 I'm actually still in trouble for something. I wasn't going to tell the story, but I'll waste a minute or two. There was a spider in the bathroom, and my kid's bathroom is between the two rooms. I have a seven-year-old daughter and a five-year-old son. And my wife sent me in, and she half-jokingly said, don't come out until you've killed that spider. And I know she was only sort of 25% joking, maybe. And so she went to sit with the kids in the bed and read to them. And my job was to make sure the spider is gone. And I had this bright idea, and I kind of like smacked it with a towel. And it fell, and I had no idea where it fell. I was like, well, I'm going to be in here a while looking for this spider. And I got in my head this idea that wouldn't it be really funny if I burst out of the room like the spider was on my back and just flipped out? And I burst out of the bathroom just screaming. I'm like, it's all me. I can't get it off. And I'm not a very good actor, but I must have been really convincing. I just was really committed to this role. And my wife screamed. My daughter fell off the bed. My son started crying. It took 30 minutes to calm both kids down. And I was banned for a few nights. I didn't get to choose the Netflix shows for a week or two. So sometimes I might get a little carried away. I was debating what kind of the tone of finishing up day one should be. And when you're finishing up a day, and it's at the end of the conference, you like to send people out on this kind of optimistic, high-energy note. I'm like, yeah, I see you. And here's the thing, though. You need to be here for two more days. And there's a lot of smart people who are going to come up here and try and teach you things that I want you to pay attention to. And I don't want you to get cocky. I don't want you to go away today feeling like you know too much. And so I decided to end on a little bit of a tough love note. And I put that on Twitter. And somebody said, well, Dr. Pete, I hate to say this, but aren't all your presentations either just tough love or scary? And are you just kind of rationalizing that? And I got to say, that was deeply, deeply hurtful. It's probably completely true, but it hurts a little bit. So this goes back to an idea I've kind of struggled with for a few years. And I haven't been able to wrap my mind about how to talk about it until we had some data very recently that I'm getting to in a bit. And it's the idea of we're very vanity-based sometimes in terms of the metrics we focus on. And even something like ranking. Obviously, we're a rank tracking tool in part. We look at that number, and it's about us. And it's about the traffic we get and the volume and how high the ranking is and the engagement. And we forget about the searcher at the other end. And when we rank on that page, it's a sort of promise. It's a sort of promise that we're relevant to that search. And we're going to give them what they thought they were going to get. And this isn't just a question of altruism and trying to be nice people and contribute to search quality. I think we should do that. But it's about our bottom line, too. What happens if we keep breaking that promise? If we keep mismatching those searches? If we keep looking at the numbers and thinking we did the right thing, but we didn't really deliver the promise we made? And how can we scale that idea? And how can we measure that, which is a thing I've struggled with for a long time? But it started about four years ago. This is a SERP from 2014. And this is my first featured snippet that I found I had ranked for. And they had just rolled out a couple months before. And so I was really excited. And it was a nice ego trip. Like, oh, my post got in a featured snippet. And I took screenshots. And I made presentations. And I probably milked this for about four years. So I'm still using it, apparently. And I felt good about it. And then I went back. And something kind of bothered me, which is that this was a weird post. But this is a really bad answer. So it starts out with this post. It has nothing to do with SEO. There's a real traffic driver for the Moz blog. And that was kind of Google's fault, because they weren't really very great at this yet. But this was a post in 2014. And it finally ends with a number. For the record, they made $29.3 billion in 2010. And it kind of bothered me. I wasn't delivering on the thing that people were asking. And this was a little embarrassing. This was just a bad answer to the question. It's a very straightforward question. And so I went in. And I rewrote this post. And I did a little experiment back then. And I rewrote the first paragraph. And I put some updated data in. And I got that to update. And it said, since this post was written in 2011, Google's revenues have roughly doubled. And it gives some more numbers. And I felt pretty good about that. And I used that index for four-ish years and showed that case study. But what stuck with me was the idea that this one query wasn't really that important. This wasn't something we were targeting. We didn't mean to give out bad information. But for this one thing I saw, how many hundreds or thousands of these were there happening every day? Or I was returning this kind of answer that wasn't really very good. And I'm not just talking about featured snippets and answer boxes. I'll get into that a lot today. But even my organic results, even my ad copy, how many times, how many thousands of times in any given day for my content on Moz was someone getting back an answer that really didn't match what they wanted. I'll give you a couple examples. We live here often. And this is Moz Pro. And I'm glad you live here. And thank you for living here and paying us money to live here. But what do we do sometimes? We look at this number, this ranking number, whatever tool you use. We're all making this mistake. And we go, I'm number one for all these keywords. I'm the greatest. And we're all happy. But I want to show you just a very specific example. So this is the new Google Search Console. And I'm just looking at fairly high-impression keywords that we rank number one for. Here's Moz Local. We rank number one. We get an 80% click through rate. This is a brand and product combo. It's not very ambiguous. There aren't a lot of Moz locals. Morrissey doesn't have an album called Local. I don't know what else that would be. So it makes sense that that's a pretty good intent. And if you look at the query, what do you see? You see that Google returns expanded site links. We get six site links. The second and third rankings are both Moz. Google is saying, this is dominant intent. I'm going to use the word brand. I hate to overgeneralize that. But let's call this Google's terms as dominant intent. They pretty much get that this is what they want. Well, here's another keyword phrase that we rank number one for. Do you look at page without canonical tag? 35,000 impressions, kind of surprising. 0.8% click through rate. 1,100th of the click through rate of that other number one ranking. Why is that? Well, this is a learning center page about duplicate content. It's a pretty good page, I think. It turns out the duplicate page without canonical tag is a very specific error in the old Google Search Council. And all the people typing that are looking for information about that very specific error. And we handle that extremely badly. And again, we don't target this on purpose. And this is a perfectly good page. But it begs that question of these two number one rankings are worlds apart. And one is keeping the promise and doing a very good job. And one is completely breaking the promise that every single ranking on the top two or three pages below us is a better answer and is more deserving of being there. And how do we understand this at scale and measure it and get to a point where we can serve this? Because we can't do this query by query we even notice. I can show you these examples, but we can't go through and do this one by one by one. So I want to talk a bit about intent. We've had at least three speakers talk about search intent today. I don't think that's an accident. We're all realizing how much Google is starting to understand intent and what they're doing with it. I'm going to talk about five intents, but I'm immediately going to show you a graph with only three. And I'll explain that in a minute just so you don't get confused. I didn't just type it wrong. You've all seen a sales funnel, something like this. Maybe the ADA model. I'm going to simplify it to awareness, interest, and action. And if you are familiar with search intent and some kind of classical intent, like informational and navigational, I think we can roughly map these in a way. And so at the top of the funnel we have the kind of content we're pretty good at, this sort of informational content. And in the middle of the funnel at the interest stage we have what Google has started to term investigational content. Kind of says that you know what you want. You're starting to explore. You're starting to drill down. But you're not quite ready to make a commitment. You're not quite ready to buy. And at the bottom of the funnel we have this clear action mode, this commercial intent. Transactional is the traditional term, but that's very broad. I'm going to use commercial to really mean this is paid intent, this is shopping, this is ads, where Google is saying, hey, these people are ready to take action. So I want to get into how this funnel maps into something real and what we can do about it. But first I want to tell you a little bit about ice cream. I always try to kind of pick a theme for all my screenshots, just because it makes things more fun and interesting to me and gives me something to play around with. And so we recently took the kids to Rome and took the boat around to Athens. And when we were done, we kind of wanted them to sort of process the experience. They're pretty young. And we'd been all these places. We'd been to the Colosseum, and the Parthenon, and the Pompeii, and the Minotian culture. It's like 4,000-year-olds. It's all this incredible history. And we asked my kids, we asked my daughter, and she said, what was your favorite part of the trip? Gelato. The 17 times I had gelato. And this is not entirely their fault, because when my wife picked the hotel, she noted a very distinct fact about the hotel, which was that it was about 15 feet from the gelato shop. I'm not going to blame the kids for this one. So this deck of the rest of it is sponsored by the gelato industry. It's not paid me for this, but if anyone's from it wants to toss me a few bucks, that's great. So the first kind of intent is informational. And I immediately want to split that in two. This is something I think is really important for us as search marketers. So this, we talk about a lot. What I'm going to call informational open. And that means this is informational intent that we can compete on as marketers. We can appear in these entities. So we talk about this kind of feature snippet a lot. How is gelato different from ice cream? There's an answer. It comes from a site. It has a link. There's related questions. People also ask after that. We can be in these. We can compete for these. This is open content. There's another variety of this that I think gets a little in the middle that we have to be aware of, which is sometimes there's an answer that's pulled out. So when is National Gelato Day? The third Sunday in July. You haven't missed it yet. I always try to find a holiday that is just coming up, which is amazing. You can always find a screenshot for a holiday. Just put when is Blank Day, National Blank Day, and there's a stupid amount of holidays, apparently. This is a bad answer because, apparently, actually, the third Saturday of July is National Gelato Day. And the third Sunday in July is National Ice Cream Day. Don't screw that up. It could be very embarrassing for you. But I think we've got to be a little careful of these, because Google is starting to pull that answer out. And I think that's going to impact clicks and engagement. I have to say, I swear, even though I have a five year old son, this is a complete accident that one of the related questions is what day is fart day. But he did find that completely hilarious. So I left it in. And then we have, unfortunately, what we call informational closed. And these are the things we keep trying to pretend aren't there. And this is essentially the knowledge graph. And we're seeing this expand. Cindy talked today about how far our entity understanding has come. Where's Gelato from? Italy. There's a flag. There's a Wikipedia link. You can't compete for this. This is coming out of the knowledge graph. This is closed. This is still informational content. It sits at the top of the funnel. But as SEOs, as search marketers, we need to let this go. This is gone. Unfortunately, there has been an expansion in the knowledge graph in the last few months. We've seen a growth in definition boxes and in these kind of panels on the right. So a query like, what is Gelato a year ago? Probably had a featured snippet. Now it has a definition box and this kind of sort of weird limited knowledge panel. So Google's trying to answer more and more from the knowledge graph. Again, we can't be in these boxes. We can't compete for this. They're gone for now. And it comes and goes, but this trend will continue. And most of you have seen this. Earlier this year, there was a roughly two week experiment where Google had certain kinds of knowledge cards, calculators, time and date, things that had a very definitive answer. And they took the organic results away. And they just put this button that says Show All Results. If you breathe the sigh of relief that that left after two weeks, don't. This will come back. We already see this. This is voice. We have one result. This is Google Assistant. Type any query in Google Assistant. You get an answer and below it, you get a button that says Show All Results. That already exists on your phone every day. This is coming. And I'm going to show you why in a minute. Because everything below this doesn't really matter. And so we have to give up on these. So two types, informational open, informational closed, massive distinction for search marketers. I'm not going to talk about this a lot. But I want you to understand a trend. Because I think it's important. And I think we have to watch this over a couple years. Investigational content. So you're in the middle of the funnel. You've kind of started to narrow down what you want. But you're still doing your research. Google is starting to not only understand this content, but deliver very specific features to it. So I'm going to show you a couple. Here's a query for walk-in freezers. And if you scroll to the bottom, you see this refined by brand carousel. We're seeing more of these pop-ups, especially for consumer products. What's interesting, if you click on one of these. So I click on Norlake. What it does is it prepends Norlake to the query. It returns Norlake walk-in freezer and shows me a new Google SERP. But what I want you to understand is Google isn't just showing me another SERP. Google is showing me another SERP lower in the funnel that they're now putting shopping ads on. So instead of just slapping ads on that mid-funnel content, what they're saying is, you know what? We're going to try and push you farther down the funnel so you're more ready to buy. And then we're going to show you ads, because then you're probably going to convert better. So Google is already trying to get us to move down that funnel, even within search. And so we have to be understanding that within our own sites. We have to be targeting the intent better. There's another one. This is for freezers. These are mostly on head terms right now. There's this best of carousel. And it has articles, which is interesting. But if you actually click on the product name, you're taking to a query with that product. And look at what's on the right. That is not a knowledge panel. That's a PLA. That's a paid shopping panel. So again, Google is recognizing that investigational intent and trying to push it down the funnel. So we have to be aware that some of this content is starting to get into paid search territory now. And then we have commercial. And commercial doesn't necessarily mean a very specific product. It's whatever Google thinks of this. So Gelato Spoons, I think Google is saying, you know what? That's not very expensive. You're not going to really do months of homework to order some Gelato Spoons. You just want to buy something. And they're treating that as commercial intent. And there's Amazon products. And there's a very clear kind of intent marker here. And then, of course, brand. I'm not going to get deep into brand. We all know this. Here's Talenti, a Gelato brand. Hazelnut is amazing. I don't work for them, but you should have it. They have six site links. They have that knowledge panel on the side. Their Twitter feed is the second result. Very clear dominant intent. Brand is weird. It's hard to put it in that funnel model, because, truthfully, we have a couple things. Something like Mozlocal. That's almost a brand product combo. That has pretty clear intent. You already know about us. You've already got some ideas in mind. You may be shopping around for a product. If I type in Apple, that's pretty top level. Google doesn't know, you're a household brand. They don't know if you want their news, stock information, you want to buy a product, you want to eat an apple, you want to go to Apple vacations. So brand is tough. I'm not going to get deep into it, and partially because if you are the brand, you need to be there. If you're not these days, you're probably not going to compete. I want to show you how this translates into not just hand-waving in pretty words, but real data. Over the last few months, Russ Jones and I have been doing a large-scale click-through rate study. When I say Russ Jones and I, I think Russ has done 80% of the work. And I've done about 20%. But I'll claim it was both of us. I made the graphs. And we're trying to understand not sort of this just big click-through rate curve, but how certain features and how intent are affecting clicks. Because we've seen these anecdotes where certain features pop up, and it just has a massive impact. And we hear these stories, especially from large brands. But we want to try and codify that. And so I want to show you how the way Google has started translating intent into features really matters. And it matters to your bottom line. And it is not just these fancy things that you can ignore. So you've all seen something like this. This is a generic click-through rate curve across a whole lot of data. And it starts in that 30-ish percent range at number one. So that's the rankings across the bottom. And it dips pretty into kind of a power curve. It dips pretty quickly. And then it kind of levels out. This is useful to a point. I think this is more useful than just counting one through 10. And we all sort of understand this relationship. But this is the average of a very, very large number of very different things. And that's what I want to show you today. This is actually the curve for just surps with organic results and nothing else. So if we only have 10 blue links, which isn't that common these days, we actually see a pretty big jump in number one and in number two. We see something in the mid-40s. So already that's something that's kind of different than the average. But it's also telling us that there's a whole lot of other things that are dragging this down. So we're going to use this as kind of the baseline curve. And I just want to show you how different things are right now. This is a brand curve. And we're going to define brand here by just surps that have expanded site links, where Google has said there's dominant intent, no matter how many of them there are. You notice that I was hoping to squeeze all my graphs into 0 to 50%, and this doesn't quite fit. You want to guess what it's going to go to inside your head for a minute? 80% across all brands. These aren't just major brands. Anything with expanded site links, on average, had an 80% click-through rate in the number one position, a massive drop in the number two. Almost double the organic only curve, which was having a higher click-through rate than the average curve. So a massive impact of brand intent here. This is what I'm going to call informational open. This is just surps that had featured snippets and nothing else. It's a little weird because we use quickstream data here. Each of these curves are at least 10,000 surps with these features. We can't actually tell what appeared in the snippet versus on page one. So it has to be somewhere on page one if it's ranking the snippet. So we've combined those. So what we've seen is that combination actually boosts clicks in two through four. But there's a bit of a drop in number one. I think there's a bit of a drop in number one, because sometimes the answer is just given away. And sometimes it naturally leads people a lot more. And so we have to distinguish between those two. But already you can see there's a pretty big difference. There's some opportunity in the middle. And we have to question what we're doing at the top. And we have to see how this is impacting us. This is the one that kind of opened my eyes a little bit. This is informational close. This is surps with a knowledge card of any kind. Less than roughly one third of the clicks in number one. Every single other position well below the organic only. This is a massive drop in clicks. This is why Google put that button there and took away the organic. Because it doesn't matter. If they're giving the answer away and it's a good answer and it's a definitive answer, most knowledge graph answers are definitive answers. Nobody's going to go any farther. They don't need to. And so I want to show you with real data that these are queries that we need to stay away from. We need to let go of this. Google is going to take over these. They're going to move into certain territory. And that's OK. This is the shopping curve. And I don't even know what to say about this. This is PLAs and ads. Not just ads. This has actually shopping ad ads. And number one is in the tank. And number two is good. And I don't know what's going on. So we're going to ignore that. I'm going to just put it on there because it's different. But what I want you to see, and I understand, this can get a little in the weeds on the data. And maybe you're not really in the CTR and you're not a technical SEO. It's not your thing. But I want you to look at this combination of all these curves and just understand that this matters. These features matter. The way Google is trying to determine intent, this matters. These are dramatically different user behaviors. These click-through rate curves are people engaging in your content or not engaging. And so this picture shows you that that average curve is kind of a myth. And we have to let go of that. And we have to understand how dramatic of an impact, something like brand versus the informational close versus the knowledge card, how hugely different that is. The other takeaway from this is if you ever get the bright idea to make a graph that goes off the top of the graphing cell and off the slide, don't do that, there's three hours I won't get back in my life. I'll leave that up there in another second just to get back to my time. So what do we do about this? I try to make things actionable because I know it can be a little scary. There's so much going on and so many things changing. So I want to give you a couple takeaways, I hope. We're pretty good at this. We're pretty good at the top of the funnel. Assuming we stick on that informational open, that featured snippet kind of world, we know how to write informational content. And we know how to target these keywords. We don't know what to do down here. And we tend to go, oh, that's BBC stuff. So I want you to think about two things. One is sometimes it's OK if that's paid search. And sometimes you should talk to those people and see if giving them money makes sense. Because if you have a client who wants to rank for wedding dresses and you're spending thousands of their dollars to rank for organic, you're wasting their money probably because there's PLAs at the top and there's local results. And they're soon going to be paid local and there's four ads. And the same for prom dresses, engagement rings, and all sorts of things where Google has decided this is a commercial query. And if you want to compete on that query, you should at least do the math of whether you should be paying for it and be responsible about that. And I think we all need to be better about that. But today, as search marketers, at the thing we're good at, what can we do? I think we can move up the funnel and try to shift the intent. And get it into the realm of things that we're good at, but on the same general topic. So I'm going to give you some examples. And then I'm going to show you, hopefully, how you can do the research to do this. So here's a query for gelato machines. And Google has already decided they're going to return PLAs. This is commercial. They're going to try and sell you a gelato machine. There's product reviews right after that. There's some videos. We're sometimes pretty good at this kind of brute force investigational query. So we might go and try and rank for how do I pick a gelato machine? Or compare gelato machines or best gelato machines. We've been doing this for years. And obviously, the competition for those kind of queries is really high, because everybody's been doing that. What about something like, why is gelato so expensive? First of all, it's informational open. It's a featured snippet. I can compete. But think about that query. Somebody who wonders why gelato is so expensive, first of all, probably likes gelato. And second of all, would like to find a way to make it less expensive. And one way to do that might be to buy a machine at home. So how can we find this kind of query that can target the people we're interested in that may be good prospects, but that are in the wheelhouse of what we're good at and the kind of intent we can serve and that hopefully Google won't take away in the near future or won't slap full of ads? This is how we usually do keywords. I'm going to exaggerate a little bit. You all know this spreadsheet. We take a keyword, and we slap a bunch of modifiers on it. And then we combine all the modifiers, and we do all the synonyms until we've returned 100,000 keywords, and we feel really great about ourselves because we created a giant spreadsheet. This is a very stupid approach to question content. And I'm going to show you why. I hope it's obvious. So if I start with gelato, and I start to expand it, and I want to create questions, and I say, well, I'm going to use all my question keywords. What is gelato? Who is gelato? Where is gelato? Why is gelato? When is gelato? How is gelato? And I'm going to add some modifiers. Like, what is good gelato? And what is free gelato? And I'm going to add some pronouns. And then I'm going to swap out ice cream for gelato. This gets really dumb really fast. This doesn't work for Q&A style content, and for questions that people ask. And if you send this to your client, they're going to have some questions for you. And their questions are going to be things like, why in the seven blazing fucks do we pay you $100,000 to tell us to target when is gelato, and how does she ice cream? And I'm trying to reduce the F-bounds, but your client is not as polite as I am. And they're right. This is stupid, and we have to stop doing this. But how do we pick the questions that make sense? How do we pick the questions that actual people ask? Well, we cheat a little bit. And we use some of the work Google has already done. And we use hopefully some smart tools. So I want to show you two things on this SERP. So here's a SERP for how do I make gelato at home. First of all, I want to show you that this featured snippet, this is the kind of thing we want. We don't want that quick answer. We want this kind of thing that has multiple steps. Each one of the steps ends in the dot, dot, dot. There's that more items at the bottom. And it naturally begs us to want to click into this and learn more, because this is not enough. That's the kind of thing we're targeting. But on that same SERP, and these often come hand in hand now, you'll see that people also ask box. And you know, hopefully, that if you click on one of these, like, what is the main ingredient gelato, you get something that kind of looks like a featured snippet. This is driven by a similar engine. But you may not know, and Brittany has done some great work on this. When you click on that, you also get two more questions. And you can keep cooking on this and get more and more and more questions. And it tries to understand the intent of what you're doing. And so this is really interesting from a keyword research standpoint, because these are driven by actual search for behavior. So these are questions people really ask. And so my question, my query was, how do you make homemade, or how do you make gelato at home? Already it's telling me, well, how do you make homemade gelato? A little different phrasing that I need to be aware of. How long does it take? That might be interesting. How do you make lemon sorbet? How do you make chocolate gelato? So I already got some modifiers that people really use. First of all, I need to think about sorbet and ice cream and raspberry and lemon and gelato. And I've got some different churns on this concept that really are useful and that I really could combine into useful questions. And what will happen if you eat ice cream every day? I don't know how that fits the funnel, but that would be great content. That would be a lot of fun. And so you can get some really interesting questions from this. Google is also starting to add these sort of filters to some featured snippets. They're fairly new. They're starting to expand. And so for how to make gelato, I get these kind of little tabs, like ice cream without an ice cream maker, chocolate without eggs. And what they're really saying is that those are kind of things that you can add to that question that still makes sense. So this almost matches the spreadsheet, but Google is basically telling us these are the ones that actually make sense to people. So as this expands, I think it'll be interesting fodder for kind of mining that and seeing what people are asking. My only Moz product, shout out, but just because I love playing with this so much. You can see, I didn't see this til after, but yeah, one of my list is ham questions. SEO questions and ham questions. Sometimes you just love ham, sorry. This is Keyword Explorer. This is the feature we kind of vary, unfortunately. I generally start with kind of a not too long tail or not too head term. So I'm going to start with gelato machine. And I'm going to skip, OK, we get gelato machine, Italy, commercial gelato machine. We're going to ignore that. We're going to see all suggestions. There's this pull down, display keyword suggestions, that. And at the very bottom of that pull down, unfortunately, is something called our questions. And that just returns questions. And I love this for keyword research, it's fun. So we get some questions like, what's the best gelato machine for home use? How do I use a gelato machine? How do I make gelato without a machine? Well, that sounds weird, but when you think about it, we also use the thing as kids in the Midwest where you had like the ball and you put the ice in it and you shook it up and you tried, I don't know if that's out of universal thing, you make your ice cream and it sucks. I mean, it's kind of watery and weird, but it's putting your kid in, it's fun and you love it. I've seen some of the manufacturers, why not show people that? Why not put some recipes on there? Because they're going to try it. If they don't like it, they still may come back to you and buy something because they're not going to be very satisfied with that slushy mess. And you gave them some information, you built some trust. But this one I loved, how to open a gelato business? Now, if you're just doing pure relevance keyword research, you're going to dismiss that. But think about it, what do people who are opening a gelato business need? They need equipment. And what do they have or what are they seeking? Money. So this is a fantastic target market and this is not something you would ever get by just slapping keywords together. So I think there are all these interesting questions out there that shift us up the funnel but are still tremendous target audiences. And so I want you to be creative, I want you to use these tools and really dig deep. And this is not as easy as creating that giant spreadsheet, but this is stuff that I think can really generate interest and good customers. Really quick, there's a lot of variations so much to the study. I think stats got to study. This is our data for featured snippets if you are targeting answers. You're still probably going to need to be in the top five or so. Theoretically, you can be anywhere on page one but very few come from six through 10. Most come from one through five. You remember that CTR curve I showed you where being two through four is kind of a boost and being one is kind of dubious. I really think this kind of two through five is a sweet spot. So if you rank two through five for something, it's great to go after that feature snippet because not only will you potentially get the CTR boost but you can hop over number one, it might be easier than trying to compete for the number one spot. So I just want to point out this is kind of the sweet spot to go for. If you're targeting the question you rank 78, you still have to do your job on the organic side. You have to move that up. All right, pointy funnels. Nobody calls them this. I totally made this up. This is what other people call the inverted pyramid. And actually, Steph got in this a little bit and it's funny that my talk is sort of why you should listen to what Steph said. So I hope you wrote it all down and we probably should have done it in the other order but you should do all the things that she said please. So I'm going to give you a technique. This is something journalists use called the inverted pyramid. And the idea is that you lead with the lead. You give them the answer. You give them the summary of the story. Then you get into the details. Then you get into the context. Well, they did this for two reasons traditionally. One was that people have short attention spans. People have very short attention spans on the web but you know what, when you bought a newspaper because you're passing the newsstand and someone was yelling out the headlines or because you read them as you walked by, you didn't have a lot to go on. And even no matter before internet media, people had short attention spans. It's not new. The other thing is that when you wrote an article you didn't know if it was going to be on page one or page 72 and where it was going to get cut off. So you had to put the most important things up front. And this is true of web users. We don't know when they're going to stop. So give them the most important things up front. And I think this fits the Q&A model really well. Lead with the answer. Lead with a clear answer. Dive into the details after that. And then you can get into things like context and data and even sub questions. I'll give you an example. No more. I've gotten off the gelato topic, I apologize. I had not custom created deep gelato content yet so I couldn't give you an example from that. This is from the Moz Learning Center title tag page. We start it right out with what is the title tag? Title tag is an HTML element, et cetera, et cetera. We talk a little bit about how it relates to SEO. We give them the answer right away. We dive a little deeper in that answer. We give them some samples, some optimal formats, optimal length. And then we get into sub questions. And this is a very deep page. And that is rewarded with a feature snippet. What is the title tag? You see that exact first paragraph from the page. So we're telling Google right away, this is the question. This is the answer. We're summarizing it. We're making it easy. We're letting people dive deeper if they want to. I think people have two fears about this. They have three, and I'm going to get them in a second. One is that we're afraid Google might take it away. That's a very realistic fear. A year ago, what does SEO return to feature snippet? Today, what does SEO returns the definition box and a knowledge panel? That sucks. That's how it is. I don't want you to fear it too much for two reasons. That same page, what is the title tag, that same question, also ranks for what does the title tag do? So there's going to be variations. And there's going to be more flexibilities we get in the natural language. That same page ranks for how do I write a good title tag, which is actually much deeper on the page in a sub question. And so I don't want you to be afraid of that either. Don't be afraid of pages that are long form. Sometimes that shorter form content's going to make sense. Sometimes you want to really do a deep dive that has follow-up questions and natural follow-ups people would ask. This is from the middle of that page. It returns these critical recommendations. And this is exactly the kind of snippet we want. Because there's six bullet points. They each have the dot, dot, dot, the ellipsis after them. And they kind of beg people to want more, to go to that article and find out what we're talking about, because they can't get everything out of this. If you're afraid that targeting that question, I know this is a different model for us. I think we look at it, and this is our own fault, too. We go to Keyword Explorer. We go to SEMrush. We go to whatever we do. And we type in that question. We see that it gets zero to 20 visits. And we go, I'm never going to target that, because this low volume. This same page in our Keyword Universe database ranks for over 300 keywords. For many variations of that same question. And keywords including just title tag would no question at all. So don't be afraid the targeted question that's useful that makes sense, because that question will naturally spin out into dozens of different keywords, hundreds of different keyword variants. How do we tie this together? Hopefully it's not too mysterious at this point. This feature snippet, what is the title tag's national element, is right there on that page. As soon as they make that jump to our page, they see that the promise has been kept. We're a good match, and we follow through. I think we're sometimes afraid to do this, because it seems too obvious. It seems like we're beating them over the head. But this is what in UX we sometimes call a scent trail. The bloodhound has got the scent. They picked it up. They're on the right track. We don't want them to lose it. We don't want the user to lose the scent. We want them to know they're in the right place. And we want to make it obvious. And they don't pay attention. And they don't read. And they don't dive deep. So we have to tell them right away, you're in the right place. We've kept that promise. This has done OK for us. This year, this page has generated about a quarter million visits. And so we've seen this a lot. I don't want you to be afraid that you're hyper-targeting or that by answering your question you're getting too narrow, or you're not serving some big head vanity keyword. Because that vanity keyword that has high volume that seems impressive to your boss might not be doing what you think it's doing. This stuff can perform very well. And I think we can take a tip from Google. And we have to do this carefully. And we're experimenting with this. And drive that funnel, too. Here's the bottom of that same page. And this is a long page with a few follow-up questions. And what do we do at the bottom? We tell people, hey, you know what? Here's some other resources you might be interested in. Here's a link to your 30-day trial. If you like this kind of stuff, this is what we do. Maybe you want to try it out. I think we've become almost too obsessed with the bottom of the funnel. And as important as conversion rate is with only content that converts. And I'll tell you something we discovered a couple of years ago. And I wish we'd codified this better. But we discovered that for a purchase, Amaz purchased as a self-service software, took about three or four touch points. So people wouldn't just do the free trial and magically jump right into bias. But people who had seen the beginner's guide, people who had seen a resource like this, people who had read a couple of blog posts, listened to a podcast, watched a whiteboard Friday, and then done the free trial. Those are the people who would convert. So don't be afraid to serve this kind of top of funnel intent. If it's naturally leading you to the right kind of people and to the qualified leads, because then you can move them down the funnel too. And we can already see Googles trying to do that. And we should take that lesson and say, you know what? Hey, if you have gotten all the way down to the bottom of this page, you know you're in the right place. You're the right fit for us. And those are the people we should be going after. All right. That's what I got.