Best Practices for Digital Advertising Targeting with Marty Weintraub - OMCP
- Updated on: 2017-06-30
- Read original article here
Many marketers have over overlooked the deeper targeting options of today’s advertising platforms. Twitter, Facebook, Google AdWords and other platforms are offering targeting data like the industry has never seen. Tie in personas and influencers and you’ll surpass your competitors in market acquisition. Where do we start? What are the best practices? Find out here in this interview with author-expert Marty Weintraub, who is light-years beyond most marketers, and is brilliantly generous with his experience, knowledge, tips, and enthusiasm for digital advertising.
The OMCP Online Marketing Best Practices Podcast is where top authors and industry leaders share authoritative best practices in online marketing which are covered by the OMCP standard, competencies, and exams. This is an OMCP pilot program that may continue based on member interest and support. Staysubscribed to the OMCP newsletter to get new episodes.
Download your MP3 of “Digital Advertising: Best Practices for Targeting” here.
Michael:All right. Welcome back to the OMCP studio, and with us today is Marty Weintraub, entrepreneur, author, speaker, and wilderness guide. We’ll get more about that later. Marty is the founder ofAimclear, author of “Killer Facebook Ads” and “The Complete Social Media Community Manager’s Guide.” He is, in many ways, the world’s leading expert on Facebook and digital advertising. I’m your host, Michael Stebbins, and today, we’ll be discussing best practices for digital advertising. Marty, welcome to the “OMCP Best Practices Podcast.” How are you? Marty:I’m better than a barrel full of monkeys in a rainstorm. Michael:Which, I understand, you’ve had up there in the Minnesota area. Listen, I’m really excited to jump into these best practices. Before we get started, I think there are a few people out there who haven’t read your books, read your articles, or heard you speak. Tell our audience who are you and what is it that you’ve been doing lately? Marty:Lately? My name is Marty. I live in Minnesota. I founded a company called Aimclear. It’s integrated marketing agency. For perspective and not to be boastful, Aimclear won “Best Large Integrated Agency” at the U.S. Search Awards. Michael:Congratulations. Marty:Thanks. Hey, even though it’s the best, it’s not my fave. My fave is winning “Best Use of PR in a Search Campaign” for our own PR against big brands and big agencies. The tactics they’re from, we’ll be discussing today. Michael:All right, Marty, I gotta ask. I know there are several things about you that are just fascinating. I know that you are a wilderness guide. I know that you have musical skills. Lately, you’ve been into photography. I’m just going to pick one of those. You have a piano on the main floor of your agency. Tell me about that.
Marty:Well, I’m a musician by trade. I went to Berkeley College of Music in the ’70s and then went on the road with a band, and that’s what took me to Minnesota. Landed in Minneapolis in the “Purple Rain” generation in the early ’80s and played in bands for years and years and did jingles. And my favorite story about that piano, aside from all the amazing musicians, amazing, like, well-known musicians who have played it, whose karma remain in the keys in all likelihood, that piano has been part of making records that it sold millions of CDs. It’s just my old friend. I’ve had that piano with me for many years. Michael:Generous of you to share it with people at Aimclear and your clients. So beyond music, Marty, I know that you have helped countless businesses and set up strategy for digital advertising. I’m just going to name a few just for the credibility for our audience. I know you helped Uber, eBay, Airbnb, Dell, LinkedIn, Amazon, Quest, Intel, Travelocity, Macy’s, GoDaddy, 3M, and a ton of others. Earlier this week, you and I picked out five, only five of your best practices for digital advertising that are, of course, core to the OMCP competency standards. So for our audience, I’m just going to read them off, and then I’ll ask you about the first one. I love how you worded these, too.
These are part of the standard just with Marty’s words around them. And, Marty, I was reading your blog at Aimclear, and we’ve read how psychographic data goes beyond what’s used in Facebook advertising and can be applied to socialcontent, even business plans. What are some of the steps to start embracing psychographic data?
Marty:Okay. That’s an awesome question. First, let’s take notes, students, that the bullet points that describe what we’re going to discuss today, I’ll begin with action words. So make sure when you’re pitching your own content to people and someone asked what the benefits are, start with an action word.
So the first thing to think about psychographics is the advances that have been made in the last few years. In the early days, it’s like a graphic targeting, aka social targeting, with Facebook being the first real channel to emerge that gave us that ability. It was pretty basic. You can target somebody who was interested in something or showed an affinity for something. The way Facebook knew that was some known attributes and some black box. It’s a lot more than who likes a page. It has to do with a much more set of comprehensive behavior and data. But basically, if you said they like basket-weaving and botany, that increased the audience. If you said basket weaving OR botany, OR botany, the big letter being O-R, OR, the OR operator, that made the audience bigger, bigger. If they’re interested in scuba diving and filet mignon, scuba diving OR filet mignon, then the audience just got bigger because now, it’s all the people that like scuba diving and all the people that like filet mignon. So the greatest advance in the history of psychographic targeting is the AND operator, one more letter, A-N-D, AND, the word that changed all our lives. And we wished for it for years because you could do layers. You could say basket weaving AND botany, basket weaving AND botany, and that didn’t increase the audience anymore. It decreased it. So that’s the greatest development.
And that gives us access to some really amazing targeting layers, if you will. Like take this one, 7,200 people in the United States, their behaviors are that they take real estate investments or interested or somehow execute. We know they’re gonna make various types of real estate investments AND, there’s that big word, AND, they have to have an income that’s over $250,000 or have a net worth that’s over a million AND, another layer, they have to be interested in Florida or the Gold Coast or Boca Raton or in Miami or etc. AND they have to be interested or somehow show affinity for personal pension schemes or self-invested personal pension, which are things that you do before you retire in America. And then they’re between ages 40 and 61. And I’ll provide you with this graphic, Michael. So in the United States, they’re 40 to 61. They’re gonna do real estate investments. They make over 250k a year or have a high net worth, and they’re interested in Florida and Boca and stuff, and they’re showing all kinds of affinity for retirement things. So if I’m gonna sell somebody land in Florida that’s a retirement mechanism, I mean, that’s a serious target right there. So the ability to layer with the AND operator in between buckets, even though I’m discussing Facebook and Facebook partner data right now, this will apply to future iterations. You could basically divide every psychographic display targeting channel by saying how can you layer them. Facebook is also unique because it has an exclude. While there’s other platforms that have exclusions, too, now, Facebook has the exclusion one that all of you have access to today. So you can do negative psychographic targeting essentially which means that, well, for instance, there’s 115 million people in English-speaking countries that are interested in free things, free things like free baby stuff or free anything. I don’t really want to have people who are interested in free things if I’m selling something for money, so I can exclude them. And by the way, if you sell freemium software, still exclude the people that actually want it for free because you don’t want people who want it for free. You want to give free stuff to somebody for a little bit of a trial period who’s going to buy it. So you could put negative keywords, negative social psychographic segments. I called it keywords. That’s something important to keep in mind here for you, search marketers. No matter how sophisticated psychographic targeting gets, we still require words to describe it. So if you’re a search marketer, you knowing the words to exclude is very powerful. So one thing you could do is you could go to your negative keyword list in AdWords or Bing and paste those words, comma-delimited, into Facebook’s exclude bucket and see if any of them parse, and some of them will.
So one strong takeaway so far is map your negative keywords from search pay-per-click or display pay-per-click where you have negatives and see if you could get a match in Facebook’s exclude box. It’s really powerful. The semantic work is done. And when people talk about integrated marketing, well, that’s an integration between search marketing and psychographic, aka social PPC, aka social psychographic, aka DMP, or etc.
It’s important to root your personas when you do social targeting. Like, say, I’m selling a restaurant that’s pretty good. It’s, like, in San Diego, and it’s four stars and three dollar signs, and it’s not four dollar signs or five. And it’s not five stars. It’s not a Beard award-winning restaurant. It’s just nice. It’s a place that you or I would just, like, gladly eat. And that would be described as a non-quick serve mid-range restaurant, non-quick serve mid-range restaurant. There’s a partner category in Facebook ads from Datalogix which is owned by Oracle. Oracle’s been quietly buying up marketing data and platforms. Anyway, it’s a DLX USA. It’s a USA. The purchase behavior is that they’re gonna buy more mid-range restaurants, non-quick serve. They spend more there, and the partner data source for Datalogix is MasterCard. So there is credit card data that says these millions of people are going to go and spend money at a non-quick serve, so it’s not fast, mid-range restaurant, and we know this because that’s where they spend their money. Okay, so layer number one, that’s your route. That’s the thing this target needs to be. And then you can clarify with active filters. Like, use your words, search marketers, eating at a good restaurant. If somebody has an interest in culinary tourism, that’s super interesting because if you’re interested in culinary tourism, it means you like to go places and eat food or dining out with or etc. In different types of targeting, look into Facebook where you type it in and it fills it out, shopping for, shopping with, understanding, learning, searching for, finding, getting, attending, going to, etc.
So now, our first layer was the persona route. And the lesson is to root your personas. Then you clarify the next layer with active filters and then decide if you need any financial clarification which is great and easy to do in America. I already have MasterCard data about spending as the root for this, like kinda don’t need to do any, but lots of other products. It’s great to know if they have money or not. And you can target, you could create a filter layer that’s made up of how people use their credit cards or what types they have or their income or the value of their home or their net worth. And there’s also really cool proxy ways like, you know, somebody is interested in a Rolex GMT-Master II or a Porsche 911 Turbo Yellow or Club Med, what does that mean? Also, if somebody is a brain surgeon or an attorney, what does that mean? So we mark it a lot in other countries like Australia where you can’t get that data, and we look for proxies like that.
So the first two lessons about…the first three lessons are
And then finally, if you don’t have access to the financial, there’s sometimes personas that have more than one route, like a behavior and a class, maybe. The coolest root persona routes in Facebook are in behaviors because they tend to be more complex objects. You all know what veterans of U.S. military are and the honored place that we believe they should have in our society. And as society does its best to nurture veterans, we have veterans’ mortgages, VA loans. What that is it’s special programs for people who serve their country. And there’s lots of companies that market VA loans. And I’m just giving this as an example of a dual root persona. Because in Facebook, there’s 5.9 million people approximately under industry veterans U.S., which is where you wouldn’t expect to find something like this, and the description is people who were previously employed by the U.S. military.
So right, that’s the first half of the persona root, the first half of the root. But there’s a second part of it, and this is 27.6 million people in America, another Datalogix partner segment, and this says people who are receptive to online higher mortgage offers. That’s amazing. Amazing. It’s based on recent consumer purchases. So it’s based on things that consumers buy right before they do that and created to identify consumers likely to respond to offers in this category.
So the two roots that these personas…it’s a dual root persona. They’re veterans, period. Right. It’s a VA mortgage, duh. And they’re gonna buy a mortgage, or they’re really receptive to the offers. Like, put them together and that forms a class of people that you can operate on by geographic area or not. And then once again, dial in your high intent, your active, your discipline targeting and halt their interest in a house remodeling or HGTV Dream Home or “Extreme Makeover” or HGTV Home Renovation, etc. So the interesting part here is the concept of how much we have versus how much we filter– specificity versus scale. As a rule in targeting platforms, most of the time, the more layers you dial in, the more the data costs. Oftentimes, that’s because there’s more mouths to feed. Facebook just comes right out and says, “Could cost up to 15% more if we use partner data.” It’s a black box for how that works. But generally, the more data you use, the more it costs you for a click or for a CPM.
Also, the more you filter, the more you run out of people. If I say, “How many people are named Marty and are sitting in Duluth, Minnesota, talking to my friend Michael for this program,” I’m statistically insignificant, there’s just one. No big deal, might as well just drive there. Michael:And probably expensive to get, too. Marty:Probably, yeah. Just drive there. But if I target everyone named Marty in the world or everyone in the world, well, it costs a lot less to target everyone in the world. And plus, Facebook can use its algorithm to serve ads more to people that are likely to click on ads. It’s an interesting question always, specificity, which is cost and smaller audiences versus scale. Also on the integration tip, if you run a lot more ads that cost a lot less to a lot more people, you create brand search. This has been shown over and over and over again. So in every audience, there’s a critical place of inflection where if you throw enough resources at it and zoom out enough, you not only sell with a much lower conversion rate at a much bigger audience, you also create brand search which any search marketer would say, “I wish I had piles of brand search.” So there’s an integration question when you think about whether you’re gonna zoom out, have it cost less to speak to more generalized people. As a rule, we like to zoom out a little and try to find that sweet spot. Michael:And I know you covered that in your book, Marty. You talked quite a bit about dialing how much brand we’re going to go ahead and put out there, and there’s a great chapter in “Killer Facebook Ads” that covers that. You know, also, Marty, in your book, I remember that you grouped targeting, well, actually, the targeting universe, if I remember how you said it, into literal, competitive, and inferred. Now, that was when you wrote the book. Is there a more developed way to understand advanced psychographic targeting?
Marty:It’s interesting that you raised “Killer Facebook Ads.” When I wrote “Killer Facebook Ads,” I knew it was a time capsule because it was the first book about that that was written. Michael:But there’s still good stuff in there. Marty:Oh, yeah. Thank you. I wrote it so half of it would never be obsolete because it was the theory of how to do it. Like, when I was in college, they taught me how to use synthesizers like keyboards that nobody ever even knows what they are now. What was important was they taught me how to learn, not what, like, so… Michael:The beauty of a liberal arts degree, by the way. Marty:Yes, yes, yes. So with “Killer Facebook Ads,” Chapter 7 about creative is immortal. The optimization one, though, how to run it in Facebook is different, the theories are, right? And the targeting stuff remains foundational. So the answer to your question, we don’t have to infer as much as we did before because there is explicit object style.
I just did a site clinic in Adelaide, Australia, with a dozen marketers where we took on their targeting channels. And Australia’s a challenge because nobody lives there, and there’s no data. 24 million people in the whole country, which is like any large city in the world, and they’re really spread out which makes it really, really kind of messy to deal with. And there’s no way to target people on like how much money they make. So you have to go build buckets of extreme luxury black brands and infer or create a proxy for it. Don’t need to do that as much as we did before, but the sensibilities and being able to do it matters. What’s most important now as you wire together buckets with the AND operator in between is to make sure that the function of each bucket is pure. If one bucket is they show affinity for subscribing to any magazine, the second bucket is they work in the trucking industry and has to be trade shows and wholesalers, and certifications has to be strong trucking business null signals. And then the third one is about decision-maker. They have to have a job that has procurement or supply chain or ire or purchasing or things like that. And so the most important thing, like to keep yourself good no matter where you are, is make the function of each bucket pure so you’re working with known theories. And then to zoom out, all you need to do is remove a bucket. That’s the greatest trick. And so the proxy part, when you get into low data situations, like, when…sometimes, you just have to assemble stuff. Like, there’s no likely to subscribe to trucking magazines. Michael:Right, but you can make inferences with the data that you got, right? Marty:Yes, yes. Michael:Like if they’re…you know, I don’t know what truckers buy, but, I mean, if they’re buying, you know, truckers’ boots or flannel shirts or whatever the stereotype is, now, we can sit there, and if we can envision our target through persona or whatever method you’re using for that, we can say, “Okay, it’s likely that if they’re into these magazines, that they’re likely, you know, in the target category that I want.” Marty:Right, right, right. So the inference is it used to be that the inferences, the proxies we needed to create to get to a targeting place…like, here’s another one. We wanna sell the minivan so we target pregnant people, or we wanna target people for delivery pizza so we target them in between on the matches for the soccer tournament when data says they buy more pizza. And, like, inferences can come from internal or external data, but the point is is that being a great psychographic marketer is about going, “Oh, this reasonably means this.” They’re interested in Danny Sullivan’s Search Marketing Expo PubCon Affiliate Summit, and can’t you, or we’re in, or double-click, or etc., etc. Like, what do I know about them? Michael:Listen, Marty, I want to continue on to influence or amplification, but I also want to make sure. Is there anything else that you want to add that will help people understand advanced psychographic targeting?
Marty:Oh, yeah, for sure, for sure. So what we just discussed is it’s the standard part. It’s the best practices part. I always think that best practices are making sure we understand emergent best practices and take them for a spin. So in the advanced psychographic tips, there have been a few developments that are pretty cool, look-alike modeling, engagement retargeting, and filtered retargeting. They’re all really interesting and they’re all really integrated. Look-alike modeling is where Facebook looks across known data for a custom audience that you create. A custom audience is a cookie pool of an increasing body of cookies that is created either from people that go to a certain page, certain cluster of pages, or any part of your website. And then they can be retargeted by way of a custom audience. And Facebook custom audiences can also come from email addresses, email addresses. Michael:Sure, we do that in Twitter, actually. Twitter will have…let’s upload email addresses and make that targeting. We can exclude and then we can say, “People who look like my followers.” And it actually seems to be pretty good so far, at least for us. Marty:It’s exactly that. What’s really interesting is if you run look-alike modeling on your Thank You Page or any other sales symbol, or if you take the email list of people that have already converted and then find other ones that look like them and market to those people in Facebook or Twitter. So what’s really cool about look-alike modeling is that you can filter that, too. You could make a whole bunch of people that are like the original people, and then you can apply another targeting filter. Like, only target the ones who make 50k or buy outdoor sporting equipment like binoculars. So the tactic is to go to your Success Page, your Thank You Page or take a customer list as emails, make a custom audience, and you don’t care about those people because they’re already your customers. Unless you’re working on retention in this channel, of course, you care about them, but you know what I mean. But then look-alike it, but then you can go, “That’s just great. Now, make sure that they’re this.” So, filtered look-alike is really cool, like, really cool. One other piece, engagement retargeting, when we think about custom audiences, we think about creating them from people that come to your website or people that come from an email list, right, a custom audience or a look-alike they’re from. You can also build custom audiences with people who engage with your social profile, and that works in Facebook and YouTube, Facebook and YouTube. So that means that you could put a video in Facebook, have Facebook host the video, not give users any way to leave Facebook to your website. So it’s gonna cost you a small amount of money because Facebook’s cheaper if you stay in Facebook as a general rule. And then you could build a retargeting list from people who watch the video a certain length of time. And there’s other parameters to define engagement including interacting with a lead form, etc., watching videos of different lengths. So if you’re a marketer where you know it’s gonna take more than one touch, it sometimes makes sense to make that touch in social for really cheap. And you could afford to be magnanimous and not make a hard sell or be very call to action-y about it because you know it’s gonna take more than one touch. And then build that list. And then the next time you see them, hit them with a unit, like a page post or something unit where it’s likely to send them to your landing page or your content. Now, that’s really interesting because it’s retargeting, but I go, well, that’s psychographic audiences because where…remember, the top of funnel targeting in that scenario is complex, like we’ve been describing. Then you capture that as a data preset, if you will, in a cookie pool with engagement retargeting, and keep track of what that audience is. And then essentially, you bought the data, the combined data of the first touch and you’re curating it in Facebook. So you’re buying the data. You’re keeping the list in Facebook. And then when you send them to your website, you could keep the data again either with another Facebook pixel… Michael:Or in your own cookie pool. Marty:Right, or in Google. Basically, sometimes, you go, “We’re gonna do the first touch as a video and spin across 12 sets. Then I build out 20,000 people that do that. Then I run a direct response ad to those people or a site driving ad that sends external, and I send them to my website and build a remarketing cookie pool. And then I run a search campaign that those people.” Like, it’s all weird and blurred in a beautiful way, like, which part is psychographics and which part is search? And we’re using words to describe the psychographics and we’re running search campaigns to social audiences, and that’s what integration is. Michael:Yeah, it’s a beautiful thing. It really is, Marty. And so I think what we want to do is make sure that people spend…all of us, as advertisers, spend a long time inside of your audience selection thinking about, pondering, making connections, and you know, deciding how you can use those Booleans in your exclusions and your layer-ins to really target the right audience. I love the idea of having, you know, early funnel and using Facebook’s retargeting first and then jumping in to pull them into our site and putting them in our own pool. Very, very powerful combinations. Anything we want to cover on advanced psychographic targeting before we move on to influencer amplification? Marty:I think we’ve been pretty advanced here
Michael:It’s okay. You’ve been generous, too. Thank you. Marty:Fun little tip would be that Twitter lets you target people who are individual people as long as it’s a thousand at a time. You could import thousand packs into Twitter and target those people, not people that look like those people, which is really interesting especially from the PR perspective because if there’s 1 person I need to see my thing, I’ll find 999, where I wouldn’t mind if they saw it, to target the 1 person. And it doesn’t cost much to do it. Like, really, we’ve gotten down to individual targeting as long as we’re willing to face some impressions as expendable. Michael:Which doesn’t hurt because you might pick up somebody in the same time, and that’s very clever. We used to do that for schools that we wanted to cover the market motive curriculum that we were doing years ago. And we would pick, you know, anybody who worked at that school, and we would just put up the impressions for them, saying, “Hey, by the way, you really should have this in your curriculum.” And it worked phenomenally well. Marty:It’s really neat. Hey, the early days of psychographic targeting were just, like, amazing. In 2008, we were targeting apprentice electricians of a certain age to sign their old-school boss up for MerchantCircle. Michael:I love it. Marty:Oh my God. Wow. Michael:All right, Marty. So you have this phrase, influencer amplification. We know it works. Tell us how to nail influencer amplification.
Marty:Yes, social psychographic editorial calendar-based influencer distribution. Influencer marketing is maligned a term as anything. It’s like rehab. There’s every kind of rehab. There’s ovarian rehab. There’s digital rehab. There’s drug rehab, etc., recovering from something. An influencer, big data, what does that mean? Could be, it’s whatever we say it means. Basically, big data, and influencer marketing is like that. However, if you can kinda generalize it, you go people who lead thinking, thought leaders, high authority users in any channel where if they say something, it means a lot, and a lot of people will interact. And vertical media roles, like, they’re a morning show host who are interested in marine, whatever, or sports or assignment editors or… It’s really interesting. There’s about 300 roles. I’ll get into that with you in a little bit. But also, it’s important to expand your mind as to what influencer means. How about the union labor boss? If you’re selling a billion-dollar water treatment plant to a whole community, you’re targeting labor who wants the job. You’re targeting professors who might have an interest, targeting treat people who are green and sustainable who definitely care about installing a water treatment plant, the mayor’s office, the city planner, like, community leaders. And you can affect all those things. Thought leaders, high authority social users, vertical media roles, and community leaders, so that’s what it is. Why it is is classic reasons. Links, citations, chatter about your brand, all things people need for various reasons from SEO, algorithms to PR. News, mentions in the news or interviews or placing articles or customer support or to prove out support, proof of performance. Branding, retention, investor, people, like, do financial reports or other money stakeholders. Community relations, how we do with where we live in the physical and the metaphysical kind of way of the world, relating to your own employees, internal relation, media relations, crisis management, media crisis, internal community, investor, customer relations. This is all PR. Basically, Google’s algorithm, with its continued respect for links, respects PR. The difference, to my mind, between SEO, non-technical SEO, non-technical SEO, PR, influencer marketing, street cred, successful reviews, PR, like all that stuff, it’s all kind of the same really. And there’s different ways you could get at that, and different ways you can get that message to those people to do the good that they could do. And when you see different people speak about influencer marketing, they each have their place and the way that they do it. Like, one person distributes to celebrities, one gets writers and sells out space and place articles in Forbes. They have dofollow links and other things like that. Michael:And implied relevance, too. I mean, you know, if I’m getting covered in a union newsletter and the context is strong and it reaches a bunch of other folks, that’s a signal because it’s implied relevance and almost implied endorsement which I think the algorithms do pick up.
Marty:Right. So there’s a few ways to divide up the influencer tool. Things like these, feature writer, writer podcaster, writer and blogger, film director, television producer, content producer, web manager, announcer, television presenter, color commentary, online editor, V blogger, university correspondent. There’s about 300 of them. Correspondence and morning show hosts and editors and directors and etc., there’s about 500,000 of those in America and major European Union nations. You can divide them up in interesting ways. Oh, so that’s the first part. Those are the vertical media roles. I’ll give you the example of applying one in a minute. There’s other ways to slice it. There’s 5.9 million people in Europe who are interested in ghosts and movable-type and WordPress themes and StudioPress and etc. That’s not an absolute signal, but that’s a pretty strong signal that they have at least a passing interest in blogging. There’s 60,000 people who work for the top 50 world newspapers sorted by money, international newspapers sorted by money. And that’s interesting because if you take 60,000 people and you take only the ones who are interested in a very, very specific thing, it’s true. It might be somebody who works in the food service at that newspaper who’s interested in Berlin Fashion Week or whatever, that’s fine. That’s expendable. Like, the person I need to see it will see it. And if the food person is interested in Berlin Fashion Week, do you think he talks to the lady who’s the actual writer for it? You could also come at it by employers to U.S. business magazines like 3,100 people who work for one of a bucket of…oh, and by the way, if you’re building out this targeting, go to Wikipedia and search for a list of business magazines, list of U.S. business magazines. You could discover these things and build them yourselves. Or if they work, 45,000 people work for top 25 newspapers in America by distribution. Now, what’s really fascinating is there’s 51 million people in America who are interested in funny viral things like CollegeHumor, or BuzzFeed, or FunnyJunk, or The Oatmeal, or Memebase, or The Daily Dot, or Hyperbole and a Half. And some of these influencer segments are so big that you can target only the ones who…also, the AND operator like to hang out in extremely viral places which has been shown to increase rebroadcast of your stuff. If you take the 500,000 bucket in America and mash it up with the funny bucket for America, you get 160,000 people. And that’s plenty to keep operating on. Super, duper, duper… Here’s an example of how you use it. Say you’ve been retained by a political lobbying company that’s working against or for a bathroom bill and you want to influence people who write about LGBT issues in America and major EU nations, this is America and major EU nations that speak English, so you take the 500,000 influencer bucket of all those job titles and the funny bucket, so they have to be an influencer and they have to be in the funny bucket. And they’re interested in everything specifically LGBT, like LGBT rights in the United States or rainbow flag, LGBT movement, etc., and that’s 28,000 people, 28,000 people. And so if you put provocative content and distribute it to those 28,000 people, and then you cookie them and retarget them with your next piece of informed hyperbole or effective communication, then you end up getting links, like, all over the place. Can I give you a few examples of what that might look like? Michael:Oh, please. Marty:Okay. So say you’ve got a snarky travel rudeness infographic, like, maybe for one of the aggregators like Travelocity or Expedia, you know, “I don’t like when big people sit next to me and sweat on me. I don’t like when small people sit next to me and walk over me. I don’t like anything, basically.” And you’re just on a bar chart, and it’s everything, top annoyances. Michael:Commiserate. Yeah. Marty:Yes. So you take people who are in the 500,000 job title bucket. And they’re frequent international travelers or upscale travelers or business travelers. They’re deep travelers, maybe determined by how much money they spend on it, and they’re funny people, Cracked.com or eBaum’s World, and they make more than $75,000. So, dear, here’s your snarky, business traveler, upscaled, critic pros. Like, I’m not a travel writer, but don’t mess with me, like, I can influence in that space. These are people. They’re critics, for God’s sake. Michael:They’re articulate cynics. Marty:Yeah, yeah, they are. Here’s another one. Here’s another one. So Aimclear did a whole series of promoted posts in Search Engine Land at Marketing Land last year. And they were all about targeting influencers or very specific targeting like political Twitter targeting. So first, that all showed up in the organic search engine results because it’s legal for that to happen as long as the links are nofollows. So when you buy a promoted post in many publications, just indexes. Then we amplified it to 10,000 people who worked at U.S. newspapers, but they had marketing. It’s just like AdWords or digital marketing or public relations. They worked for them. So they could have been like in the PR department, or they could have been anywhere in those publications that are concerned with marketing, like the stuff that we do. And so we got a phone call from a wonderful writer named Jack Marshall, marketing a media reporter at “The Wall Street Journal.” Like, he wasn’t a contributing writer. This is a columnist for “The Wall Street Journal.” And he said, “I’ve been noticing your content about targeting journalists who work for major newspapers and stuff, and I’ve been seeing your content for years. And I’d like to write an article about companies targeting journalists with ads on Facebook.” Mike, you’re getting this, right? This is a writer who’s writing an article about targeting journalists and ads on Facebook. Michael:He’s targeting inception. Marty:Right, right. He found out about it or was reminded of what he knew about us from seeing us in the past by promoted ads to him, targeting journalists with ads on Facebook. And then that resulted in a couple of paragraphs of quoting me and links to our website that were dofollow. So, like, I just have to ask the question, like, that’s a big deal for SEO to be recognized in “The New York Times,” and it’s a big thing for brand search and stuff. It’s massive. So then were the Facebook ads that did the amplification or the content where I paid Facebook, was that Google SEO? Michael:No, it comes down to influencer marketing, to be honest. I mean, you actually got things early in the funnel and multiplied it to get the attention of people who mattered. Marty:It just goes to show that these tactics don’t live in an island, right? These tactics do not live in an island. Like, which part of that was PR? Which part of that was SEO? Which part of that was ads ads? And you could break them out by channel, but it’s integrated strategy. And, you know, if we had been on it, we could have been calling writers as well or we could have been calling in personal relationships. But when you’re always getting inbound PR like this, there’s not any time. Like, we’ve been cited in so many different publications with people that came to understanding of our company from magnanimous, no Sal, just giving. Your hand is extended. We’re not going to try our product. We’re just saying how we feel. In the immortal words of “The Monkees” television show from the ’60s, “Hey, hey, we’re the Monkees, and we’ve got something to say.” Michael:Marty, I want to jump up to retargeting and what that means. Any last wrap-ups on influencer amplification before we do that? Marty:I’d say no. No, I think that that’s pretty well-covered. Michael:Pretty generous, and we’ll have all that in the show notes as well. So, okay, retargeting, obviously tons of value. I gotta ask. What do you mean by “two hops, baby”?
Marty:All right. So most of the time when we go into really large systems, we look at how they’re segmenting their retargeting. And it’s not functional. Like, they’re leaving money on the table. You know how retargeting works? Oh, and we’re gonna take that around to the concept of team-curated cookie pools, team-curated cookie pools. So you know normal retargeting? I’m searching for a flight from Minneapolis to Heathrow, and then I get pretty far into it. I decide not to buy, and I go into Facebook, and I see British Airways, great fares available. Well, first, that’s a really great example of dumbass retargeting because why doesn’t it say, “Marty, we have the best damn Heathrow flights for Minneapolis on earth, and we’ll give you a cross grade to your Delta loyalty since you’re a diamond.” I mean, they couldn’t know that, but it could know quite a bit about me. Like, that’s a big difference. And it’s a big conversion difference, too. Unlike Airbnb where I’m searching for Hayward, Wisconsin, and I go to read the newspaper or the “Star Tribune” and it says, “Hayward. Marty, you belong in Hayward. Book now for $49 a night.” Like, I mean, one of them is segmented retargeting and one of them is dumbass generalized retargeting. You first know segmented retargeting because if you search for a car battery on, like, the Sears website or something, everywhere you go for, like, the next four days, you’re gonna have four boxes of Sears batteries chasing you around, hunt-you-down-and-eat-you conversion. So funnel retargeting, that’s where they go all the way into the car and they get to stage three and four and bail. So you could serve them an offer that says, give them a reason, sweetened offer, whatever it is, find the next thing you say to them, that kind of retargeting. We think new best practices, what we know. And I know this because I judged the European and U.K. Search Awards every year for having won a number in America. And almost all winning case studies have some element of this in matching data. Basically, everything that you drive in from every channel has to be tagged up. Wherever you’re building cookie pools, you need to segment it. That could be your, like, top keywords. There’s a whole bunch of different things it could be. So we literally have a thing in our contracts that say we’re gonna support the client as they build up their assets to maintain and curate segmented cookie pools at the audience and/or the creative treatment level as we agree. If it’s just one channel, if you’re just using AdWords and you wanna separate retargeting by keywords, you could do it at the targeting level. But if you get cross-platform, there’s no apples to oranges between Facebook targeting and Google targeting or LinkedIn targeting and YouTube targeting. So sometimes it makes sense to segment cookie pools at the creative level. We’re targeting the accountants who work at home. That’s who we’re targeting. However, so just understand that it’s not a straight up metaphor if we keep track of all the targeting. And we build out lists of audiences in Google remarketing, Facebook ads as custom audiences. And LinkedIn, now, has a native retargeting audience. And some DMPs will do that for you. Wherever you keep track of your audiences, all the world’s a list. I mean, I would like to just sell something now. I wanna sell something now, direct response, an influencer that made that happen, a social pay-per-click top of funnel, search engine optimization. I want a cost per click or a cost or a cost per acquisition or a cost per install or a cost per download conversion. And I want it now on one touch, thank you. But if I can’t sell something now, I should sell something later, and that’s what this is, retargeting, filtered retargeting. Retargeting between search and social where you plan on it being two hops, two hop thinking. And if, I guess, I can’t sell something now and I can’t sell something later, I better make some content. The joke is the content marketing is for people that can’t sell in two touches economically.
Michael:Or maybe my product and offer isn’t what it should be. Marty:That’s right. Michael:All right, Marty. How do teams curate this?
Marty:Well, if you work for the social department and you’re amplifying your content, then you’re turning Facebook custom audiences over to your direct response, Facebook marketers, and you’re pinging the Google people to say, “Hey, I’m sending these tags and build a remarketing list of all my maxillofacial oral surgeons,” right? Just sell them dental adhesive or whatever. So you make it happen by…if you’re the search people and…here’s a test list, right? Here’s some example tests, some example tests. First, if you’re doing search pay-per-click, separate out and retarget the big keywords. In almost every case study, there’s one massive or a few massive keywords or modified broad match roots with negatives that are responsible for a lot of the conversion. The first question I personally ever ask in any marketing audit, the first one, sort the whole, everything you’re doing by where you’re spending your money and show me what you’re saying to who, where you’re spending your money. Like, that’s the very first thing I wanna know. So I take the place where I’m spending my money because it’s effective, and I retarget those separately. It’s often effective to put the query from the query report in the creative of the retargeting. But anyway, so that’s a really simple way to understand segmented retargeting. We’re just giving the expensive shit a different treatment. Just really simple, like, if I have a balsa wood airplane or a beautiful new Cessna, like, which one do I care if I crash? I mean, like, treat your expensive inventory. Here’s a more complex one. The keyword is Nikon binoculars sale. And I’m gonna retarget that keyword or that keyword cluster or that ad group into social. I’m not gonna retarget everyone, though. I’m going to filter it by adding Facebook Ads filters. I’m gonna only retarget the ones who are interested in 300 geeky species of birds or Audubon Society or birding, and I’m gonna serve them creative that puts the bird huge in my binoculars and then say, “Nikon binoculars sale for birders,” or whatever it is we say. I’m separating and retargeting my big keywords into social, filtered with vertical creative. Creative is the last mile of all psychographic targeting and filtered retargeting. Here’s another cool example. I drive them in for the keyword, “Nikon binoculars,” and I build an RSLA list in AdWords for that big keyword. And I retarget it to keyword searches for informational searches that are uncompetitive like “Grand Canyon hiking trail.” And when I run the retargeting out to that audience for people, RLSA is just search-filtered retargeting, what I get is…the big query is still attached to it. It’s only the audience who searched for a sale, the one that cost, what, \$8. I don’t know how much it costs, but the retargeting keyword costs like 30 cents. Nobody buys it or cares. But it’s a unit. And then I could go to the ad, say, an expanded text ad, and I can say something about the Grand Canyon and make it feel right. So filtered retargeting mashes together one channel and the same channel or cross from search to social. I mean, that’s what integrated is. Like that’s what integrated is. Michael:So I know, Marty, you’ve covered some of the best practices for digital advertising success that are in the standard, and many of these are on the exam. Outside of that free-forming, any other tips or general guidelines that we should practice?
Marty:Remain human as a marketer. Like, a lot of it is about human motivation. You know, say, you’re selling something that’s a really fast-moving item like a ticket to a concert. You know, you run a top-of-funnel ad. You know, retargeting isn’t gonna really do a lot for us because tickets are gone, like, in, what, seven minutes for the big concerts, whatever it is. So you know that you only get like that bite at the apple right there or a few minutes to work with them afterwards. That’s as opposed to selling a mining truck or a Lamborghini or a 3D modeling printer. The process for selling everything is different as is the sales cycle and the latency of how many touches it takes until when. And ironically as the marketing universe matures, once again, it’s about what you say to who and human motivation. I’m largely replaced at Aimclear. We have a CMO who’s incredibly accomplished and owns all the awards that we’ve been getting recently. I serve as a mentor to some of the people that work here and the whole company for marketing. And what I consistently find myself saying is, “Are we sure that, like, us saying that to this person is actually…” Like, it’s important to always get back to the metaphor. And if I ran into this person in the place for blah, blah, blah, and then if I saw him a second time and you didn’t do it, what would I say? And more and more, I personally am talking less and less about the technical specifics and more and more about the human things and creative. Ultimately, creative is always the last measure of targeting and retargeting, like what you say to who. And so make sure that you don’t sell that process short. Knowing how to do it qualifies you. Knowing how to do it is like getting a record deal. That just means you get to compete with Lady Gaga or whoever, like, you’re getting a…really mastering this stuff is getting a record deal. Then you have to go there and be the best in the world at what you say to the other person to get them to take your action. Michael:Yeah, you’ve got the camera. Now, what are you going to say to it? Marty:Right. Michael:All right. Well, listen, that is all the time we have today. A big thank you to Marty Weintraub.Check out Marty’s Book Killer Facebook Adson Amazon. Follow him on twitter at@ martyweintraubcatch his blog posts atAimclearBlog.com and of course reach out to Marty’s team team at hisaward-winning agency Aimclearto see if you boost your advertising results your brand.
Marty, any other places people can track your speaking or what you’re up to? Marty:Yeah, you can email me, marty at aimclear.com. I won’t market to you. Also, we have a super-secret psychographic social targeting weekly tip sheet that you’d be astonished. It’s amazing. We don’t use it to market. We just send out a thing. I guess it is marketing because about 30% of our clients ultimately somehow touch that list. But let’s put it this way. The team here is really smart, and it’s where I go to see the most important stuff that happened. And it’s bite-sized. So you can email me at marty at aimclear.com for the super-secret psychographic social distribution weekly tip sheet.
I have five different European appearances and workshops and speaking and then more all across America. We’re doing a full-day workshop on psychographics at PubCon this year. Just reach out. I’ve given my email address to over 45,000 people in the last three and a half years. And believe me, [even] you would follow. Michael:Yup. Well, I appreciate the generosity. And again, thank you very much, Marty. Now, I’m your host Michael Stebbins, and you’ve been listening to the “OMCP Online Marketing Best Practices Podcast.” OMCP maintains a certification standard for online marketing industry in cooperation with industry leaders just like Marty.
Join us inside of OMCP to maintain your certification. Get special offers and join other certified professionals or universities and training programs that value the OMCP standards. So we’ll see you inside of OMCP as a certified professional, where we’ll be improving our professional careers and teams together.
