Advanced Pay Per Click (PPC) Certification Program

Certification Training
2380 Learners
View Course Now!
42 Chapters +

Remarketing Tutorial

5.2 Introduction

Hello. This is Brad Geddes, the author of Advanced Google AdWords, the founder of Certified Knowledge, and the PPC Faculty Chair for Market Motive. In this video, we're going to look at remarketing. We'll look through how it works, how to set it up properly, and some best practices for making your marketing campaigns effective for you.

5.3 Overview of Remarketing

So first, to recap how remarketing works. First off, you will create lists, you'll define your audiences, and then you'll place a code on your website. Second thing. What happen is, the audience you've defined, they'll visit your site. When they visit your site, a cookie's placed on their browser, which is related to the list you created. Then the user leaves your site. So now what happens, is when the user is somewhere in the Google Display network, you can market to just that particular audience list, they can see your marketing ad. So if they see your ad, they then click on your ad again and they come back to your site. So, that gives you another chance to, attempt to convert the user. So when we look at remarketing, what's important it's how we define our audiences, so our code's placed correctly. Then we need to look at our ads themselves, what ads are the users seeing, and then what behaviors do they come back to. So, we'll cover all these items within this video.

5.4 Using Google Analytics for Remarketing

Now in setting up remarketing you can do this either in Google Analytics or in AdWords. They're slightly different how you create them so walk through the basics of both. So in Google Analytics, you can go into your administration area and then into their marketing section, and you can create lists within Google Analytics remarketing area. Now when you create these lists, you have a lot of options within Google Analytics. You'll select your profile, and you can make a list that's all your site visitors. That's normally not recommended for reasons we'll get into in a little bit. Then you can make a list for people who visited a certain section of your site. It could be your shopping cart, your home page, within category pages, or you can make a list based upon someone who completed a conversion goal. Now one of advantages of using Google Analytics over AdWords, is you can market off a conversion goal. So a conversion goal in analytics. It could be your relic, a thank you page. It can also be something like, someone who spent three minutes on your site or visited five page views. So with using Google Analytics remarketing, you can do interaction type of remarking, while in AdWords you can't we'll cover AdWords in a second. And you can even create specific segments and filters on your site and apply people to a list, based upon filters and profiles that you're creating within analytics. So analytics is really powerful to use remarketing. Now, the issue with Google Analytics, is you must modify the Google Analytics code to not call in the typical google/analytics.com ga javascript. You got to call in instead, a double-click.net javascript. So that if you never have installed Google Analytics, it's really easy then to modify the code to begin with and then install it across your site. If you've installed Google Analytics already, you've done a lot of code customization you might not want to change it. If you're using a plugin based upon your content management system to have analytics fire across your website, your plugin may not support changing the code around. So with Google Analytics, if you can't change the code then you need to use the AdWord scripts.

5.5 Using AdWords Script for Remarketing

So with AdWords, then you can go into your shared library to see all your audiences. And if you created your list within Google Analytics, you can still see your list inside of AdWords as well because AdWords is how you're creating your ads and landing pages for these lists. Now within AdWords you can also create lists. So in this case, you'll get a simple script that you can put across your website and then call in users based upon certain URL strings they visited. So creating lists in Analytics is really simple, but it's mostly based upon URL options. The URL contains something or it doesn't contain something. So with AdWords, all the scripts are just based upon URL matches. Where Analytics, they can be URL matches or goals, and even certain filters. Now within Google AdWords you can create a list and you can even do combinations of lists. So in other words, you could have one list that's people who visited your site. You could have a second list which is people who completed an action. And then you can make a custom combination that says people who visited this part of our site but did not convert. That's who we want to market to. Or we can say, people who visited our site is our list, but if they're also in the secondary list of completing an action, don't show our ads to those users. So, it's sort of like positive and negative keywords with lists. You can make one list that's positive, another list that's negative. You can make two lists that are positives. So really, what it comes down to is how your website is set up. So now, once you make these lists based upon your URL strings or goals or whatnot, depending on if you're using AdWords or Google Analytics to create the list. Then you can go into your ad groups and you'll have to go to the Display Network section under Interests and Remarketing, and you can then add your list as additional targeting options within your AdWords account. So you can have your positive list just like you would have your interest tags for interest targeting within the Display Network. And then you can also do negative lists either from a custom combination or just making them an ad group level or campaign level exclusion, with the exclusion availabilities within the Display Network. There's a lot of different ways to get to the exact same outcome, whether you want to use positive lists and negative lists or just custom combinations where you've already included and excluded users. So either way gets you to the same outcome at the end. So the one issue with remarketing, though, is you have to have at least 100 people in your list before the list can become active. Now you can also list how long the cookie lasts for, so you need a 100 cookie set. So if you have a page that gets 50 visits a day and you make your list only one day long, it'll never go active because every day your 50 cookies expire. Now if you have a page with 50 visits a day and you make your cookie duration at least three days, then you'll always have at least 100 people in that cookie pool, or that list, so that that list stays active.

5.6 Determining Lists

So what we really need to do when thinking about our list is consider our offers. What's the ad we want to show to a user, and what's the landing page we want to use based upon their previous behavior on our websites. Now the biggest mistake with remarketing is someone sets up remarketing list. They put it across their whole site, and all they do is say, if you've ever visited my site, then show an at. This is a terrible way of doing marketing, because you don't have any segmentation. Even if users actually converted, you're still remarketing back to them. So what you want to do, is at a minimum, do some segmentation. Even if it's a basic one that's, they visited my site, that's my positive list. And they converted, that's my negative list. But when you really start thinking about your users, they all have different needs. You're going to have some people who are shoppers, some who are actual buyers, some who are purchasing agents, others who are doing research to give information to a third party. You'll have some people who are researchers and look at all kinds of information they convert others who are impulse buyers. So your users are all different.

5.7 List Segmentation

So, what you want to do is go back to your site again and start thinking about your ad and landing page that you want to send a visitor back to. And then, based upon that, start segmenting your site. So for say market motive site. Your one set of big segmentation is what type of learning are they interested in? Is it SEO? Is it PPC? Is it social media? Is it analytics? And so forth. That's your big segmentation list, often it's your category list or your navigation on your website. Now we have a secondary section we can also do, so self paced online courses verses coached courses versus groups. You could also include contact us, so. What you really want to do is start thinking about the reason someone visits your site. What do they need to know? And what are they interested in before they convert. And then start thinking about how you get a user back, based upon the behavior they've shown on your site. But they didn't finish a conversion. And that will help you start thinking through how you want to create the lists to show a different ad to a user. Because if you don't segment the site, in this case we can just show a MarketMotive ad to everybody. But if someone went to the social media section you'd rather show them an ad about social media training versus just online training. because now you're connecting to a user based upon their primary need and want from visiting your site in the first place. So when we start thinking about our segmentation. We also need to start thinking about how long we set our cookies for. So looking at days to purchase is really important. You can pull this data from analytics. So this is a site's information which is a membership based site, and they have a seven day free trial. So when you look at days 0 to 5, very low conversions because it's a 7-day free trial. And this is days to actual revenue, not just signing up for a trial. Day 6, huge amounts of conversions. Now what happens, though, is many people take more than five months to convert. So then you probably have one list which is going to be seven days and really hits people who sign up for a free trial but haven't yet converted. Just remind them, hey log in, try it out, make sure you understand the benefits of this program. And that's one list of converters, but they haven't actually bought yet, they're free trial users. Then you might have some other segmentation lists based upon the different benefits of the program that a user has shown interest in that last much longer because it takes a long time, in this case, for many people to convert. So, when you start looking through, here's my segments, then here's how long it takes someone by segment to convert. That gives you that initial information. Then you take that, so you have days to purchase by segment, and then go to your analytics program and set the days to purchase as your view state for your time range. Because then what you want to look at is does this page, or even categories of pages. It just doesn't have to be a single page. It could be multiples. Have enough visitors in that time frame to even go active. Does it get at least a 100 visitors in our time frame? If the answer's yes, great, you can make a list. If the answer's no, then you either have to combine more pages together to make the list go live, or you need to increase the cookie duration to make the list live. So, once you have, here's your segments. Here's your days to purchase. You've looked at your page views to make sure you can set enough unique cookies for those lists to go live. Then you'll go back into your account, whether its AdWords or Google Analytics and start creating all your lists. Now you don't have to make these lists live right away. You can make them see how many people actually join your list and just start collecting cookies before you even make something live. This is really useful say at holiday time where you might even start collecting shopper lists over time. You don't make the list active, and then it hits right around some holiday or special event, and you push the list active. So you don't have to have a list being actually marketed that moment in time to collect the cookies. A list collects the cookies. To make a list live, you then need to go into the display section and into interest and marketing. And then set the list as one of the targeting types. I find that it's usually best to have one list per ad group. This lets you then control the ads for that list and even do some ad testing for each of the audiences. I also like usually calling the campaigns remarketing and segmenting, remarketing from other types of display advertising. So you can control the budget and look at the data separately. 'because with remarketing, you're bringing a current user back to your site. Someone who's already been on your site. With a lot of the display options you're acquiring new people who have never been to your site. So remarketing's very different because you can make a list based upon someone who came to your site from an email that you sent out. You can make a list based upon someone who came from organic search or social media. So your lists in AdWords that you're using for marketing can be really targeted, and they may have never come to you before via an AdWords click. It might have been a completely different channel. All that remarketing cares about is they were on your site, they were in a URL or hit a goal or what not that caused them to be on a list. Now they can see your ad. So you can have an average account, but to only remarketing, and you're remarketing based upon other ways that you grab the traffic and brought it to your site initially. Then you use remarketing to reinforce the messages or to bring back the non-converters.

5.8 Ecommerce Example

So we previously looked at an example from Market Motive which could be a membership site based example. A small business example. You have contact types. You have categories, so forth. Let's now take a look at an ecommerce example. So here we have LANDS' END. And so we think again about our segmentation, all visitors would not really let us show a very good ad. So our first level of segmentation is our main navigation. Women, men, kids, swim, shoes, etc. Now even that is still fairly generic. So then we may instead say all right within the women's section our categories are swimsuits and shirts and blouses and sweaters. So now we can show an ad back to a user based upon the type of apparel they were looking at. Now what'll happen if you go to that much segmentation is that the same user will be in multiple categories. They're going to be in sweaters and swimsuits because they looked at them both. So you have two options with this. One, you just ignore it because they were interested in both products, they looked at both products, show them ads from both lists. So that is an option is to ignore those conflicts because a user has shown interest. Second, with display Network Google still does ad rank. Your max CPC times your quality score. So if you find that your swimsuit category marketing list converts better than your sweater category marketing list, your CPC's will probably be higher. So then, you're naturally going to show the list where your CPCs are higher the most often to the user. So a user, if they're in multiple lists, you could ignore it, or you could let your CPCs dictate it, or you could go to a crazy level segmentation of swimsuits, but not sweaters, sweaters, but not swimsuits, so forth of exclusions. For your custom combinations. That gets more complex than most people want to get into, because usually the amount of effort to control remarketing that closely, is not worth the outcome. So in that case, you might say if they went to a product category, we're going to bid X amount. But they made it to our shopping cart, or maybe a form page. They're more valuable. They not only looked at our product, they added it to their cart. And they even went to checkout. So but they didn't finish checking out. so in that case that's a different list because we're willing to pay in more for someone who got that close almost converted but not quite. It did not quite give us money but they were ready too. Maybe they got distracted there was an issue where they didn't see a shipping fee or they weren't, they weren't a short term policy or something else. So in cases where someone got to a forum or shopping cart and they didn't convert, those are usually more valuable than someone who looked at a category page. So then, when we looked at our ecommerce example we have our categories. Then we have our shopping cart abandoners, and finally, we also have people who converted. Now, often converters are negative lists. They shopped, they bought. They're done, they can be a negative list. And this is useful for things like contact forms or what paper downloads will a user really has their information and they're just not going to convert a second time. There's not a reason to bring them back. However when you're focused on things like lifetime values, remarketing to converters can be really powerful. Now, you can do this in a few different ways. If you have a product that people buy multiple times per year, so you have a daughter and your daughter's under two years of age, you buy diapers once a month. And so you may instead set your remarketing to after every three weeks, the list goes active again and therefore, they start seeing your ads when their subscriptions are running out. You might have times where your in flowers. You look at the flower industry, in the Spring there's two main days. You have Valentine's day and Mother's day. So if someone bought flowers for Valentine's day and they had a good experience then you want to remarket back to that same converter when it's coming up to Mother's day. Because they're more likely to buy flowers from you if they had that good experience. Other times it was an ecommerce purchase and you just want to remind them, hey, you want to buy more clothes cause clothes wear out. Let's show them ads again, maybe a month after purchase or some timeframe based upon our repeat purchase behavior. So remarketing converters can be really powerful but often you don't want remarket back to them the next day, or even that same day cause they're not ready yet, again depending on your industry.

5.9 Creating Audiences in AdWords

So let's say that we wanted to remarket back to converters, we wanted a time lag on it. So what we could do is we make one list for people who convert it, and we make say, the membership duration 30 days. That's how long the cookie last for, then we make another list, it's like, We Miss You. And this may be a 90 day cookie, so now when someone converts, we put them in both of these lists at the same time. So now we can use our custom combination, and we say, if they're in our list of We Miss You, which is 90 day cookie, then show our ad. But, if they're in the Thank You list as well, don't show the ad yet, now the Thank You list is only a 30 day cookie. So now what happens is from day 0 to day 30, they're in the Thank You group so they won't see your ad, on day 31 the Thank You cookie expires so now they will see your ad from day 31 to day 90 and at day 90 the We Miss You cookie expires and so they don't see your ad again. And so remarking to converters can be very powerful once you start looking at how do people do repeat purchases or maybe they don't now and you want to try to try to get them to do repeat purchases. Remarketing converters are really powerful, especially when combined with say email campaigns, which are also geared towards secondary and tertiary conversions within a time period.

5.10 The Ads & Landing Pages

Now, once you create your list and you have your list live and you've made your ad groups, you need to think about your ads. Now, with the exception of remarketing to converters, so someone who abandoned your site but did not convert, you don't want to just show them that exact same offer over and over again. Because they didn't convert the first time. So also, when you think about your offers for non-converters on remarketing, you want to test different offers to see what is going to resonate with them. And everyone's a little different, so often, with remarketing, you may keep two, three, four different ads running at the same time. And you may test 'em, get rid of your worst ones, but you may never have, like, a best ad. Because what happens if you only have one offer, the same user sees that offer if it doesn't work for them, they'll never convert. So, instead, by having a few different offers, in a re-marketing group, when they don't like offer one, and two, maybe offer three resonates. So, it's very useful to keep a few different types of ads in your remarketing groups to see what's going to resonate with the users. The same goes for your landing pages. Often remarketing sends a user back to the same page they were on when they did not convert. If that page didn't resonate the first time, rarely is it going to resonate the second, third, or fourth time. Now, there are exceptions. Someone wanted to buy, but their phone rang and they got busy and forgot about it, so forth. So things do happen, but try a few different landing pages that are correlated with your different ad offers. So, in the case of like market motive. If you're interested in paid search training, you may have one offer that's visit one of our workshops, see our information live. A second one, take a free preview of some of our videos. A third one, the main page, or page search. You can make another one that's white papers and so forth. So, don't just try the same thing over and over again. Try some different types of offers to see what does work. Another mistake with offers is making the very first offer someone sees a discount. The reason why is shoppers are becoming very savvy online. And there are groups of people, in fact there's even Facebook groups for this, where they keep lists of websites where you should go to the website and then just abandon it because you'll get a percentage off. Or you should go to this website, add things to your shopping cart and then leave. Because then you'll get an offer for 10% percent off and you didn't have to search for coupon codes. And a lot of poor remarketing is teaching shoppers to abandon websites to get a better offer. So this is not a good behavior to train users to do. Abandon your website. So Often your first offer should just be more about the product, different benefits so forth, and then you might have one that you show to a user later, not immediately which is a discount. So just be careful of always offering discounts immediately. This is common with sites with email as well, where if someone imputed their email address into our shopping cart, but they didn't convert, and its been 12 hours, and the first thing we do is we send them an email saying come back and we'll give you 10% off. There are lists like this that people keep of sites that you should enter your email address but then don't convert. So just be careful with always offering discounts first, try some other benefits to see how that works.

5.11 Frequency Cap

Now for re-marketing, one of the issues is most people who were involved during marketing are not display buyers. They're not used to buying CPM. And, and then what happens then is some people feel stalked online because the ads aren't frequency capped. So a user may see your ad five, ten, 20 times in a single day. So you want to frequency cap your ads. A frequency cap is how often a user can see your ad in a time frame. This is a campaign level setting within AdWords, so you can go to your campaign setting if you only have re-marketing in that campaign, then it only affects your marketing campaigns. And instead of cap of how many impressions a user sees in a time period per day, per week, per month. And then you can make it per ad group or per the whole campaign. And so now the user won't be bombarded with your ads. They'll be reminded but they won't feel like they're constantly seeing your brand everywhere, which can also cause negative connotations to be associated with the site.

5.12 Case Study for Segmentation

What really comes down in marketing is making sure you have good segmentation, and then ads running based upon the different user segmentation. So this was a company who they start re-marketing and they just had one list. It was all site visitors. And they ran it, so well marketing is terrible. I mean, across display they were getting 2.89% conversion rates, but their marketing was 0.37% conversion rate. Much lower. And their CTRs were in marketing were a abysmally low, and it's because they just had that one list that was all site visitors. I mean if you have 1.8 million impressions, and you only have 282 clicks, you haven't done a good job of either the segmentation or the ads. So then they went and they added just some negative lists of people who converted or didn't spend much time on the site. That alone caused their conversions to go from a really low percentage, 2.65%, still not as good as the rest of their display, but much better. But one conversion to 38 conversion. So then next step was to really do true segmentation based upon where people were in the site, what they were interested in. And now their normal display network is 2.89% conversion rate, but their remarketing, which should be higher than display overall, is up to a 5.86. And that one conversion became 675, just due the good segmentation. So if you've done remarketing and it hasn't worked, you need to look back at how was it executed, because if it was executed just by all site visits or the ads were never were really tested, it was sent back to the same landing page someone didn't covert before, often you're not going to see good results. So you do need to do your segmentation, just like you would with your normal search ads; here's my ad, here's the keyword, here's the landing page, they're all correlated so that a user is seeing an ad based upon their information intent, which is their search query. And they see a landing page based upon the ad and their search query. It's the same thing remarketing, you if you show an ad for every single person, regardless what they did on your site, you're not going to do well. The more you segment, the better off your results usually become. Now, this does come in to eventually where you have to think about the possibilities versus the execution time. So for instance with Kaplan, you could segment based upon every area of study and every program within that area. This can quickly become thousands of remarketing lists. So there are times you need to look and say, all right, we can, this much segmentation makes sense but once we pass that we're making an incredible amount of work without a lot of returns. So there's that time versus possibility that definitely kicks in. But honestly if you'll catch some of these beforehand by just looking at first how long does it take someone to convert. How many pages do we get in that time frame? And can we even make this list active? For small sites, you'll have some answers that are no. For big sites, pretty much everything would be a yes. You'll get at least 100 visitors in a month, you know for every program, for say like Kaplan here. So, now you have to look at, here's the possibilities, but it does become exponentially larger as you do more and more segmentation. So there's some planning to be done about how much do we segment versus the returns. And you can do a few tests to say, if we went to this level of segmentation. What's the result, versus just as higher level segmentation, which is pretty easy for us to execute. Let's run them both. If our low level of segmentation, we went really crazy on it, is that much better, then we'll spend the few days setting this up. If it's not, then we'll leave it at the higher level of segmentation. So segmentation's important. But you always have to take into account how much time will this take to go to the full possibility of levels.

5.13 Recap

So to recap remarketing. You know, first you really need to determine are you going to use the AdWords scripts or Google Analytics as your main platform for setting cookies. That really is question one because that will determine where you're setting up all your list. You're still going to do all the execution work of the lists in AdWords. But what you're going to use to collect the cookies, the technology questions first. Then, you need to segment your website into lists, or segment into categories of types of users. Then, consider your average buying cycle time by those segments. And in some cases, you may not have a difference by segment. It just is a site. Makes it even easier. So, do the analytics work to see how many page views do we have for these segments by our average buying cycle time because this will help us determine our cookie lengths that we create when we set up each of these lists. And don't forget, you can make negative lists. Commonly, it's converters but it could people who went to a certain part of your site and that site just doesn't get many conversions. Or you might have a big website where you also have a support section for current users. Well, if they're in the support section for current users, that's a good negative list. If a membership site post-login is a good negative list because they're already members. You don't need to show them remarketing once they've logged in. So negative lists can be really useful in getting rid of either current users of your product or people who are unlikely to convert to your product. Then once you have your list, usually you want just one list per ad group. This let's you really do good ad group testing per ad group and good landing page testing for those ads per ad group. And then, often, it's useful to create a new campaign just remarketing. So you can segment out this is our remarketing traffic and then this is our display. because with display it's new customer acquisition then many times who have never had their brand interaction with you. Remarketing, someone's been on your site. They've have a brand interaction. Different type of user. So when you really think about remarketing it's about segmentation, then showing the right ad, showing the right landing page, to see what combination gets someone who did not convert previously to convert and become a customer the second or third time they visited you.

  • Disclaimer
  • PMP, PMI, PMBOK, CAPM, PgMP, PfMP, ACP, PBA, RMP, SP, and OPM3 are registered marks of the Project Management Institute, Inc.

We use cookies on this site for functional and analytical purposes. By using the site, you agree to be cookied and to our Terms of Use. Find out more

Request more information

For individuals
For business
Name*
Email*
Phone Number*
Your Message (Optional)

By proceeding, you agree to our Terms of Use and Privacy Policy

We are looking into your query.
Our consultants will get in touch with you soon.

A Simplilearn representative will get back to you in one business day.

First Name*
Last Name*
Email*
Phone Number*
Company*
Job Title*

By proceeding, you agree to our Terms of Use and Privacy Policy