So going back to that Google Analytics example, imagine that you have the ability to segment customers in any way that you would like on your electronic store or e-Store. What are the ways in which you could find customers that you feel like would be more interested in your advertisements? Fjallraven is an emerging winter clothes shop that has just recently grows into prominence in the United States. It's actively trying to grow its US market and it could use help because a lot of people in the United States don't know about Fjallraven it might be a ripe opportunity for them to use digital advertising. Let's play a fictitional example here and pretend that Fjallraven is launching a new line of coats. These coats are going to be warmer than the existing codes in there actually going to be lighter to wear. So we really feel like some folks who may not have bought our coats in the past because they were so bulky may consider this product. What a way in which we could use retargeting to reach that most optimal group of people. Well, if we just look at the group of people who looked at coats on our website but didn't actually buy our coats, that could be a good group to start with, right? This is where retargeting starts to shine. These consumers have already shown interest in coats, but for whatever reason they didn't buy it. If we launched a new coat product that could resolve a lot of the concerns related to the previous iteration of the coat. And we have a real reason to think that this group of people could be really applicable to a retargeting approach. So here's an example of a Fjallraven ad that's designed to do just that, it's got a new type of Parka and its advertising it to people that have shown interest in their Parka in the past using the retargeting approach. Now, let's take another example, consider the apparel company Vineyard Vines. Vineyard Vines has a very cutting-edge social media team does paid social they also have a team that's dedicated to doing emails. So regular updates about products and sales. What they learned is that there's a group of people that never open their emails. Or group of people that were just refused to check the content of their emails no matter what the subject heading is or no matter what's in the actual email itself. They realize that those people remained active on their actually e-Store, they continued to buy products online, they just never check the emails or the updates. So they really wanted to reach these people so that they can begin to engage them in even more type of e-commerce. So how did they do it? Well, they decided that they would use it a retargeting approach. So they embrace the technique that we've just talked about. Vineyard Vines created a series of Facebook ads that retargeted people that often bought stuff on their platform but never actually open their emails. And what was the result? The result was amazing. Those consumers were especially interested in Vineyard Vines and took them up on their special offers. It just shows you how sometimes reaching a consumer in one medium versus another can yield a completely different result. So taking a step back Vineyard Vines inserted its own special ingredient into its Facebook ad campaign. Instead they realized that the best list of people to reach was already in their hands. So they sliced up their customer email database into a segment that they really felt like would engage and they tested it, and the results were great. So the Facebook ad manager gives you a couple options once you've uploaded your custom audience. You can just simply say I only want to serve ads to the people that are in this list and really limit your reach. This is only effective when you've got a large customer email database, right? So if you only got a hundred consumers Facebook won't accept something that's small and they may let you use it for other use cases that we'll talk about later in particular lookalike audiences, but they're not going to let you just reach a hundred people. This kind of like ultra-precise micro-targeting is sticky with consumers, there's a privacy issue there, right? If you could advertise one-to-one at some point then the idea of anonymity really starts to break down and consumer start to be really upset. So you have to have a large list to really say that you only want to advertise to that list. But you also I could say I'd just like to pay more when I know that a customer is in this list. So you can actually tell Facebook hey, yes, serve it to these people that could be interested and you can dial in your targeting parameters as best you're able. But then you can also say if someone is in this list that I want to pay more to serve that ad to that person because I'm especially sure that that person could be a great fit for this ad. Or I'm especially sure that this person will engage with that ad. Overall whether you using Twitter or Facebook it's incredibly easy to get this approach set up. So let's think critically about some possible downsides to running a remarketing campaign. I think the first major downside is this idea of being introspective or looking inward at our own list. Often the customers that love us already use us to the extent that they are able to or that they would want to. So if we just look inward to our customer database, we may find that it's saturated and that it won't actually drive results to an extreme extent. Because the people that are already are buying and we don't really need to reach them again. In this way retargeting can actually take money from one place and put it in another place. The second major downside is cost. Custom audiences tend to cost money, and what I mean by that is when we leverage custom audiences and we tell a platform that we'd only like to advertise to those people, it generally costs more, why? Think about it this way, if we create an ad with targeting parameters and Facebook says that the estimated reach is 2,500 to 5,000. Well, we have that group of people and Facebook can really serve that ad at its leisure to the people that are most convenient. However, once we start to define our target audience down to lists of people that we know, it's a lot harder to find them on the platform, right? Even if we have a list of 5,000 people only 2 to 3,000 are reasonably expected to use Facebook at all regularly. And it'll be hard to find those people inside of all the other parameters that we might divide for our target audience. So it's actually harder to find people, and if it's harder to find people, Facebook is going to charge you more to do it. And it's just a matter of supply and demand, there's not much supply there so Facebook is going to have to charge more to do all that. Finally, I think the biggest and most worrisome complaint when we think of retargeting is this idea of burnout or annoyance. If you ever seen the Amazon ad that follows you around the web. Let's say you bought a blender and then almost bought a second blender. How long does that second blender follow you around until Amazon gives up on this cold case? This is really a good example of retargeting gone awry, consumers will become annoyed with you if they see an ad every time they visit your website. Just think about it like this, they'll begin to be creeped out. So in these ways we want to make sure that our retargeting campaigns are of annoying. How can we do this? By limiting the number of average frequency that the ad is being served, right? If we were retargeting a group of people, we don't want the frequency to be so high that people are just inundated with messages every time they visit our website. If for every one visit, they see 10 of our ads they'll be sure to be upset. So when you're using retargeting make sure that your frequency is doubt in a way as to not annoy consumers. Remarketing like every new digital advertising concept is exciting and can possibly generate more return for your spend. However, like many new bells and whistles in the digital advertising world, we want to make sure that this actually works before we go ahead and throw all of our money on it. So how do we evaluate this to make sure that this tactic is actually working for you? The number one way to do is to really just create campaigns with A/B Testing. So what I mean by that is create a campaign with traditional targeting parameters is best as you're able. Remember, that whole process is iterative where you're going to constantly try new parameters to see if they matter. But once you've gotten a good set of parameters locked in then create another campaign where the second audience is essentially your lookalike audience. So you're making sure that for these ads that they're either only being served to your lookalike audience or that you're blending it to an extent so that you could see a difference. And really at that point it's compare campaign A to campaign B across your various KPIs to see which it is working better. You have to ask yourself is my custom audience significantly better for the metrics that I care about? If all you care about is cost per action, then assess cost per action for the to. Remember, that you may have to run this test a few times. That is you may want to run it for a week, or you may want to run it for multiple days across the month to make sure that the one-time result isn't just due to happenstance. At the end of the day you need to be critical, if your KPIs are going up, yhen you shouldn't be using retargeting. There's going to be an additional cost associated with retargeting and you're just hoping that your performance goes up to an extent that that's outweighed.