rewrite all the content. It’s like, but you had the content there.

[37:13] The design was better. So just move that content over and worry about the content later. But the fear always being, I’ll never get to it. I’ll be lazy. Well, okay, now nothing’s changed. So now you’re exactly where you were and it’s so much later. There are new styles out. So now you’re on like a seven year old style instead of a five-year-old style or a three year old style. You could have done these things. So sometimes we can be our own worst enemy in following our own advice on that one. And that’s where I failed to follow my own advice. So I try and be a little sympathetic with clients when they failed to follow my advice as well.

Doc [37:47]: Several years ago, Ian Lurie, of Portent, put something up on his blog. He referred to it as DFP, distance from perfection, which I globbed onto, I thought that was a great way of describing it because you can’t optimize. You can improve, but you’re never going to hit perfection. So what you do, is you hit those little nickel and dime steps to get you closer to perfection. And especially these days. There was a time, not that many years ago when there was one or two things you could do that was going to have a dramatic effect on your rankings with Google. Now there are 72 things you’ve got to do to have an equal effect because there are a lot more ranking factors involved. There’s a lot of different weighting that’s taken place. So you’ve got to attack these little battles and continue to improve because you’re never going to actually hit that optimal point.

Dave [38:50]: That’s a great point and you never will. As long as you can come up with some good benchmarks, I’ve been part of a couple where I’m like, it won’t be perfect in design, but we’re having to totally change the build. The framework is totally different. The build is totally different. So let’s just make sure all the boxes and positions of things are in the right spot. And let’s just make a sprint and just get that done. So then we’re at least in the position to start to make changes. Because right now we’re trying to do eight things at once. We’re trying to change design. We’re having to deal with a million different opinions. Let’s just make it look exactly the same. But in the framework we need it in because we’re moving to JavaScript or something.

[39:30] Whatever you’re doing, just move it over there and just make it look exactly the same because then at least it’s in the framework and then your SEO can start to go, well, there’s the impact of that framework. And I can figure out my thing. I might even argue for that and go that’ll let me know what the framework did. And then we can start to figure out what else to do while I know what this did, rather than waiting for this magic day, when all of us have all the time in the world. That will never happen. So we’ll be waiting forever rather than going, alright now we’re done that, let’s fix this box, let’s fix our search feature. And I mean, none of us are going to be perfect. Google’s right now testing whether that little like magnifying glass should have color on it.

[40:10] They’re constantly testing. Doc to your point, these minor little things, just to go, can we make users use this better? Will users like this more? I don’t know. That’s not a direct SEO thing. Other than it being on Google, but it’s not, they’re not doing it to rank better. I think they do okay on traffic. I remember once at a conference I was talking about the importance of testing. And I went through a series of slides and I’d gone, like here’s a change Google made, here’s a change Google made. Here’s a change Google made. And I went through and there were 17 of these slides. I said that was last month. That’s it, 17 changes that they had made just to their layout for users. That was a month.

[40:49] It was a busy month. Normally it’s probably only 12 or 13, but it happened to be 17, which is great that what the month after. I was doing my thing at this conference. If Google’s willing to put in that time and still after the massive efforts that they’ve put into user retention and user engagement and the massive amounts of data that they can collect. If they’re still doing that, why are we thinking, oh, we’ll just hit it out of the park on the first go. And we’re just going to wait for that magic day when we can just get this all done at once. They didn’t and they went, let’s just do incremental refinements over and over and over again until we’re dead. 

Steve [41:32]: They do have the advantage of a significant number of data points coming into the site with the volume of people that hit those pages. They can run split tests in a couple of hours, where some clients need a couple of months to run a decent split test, just because they don’t have the traffic.

Dave [41:52]: That’s very true. And optimizes is horrible. I’ve seen that and sort of walk clients through that and they’re like, well, let’s to run these four. And I’m like, that page has like 40 visits a day. We’re going to be showing 10 to each person forever. This is not a quick data collections, let’s just pick our two biggest or run a paid search campaign and just dump it there, understanding that paid search and organic search are actually going to drive different people. Sometimes you’ll just hit, like it’s so overwhelmingly different.

Doc [42:25]: Two things that drive me away immediately when I see somebody posting about a study or an experiment they’ve done, is A, no control and B, a sample size of like four sites. Both of which are totally meaningless. Add three zeros behind that and maybe you’ve got something worth looking at.

Dave [42:47]: Yeah. I think that’s been a pet peeve of all of us, when we see something become a thing, that actually didn’t have a basis and enough data. I don’t blame people for talking about stuff and going, oh, I did this thing. I would sit at a conference with you Doc, have a drink and go did this thing and this happened? But it would be casually over a drink and going, here’s a fun thing. If anybody else wants to know or test it. Here’s a fun thing that you might want to know. And here’s what I found out from it. Not here’s a big article that looks authoritative because of where I put it, going, this is going to be the key to getting you above the fold on page one from starting from page eight. 

Doc [43:27]: But no online pub would ever publish something like that.

Dave [43:34]: No, we all know the ones that would, or videos that would be made to cover those things. And the thing is, I don’t even blame some publishers because there are even publishers that I don’t think have the skill. They’re a publishing company. They’re hoping that we’ll police ourselves, which I think Google has learned from the past don’t trust SEO. That’s just a horrible idea. Give them a green bar and all of a sudden they’re selling links for like hundreds of dollars a month because we told them how much value it has.

Steve [44:13]: We may not have owned several hundred link directories in the past.

Dave [44:20]: And that’s the point. You just can’t give SEOs nice things. I don’t blame them for not telling us stuff. I appreciate that they go, yeah, there was an update. What was it about? No we’re not doing that. And I understand, especially for some of the newer people who don’t remember the wild west that was. They’re like, but why not? And it is so complex and it’d be really hard to tell, but I think in a lot of cases they’re pulling in so many machine learning systems, like ever since RankBrain and NeuralNet – It’s all just become so complex. It’s search quality, that’s all they tell me. Because there are probably about 18 engineers involved in just that one change. Maybe 1800 engineers involved in that one change. 

[45:09] And none of them quite know what’s going on, because they programmed a machine to figure out what to do. And they don’t quite know what that machine did, but people seem to like it. That’s kind of all they’ll know. And that it seemed to focus on these areas. Seemed to focus on link-spam, or seemed to focus on improving clickthrough rates and finding sites that have higher clickthrough rates and traction or whatever. 

Doc [45:31]: And sometimes it’s as simple as giving us essentially the same data that we were already getting, but using 12% fewer machine cycles. So we’re alleviating our hardware burden. 

Dave [45:46]: I love hearing you say that. I think that is the one thing that is the unsung hero of what machine learning engineers at Google have to contend with. And where so much power. I’ve read a couple of their papers on it and just the amount of effort that goes into producing nothing but cheaper. Just going, we have so many machines doing so many things. We just need to make that cheaper so that we can free up those machines to do new things. So that we can do the next thing. So I find that really interesting that you identified that as well, because it is just a fascinating, but just completely undervalued area.

Doc [46:34]: And you can put in a search query and in 12 milliseconds, you’ve got back 2.8 million results for that query.

Dave [46:44]: But there were two spam results. I don’t like that. 

Doc [46:49]: But I mean, the ability to, to even find the index in that few nanoseconds is incredible – to actually pull something for page one only that’s worthwhile is phenomenal.

Dave [47:05]: If I had a list of bookmarks in my top nav of the 10 sites I visit the most often, it would take me probably longer to look through those bookmarks that actually just Google the name of the site and pull it up. Most of my bookmarks are just like, I need to remember where this thing is because I can’t find it. It’s a notion doc or something like that. Most of my bookmarks are stuff like that. Now the rest of it’s just easier to just Google and go there it is.

Steve [47:35]: I don’t want to shut us off too fast, but we are coming closer to an hour. So maybe you’ll give Dave a couple minutes to tell us what you’re into, how people can get ahold of you, if you want them to. And kind of wrap up what you’ve got going on.

Dave [47:56]: Sure. I’m the lead of SEO at Weights and Biases. So I basically help people find people that teach machines how to think. So that’s kind of a fun job for somebody who’s a nerd. If you want to find me, the easiest place would be Twitter. It’s just beanstalkim, like the letter I, letter m, like internet marketing is my handle over there. And I’m always happy to answer questions and most of my feed is just SEO stuff. I start every morning by reading all the SEO news. Thanks, Barry Schwartz. It’s not the only person I read, but most of the stuff I find is from him because he’s just on stuff first. So most of the stuff I share there is about what’s going on and hopefully people find it interesting and feel free to reach out, happy to answer any questions for folks.

Steve [48:48]: Really appreciate you spending your day with us. That was an awesome discussion. Thank you very much. Glad to have you and I want to thank everybody for watching the show. If you enjoy what we’re doing, obviously we’re going to do the YouTube thing. Subscribe, please. Like the video. I’m sure there’s a process I’m supposed to be using to get people to do that, but I don’t know what that is at this point. And we will see you all in our next webinar. Thank you very much.

Doc: Thanks Dave, great having you.

Dave: Thank you.

Recent Webinars