The following is a transcript from a Google Partners Hangout. All of the information contained here belongs to Google and is to be used for educational and research purposes only.

 

Attribution: Strategies For A Cross-Channel, Cross-Device World – AdWords Hangout Transcript

 

Welcome to today’s Google Partner Hangout on air. My name is Bill Key, and I’m a product manager here at Google for attribution. Today’s actually the last hangout on air for Google Partners in 2015. So, therefore, it must be the most important one that gets all the credit. Couldn’t resist making a last click attribution joke as a part of this presentation (for those of you that are familiar with attribution.

 

So, any case, a little more seriously attribution topic that’s really really central to the digital advertising industry, and how we actually serve advertisers is ultimately what it’s about as helping advertisers understand what they’re getting in return for their investment in digital advertising. It’s been at the center of Google’s measurement strategy for a long time, but as the world has become more complex. With multiple devices and increasing numbers of channels, this has become a bigger problem.

 

So, if we think to what the consumer experiences, really (like today it’s really a set of moments that consumers go through on their path to purchase), we think to our own experience these moments are taking place across multiple devices. We could start out researching on a phone, get reminded of a product that we’re interested in, while watching a YouTube video, and then ultimately go on to complete a purchase on a desktop. And, we can expect both the television advertising and lots of other forms of messaging from advertisers who are trying to move us along that path to purchase.

 

So, for advertisers to measure this properly, the challenge has really become large. And, the way to think about it is that we really need to value when all the moments along that customer journey and most of the digital advertising has been built off the idea that we can track the last click or the last impression that happens prior to conversion and that tells us what’s performing. Clearly that doesn’t work in today’s world. So, we talk to day a little bit about how the industry is changing around this, and how Google solutions can help.

 

So, If we look at the industry, we did a couple of surveys in the last few years. So, it basically asked advertiser what type of attribution they are using. And, if we went back to 2012, we saw about 15% of advertisers said they were using attribution beyond last click attribution. So, a relatively small number. When I started working on this area years ago, I would go and talk to advertisers, and frankly, the concept of something beyond last click attribution is still relatively new.

 

If we fast-forward to 2015 multi-touch attribution, and the ability to look at the full customer journey, it has definitely become more mainstream. And, when we ran the same survey, we actually found that about 35 percent of advertisers are saying that they were using some form of non-last-click attribution. Now, that’s hardly a hundred percent. So, I think we’re still in the early days of this adoption, and there are are a lot of challenges that remain. But, clearly the move beyond last-click attribution is a major force within digital advertising. And, I think it’s something that everyone is going to have to find a way to work with to make the most out of their business going forward.

 

So, for the advertisers that have made the switch, we’re actually seeing some great results. This is just a few snapshot examples of the type of results that we can actually see so in the case of ideeli, they were able to see CPA’s reduced by about 14%. On the Beach, a travel company in the United Kingdom, actually saw, their ROI grow by about 25%. You can see some of the other examples that are there. But, the point is, actually switching to these new forms of measurement, advertisers are actually able to see better results. And, therefore, that what we think will drive industry going forward.

 

Attribution Solutions from Google

How can we actually help this through attribution solutions that are delivered by Google actually has a range of attribution capabilities. And, I’m gonna talk through what those different capabilities are as well as some of our roadmap for where those solutions are going so that you can plan 2016 around providing the best solutions for advertisers around attribution. So, there’s really three major buckets that we look at right now. So, one is a set of attribution capabilities within AdWords that helps you understand for Google search how the path to conversion looks before the last touch.

Attribution at Google Spans a Set of Solutions to Address These Challenges

AdWords Attribution – Google Analytics – Adometry

So, this is actually something that we’ve had for several years now was originally branded as search funnels. It’s been rebranded as advertised attribution, and there’s actually an exciting roadmap of capabilities and additions that have been added both in 2015, as well as are coming in 2016. This is a solution that simply turn key leverages AdWords conversion tracking and helps you understand how Google search is working on the last-click.

 

The next solution is Google Analytics, and this is really our multi-channel attribution capability that allows you to track not just how Google media (in the case of google Search), is performing beyond last click. But, also how the multiple touch points across multiple channels (whether it be organic, paid, email and others) are actually contributing to the path to purchase.

 

So, I’ll talk a little bit about what you can do with Google Analytics today, and what’s coming. Then, the last solution is Adometry which Google acquired about a year and a half ago. Adometry really provides a full enterprise-level solution for very large advertisers focused on holistically looking at all media across all their digital channels (whether it’s Google with many other partners providing integration, cost data, and additional modeling capabilities) to really serve the very custom in-demand needs of the largest marketing organizations in the world. And, so I’ll be talking a little about Adometry, and how that works in the border suite of Google attribution capabilities. These are the three solutions.

 

However, Significant Challenges Remain

Data Collection & Quality – Connecting Moments Across Devices – Bridging the Worlds of Online and Offline

I now want to transition to talk a little bit about the core challenges that we’re trying to address, and you will need to be able to address with advertisers going forward with attribution. So, there’s really three major ones that I see on the horizon right now that pose barriers to really being able to take advantage of attribution. We’ll discuss the first, then I’ll spend some time on data collection and quality. And, for those of you who engage with clients in deep analytics and data analysis, you’ll understand that this really is paramount.

 

It really goes back to the idea: when it goes in as garbage, it comes out as garbage. So, understanding what’s needed in terms of data and data quality to do well is pretty central to having a successful attribution strategy. Let’s talk a little bit about how you can think of data quality and data completeness and collection throughout the set of products that we have, and think about it as a framework for your overall approach.

 

The second is how we think about multiple devices and moving across devices. This is a major shift in the industry, and there’s a lot of interesting capabilities and work being done here and approaches that you can take to help clients better understand a multi-device world.

 

Then, finally, there’s a bridging the world’s online – offline and this goes in a few different directions. Understanding how online media is affecting offline transactions and conversions (as well as the reverse – how offline media actually drives online).

Data Collection & Quality

Quality Data Collection is the #1 Barrier to Adopting Attribution

 

• Cost data – Aggragate CPM’s and CPC’s

• Context data – Campaign & Creative – Keyword – Channel – Revenue

• Event data – Clicks -> Impressions -> Directs -> $

 

 

So, let’s dive into the first one, which is data collection and quality. When I look at the framework for thinking about what an attribution solution needs, there are really three different types of data that we look at. So, we’ll start from the bottom of this list. The first is event data. At the end of the day, attribution is really about two things: 1. the conversion events (whatever the advertiser values as their business objective) and 2. then the marketing events that potentially drove those conversions. About all of those things are discrete events in time that you can collect through a variety of different data collection capabilities.

 

For example, you can collect the conversion using Google Analytics. We can collect direct visits to the website. Impressions can be tracked through things like the Google Display Network or the Double-Click products. Then clicks can come from a variety of other sources as well. But, at the end, you want to be able to assemble all of that in one place, so you can get as complete a picture of all the events that happened on the way to conversion. So, that’s the first player but it doesn’t stop there.

 

One of the next challenges; you need all of the contexts for those events. So, for instance, if you have a click to the website that came from an affiliate or comparison shopping engine, you need a lot of information to understand the context of that click. So, for instance, you want to understand the campaign that was driving that, or whether there was a particular creative associated with it. In the case of search you need to know things, like the keyword and the channel involved.

 

In terms of things like conversions you actually want to understand information like revenue margin potential and things like projected lifetime value. So, even though you have the event data at the bottom layer, you actually need to add on top of that a variety of contexts data. I’ll talk a little bit about how you can approach that with the different products we have today, but this can be a complex area. When you’re working with an advertiser, it helps to get an inventory and understanding of all the different places this context data can be coming.

 

Then, the last piece, which was related to context data, is cost data. Ultimately, what we’re after here is helping advertisers understand ROI. So we might get a lot of the ROI in that equation from the bottom parts of this, but we need the investment side. You need to understand what was being spent. While that sounds simple because an advertiser might know what they spent. It can actually be complicated to get that from the wide variety of different sources that an advertiser might use for advertising. So, we’ll talk a little bit about how cost data fits into the picture as well. But, this is really a framework to think about. Even if you weren’t thinking about a particular product, if you were just thinking almost from first principles (what data do you need in order to do actual attribution), this is sort of the core set of data.

 

Tackle Data Collection At Different Levels Depending On Your Needs

So, let’s talk about those in the context of the three products that I just mentioned before. Let’s start with event data in AdWords, so building up from the bottom in terms of the clicks and conversions, this is something that you get directly from a bird’s eye using AdWords conversion tracking. Same thing for the context data there (in the sense that AdWords already has all of the different variables when it comes to even things like display advertising).  It has all of the targeting contexts that you need to understand what that media finally cost also built into that because it’s actually transacted through outwards.

 

In some ways, when you look at the AdWords solution, whiles it’s narrowest in terms of it only covering AdWords media. It’s also the most turn-key and simple. You don’t have to manage data sources from many different places in order to get that picture. So, that’s one way of thinking about the AdWords solution is that this is the simplest place to start.

 

We then move on to Google Analytics which, as I mentioned before, begins to allow you to get that cross-channel view of things which begins to actually make the collection process a little bit more complex. So, there are multiple ways of collecting the events that matter in Google Analytics attribution. There’ are visit and conversion tracking. This is when you have Google Analytics tags deployed on your website. You can track when somebody visits the website, where they came from, and you can set up conversions to track things like goals they achieve on your site as well as e-commerce transactions (in addition to doing that through the tagging).

 

We also have a solution in google Analytics called the measurement protocol that allows you to send in events from other sources. We actually have some advertisers, and I’ll talk about one in a moment, who sent in conversion data from a point of sales systems where they actually sent in event information from their in-store touch points.

 

Finally, everything that we talked about up until now is primarily on site, or something that happens within kind of an advertiser’s property or control their store. There’s a third category, which is off-site impressions. For things like display ads that happen on publisher sites, we have integrations and Google Analytics that allow you to pull in that data, as well as with both the Google Display Network and our Double-Click products.

 

So, that’s how you can think about the event level coming into Google Analytics to get those clicks, conversions, and impressions. The next area’s context data is AdWords and Double-Click data. We can actually bring a lot of the context data, and automatically through first-class integrations that we have those platforms. However, if you’re tracking media and clicks from non-Google media outside AdWords or Double-Click, we also provide the ability to either tag the campaign tags or upload match tables. To help you understand the context of those clips, are things like campaigning. So, there are multiple capabilities there. Then, finally cost. There’s a similar story where we have the cost automatically from AdWords, and Double-Click. You can use our upload API’s to bring in additional context.

 

The last piece to cover here with it’s ability is the diameter capability. We start with the site and add tags or we can track both on-site and off-site activity. And, we have event data import capabilities similar to the measurement protocol. Some of the additional flexibility here is that we can support things like late events that come in. So you might not know about an event you or have data for a conversion until, for instance, a week after the fact. You can actually send that and Adometry later on. This makes it a bit more flexible., but the complex enterprise use cases in terms of contacts data, their direct integration with multiple platforms.

 

So, we can bring in costs and context data from multiple demand-side platforms, multiple remarketing and display providers, and similarly, with Kosta, there’s the ability to bring in and reconcile costs for multiple platforms. And here, though part of the layer is providing a layer of additional services from Google to both help manage that data, and build custom integrations were needed in order to make that work for the enterprise. So, this is really taking that framework in the context and cost data, and mapping it into the solutions that we have at Google today in terms of what you can take advantage of.

 

Connecting Moments Across Devices

Current Approaches Fail to Offer the Full Picture

I’m going to move on now to the next topic, which is connecting moments across devices. How do we deal with cross-device? So as many of you probably know, today’s predominant measurement techniques are essentially based on anonymously identifying the user associated with a particular browser cookie (or in the case of apps and advertising, ID (like an ad ID on iOS). The challenge with this is that ultimately, the events on each of those devices, whether they be conversions or impressions, clicks end up being associated with those individual anonymous identifiers that don’t connect the customer journey together.

 

So, you might have one cookie that sees an ad on a desktop and another cookie that seen on mobile. But we can’t actually connect those together. This presents a large challenge in the wold of cross-device attribution, in terms of trying to understand how that all fits together.

Finally, a Way to Tie the Story Together Across Devices

So, a major way that we’re beginning to move this forward (and we’re rolling this out across multiple platforms) is the ability to anonymously leverage Google-signed-in-identity to enable us to understand when there are multiple touches on multiple devices. So, what this enables us to do is to understand when there’s an ad click on a mobile device and a conversion on a desktop. We can actually begin to tie that story together. This is really what’s powerful about making the shift from cookies to a more complete understanding of the user path. So, this is pretty central to marketing and digital marketing going forward. I’ll talk a little bit about how you can take advantage of some of these capabilities today.

 

Within AdWords, there’s actually (as many of you are probably familiar with) capabilities today that allow you to do this for last-click attribution. This is done so you can understand in particular when there was a click for instance on mobile and conversions on desktop – how that influence actually happens. We’re adding the capability in this coming year and the first half of 2016 to do that in a multi-touch; so that you can understand how, for instance, a mobile ad click on a mobile device actually assisted a conversion that never happened further down the conversion funnel. We’ll talk about that in a moment in a little more detail.

 

Within Google Analytics, there’s also capabilities today that allow you and an advertiser to provide their own identifier. For your advertisers that have large volumes of logged in identified users, we can actually take in that data and help you understand when they’re interacting with you on a mobile device and desktop device. You can actually stitch that together and understand, in aggregate, how users are working across those.

 

We are also working on a Google solution for this that would enable an advertiser to understand this even if they don’t have their own signed in identity. Then, the final piece is in Adometry where we actually have a beta going right now where you can essentially leverage the ability to understand users across devices for all digital channels and conversions. So, we’ve seen some great results from that so far, in terms of really understanding how mobile in all channels (beyond search, across email, display, video, and other channels) are really working to drive conversions across the board.

 

Example: Cross-Device Insights From AdWords Attribution

Let’s talk through a couple of examples of what you’ll be able to. This is a set of capabilities that will be coming very soon in AdWords that actually allow you to understand, for instance, how many conversion were not just the last click, but were assisted on a device basis. So, if we look at the row here at the top for mobile, we saw that there were 75 click conversions. But, what we actually see, is that there were three hundred and twenty click assisted conversions. That means mobile may not have been the last click on the path, but was earlier in the process. Under last click, you never would have been included for credit, but now we can see that it’s actually assisting. If you look at the ratio number that we provide. For those of you who had used our AdWords attribution or search funnels capability in the past, you will be familiar with that ratio you can see when there’s a high ratio of assist to last. This means that this channel tends to be happening earlier in the funnel. In this case, we can actually see that mobile, in this particular example, much earlier in the funnel than the other devices that we show. This is a new set of analysis that’s going to be made available that will really give insight to how mobile search is assisting not just the last click, but as part of the full path.

 

Another example of what we can do with cross-devices today is in Google Analytics. This is a screenshot of the capability that’s been available for some time that allows you (when you provide your own user identity) to get a breakdown of how your users interact with you on multiple devices. So, we can see the percentage of users, for instance, that are desktop only vs mobile only  vs tablet only, and what the overlap is between them. This really helps, not just with attribution, but in terms of understanding your customer base and your user behavior as a whole.

Case Study: Cross-Device Attribution in AdWords

Key Findings

• Observed a +7% increase in overall conversions when accounting for cross-device

Best Practices

• Reallocated budget to account for additional cross-device traffic and previously unreported conversions

 

The last point that I wanted to raise is just around the results that advertisers are seeing from this. This is a case study that we did with 1 800 flowers. com, where when they began to look at cross-device attribution in AdWords, they were actually able to see an increase in the total number of conversions that AdWords was driving when they took into account those conversions that actually happened on a different device where the ad click took place. So we think that this is a set of capabilities that will expand over time, both from helping you understand how Google search is performing, all the way to understand how media’s performing across devices on a whole.

 

Bridging the World of Online and Offline

There are Two Sides to the Online/Offline Challenge

We talked about data collection and quality, and we talked about cross-device. The last piece I want to touch on is how we think about bridging the world of online and offline. So, there are two different categories of this problem that I’m going to touch on. One is how we think about how online advertising is affecting offline sales. I think we all know from our own experience (particularly as we interact in the real world with mobile devices) that our activity on a mobile device influences purchases that we make in stores. That’s a major area that advertising is affecting, but creates a major measurement problem. So, we’ll talk a little bit about that. The second area is actually understanding how offline media affects online behavior. For instance, when you’re running television ads, how does that actually affect the user behavior online. So, we’ll talk a little bit about that.

Addressing Online to Offline With Store Visits and Purchases

 

• AdWords – store visits and-store purchases, attributed to Google Search

• Google Analytics – measurement protocol support for sending offline conversions

• Adometry – in-store purchases via partnerships; offline conversions via CRM

 

 

Let’s start with online to offline. So, in terms of what’s possible from Google today, looking at our three solutions (AdWords – Analytics – Adometry). In AdWords today, there’s a set of capabilities that you may be familiar with, that allow advertisers with other physical store presence or in-store purchase data to attribute that data to Google search. This is a powerful set of capabilities that allow you to really begin to get a sense of whether or not your ads are driving in-store behavior. This helps on two fronts: 1) it helps to understand what the overall impact of search is, beyond just online conversions. It also helps you to operationalize and actually  make optimization decisions about what the most effective search strategies actually are to drive that overall business value. These are available now and are a really powerful set of tools. Within Google Analytics, we have to ability to use our measurement protocol to send offline conversions into Google Analytics.

Case Study: Galeries Lafayett Measures 20% of In-Store Purchases Preceded by Online Visit

• At least 20% of store conversions are preceded by an online visit
• 20% of these visits take place online from a smartphone
• 50% of mobile visits take place on the day of purchase offline
• When in-store conversions are included, digital ROI increases 2.4x

 

I’m going to talk in a moment about a case study for a large advertiser who did precisely that, and was really able to get powerful insights into how ultimately their online advertising and presence was affecting their in-store sales. Then finally, with Adometry, we offer the ability to take in-store purchases, and link them up by partnerships into the online conversion path We also support the ability to bring in offline conversions from a customer CRM system. With both of these capabilities not only do you get a picture into how Google Search is driving offline purchases it really provides a full multichannel point of view that helps you to understand how search display, video, and organic channels are working together to drive not just online purchases but also in store as well. This means our capabilities that are available today so let me give one example here which is leveraging the capabilities that I spoke about in Google Analytics. So, this is Galleries Lafayette. Their a large department store in France. What they were able to measure, essentially by taking point of sale data, and linking it back up to their Google Analytics data was that about 20% of their in-store purchases were preceded by an online visit. Just to be clear, this is  just what they could track, not taking into account the fact that there were some users who had an online visit. But, they were unable to connect it to the in-store purchase, because they don’t use a loyalty program (or another way of staying connected to Galleries Lafayette. Even based on that sort of example of users that we saw, there was an enormous percentage of the purchases that actually were preceded by a trackable online event. Twenty percent of those visits took place online from a smartphone. So, a significant amount of activity is actually happening on mobile, and fifty percent of mobile visits are happy being on the day of the purchase.

 

This means that this is not that proximity in time actually helps to show how connected it is to the purchase process, and where the business value came in. When we include the in-store conversions, you’re actually able to see digital ROI. ROI that they were getting from their digital campaigns included posting online purchases and offline trackable purchases increased by 2.4 acts. This completely changes the game as far as understanding the effectiveness of that online advertising as well as the strategic and tactical decisions that you would make around where to invest. So, this is really a great example of how when you can begin to tie that story together it changes the way you think about advertising within the context of the larger business that has both online and offline presence.

 

There are Two Sides to the Online/Offline Challenge

Let me now talk about offline to online. So, many users (and you probably have had this experience yourself sitting in front of the television using a device like a phone or tablet) with the proliferation of those devices. It’s really become a common way of interacting that really didn’t exist even a few years ago (to the extent that it does today). So when we look at the relationship in terms of consumer behavior between TV and online, the signal has really become a lot stronger.

TV Attribution Connects Offline Media to Online Activity

One way of thinking about this and this is how we actually do about this in a product capability we have is to look at how TV is affecting the minute-by-minute behavior of users both on their website and in terms of their search behavior. We look at this chart, the blue line that’s here is essentially the observed data for searches for particular financial services brand. So you know you can see that during the beginning of the day, there was a certain amount of search activity and then the blue line is observed activity. There are obvious spikes on most spikes coincide with their TV ads. TV is clearly driving a very strong start response here. What we can do as a part of the product, is actually build out a baseline model so that baseline orange line predicts what the traffic would have been in the absence of those TV ads. There’s some sophistication to this. We don’t just average the two minutes before the spike, and then project out. We build things like a machine learning capability to learn how to expect when an ad of a certain magnitude (like an ad that’s shown to a million people versus an ad that’s shown to a hundred thousand people, what kind of response we expect to see. What that allows us to do is get a really clean ead on what that incremental bump is in from the TV. The great thing about this, if you’re a TV advertiser is that it gives you access to granular level with much more real-time insights then you’ve ever been able to get around TV before. Very often, in terms of TV measurement, you get extremely high level aggregate views of how many people you are reaching, and perhaps a view of how TV,(as a whole channel) is driving results. With this set of capabilities, if you look at the report screenshot that I have here, I can actually begin to drill into detail (like the Networks that the TV ad is running on). Here I can look at the number of impressions that were served as a part of those TV spots,and the costs associated with them. Then I can actually look at the response. You might really see that you begin to get into a form of ROI performance calculation where I can say how did these networks perform relative to each other.

 

In a somewhat extreme example, if we look at the first two rows on this chart, we can see the largest investment I had isn’t CNN. The largest number of impressions happened during this time, and in terms of the cost per attributed click, which is way over on the right there, $6.34 (which is actually significantly less than my average overall average of $10). That’s telling me that it was both large volume and was performing pretty well. ABC network, however, when we look here in this particular case is the second largest investment (10 million impressions). When I look across, it had very, very little search response. In fact, it has me over $1,000 per click (according to the small which suggests that there might be something going on with that. You can drill, and I don’t have the information right here, but you can drill. Then you can understand things like: what were the creatives being shown, what programs were there being shown against, what time of day they were running. You can actually unearth why that strategy may or may not be working, and then make investment decisions to either change the way they are advertising on ABC, or reinvest that money elsewhere.

 

So, we found this to be a really powerful tool that helps advertisers begin to make better, faster, and more optimized decisions about television, and has the benefit of really connecting with the online strategy. This is because you can understand what type of online response is being driven by her ads, what type of searches it is driving for website visits, what is the quality of those visits. This really becomes a powerful tool in connecting those two worlds together. It is really driven ultimately by the consumer behavior of users sitting in front of their TV, and responding immediately tabs. So, just as a highlight, you can actually now leverage this TV attribution product which is sold as a part of Adometry. You can engage with us on that. You can use it completely separately from the rest of Adometry if you wish. You can use either Google Analytics site data to understand how it’s driving site visits or you can actually just leverage Google search data to understand how it’s driving queries on Google search for terms related to your brand and category. One of the nice things about this it’s extremely low friction setup. you can essentially, just using the Google search data, we actually have all the data that we need to deliver their insights immediately to a client. So, this is really becoming a powerful tool.

3 Major Attribution Challenges to Address

Those were the major areas that I want to walk through today. Obviously attribution is a broad and complex space. We talked about data collection and quality we talked about approaches to thinking about cross device, and then finally went through sort of the two dimensions of online and offline. So, with that I’m gonna switch to answer some of the questions that have been coming in so that we can talk about, either these topics or any of the other areas of interest around attribution. I’ll switch here quickly to take a look at the questions here. If you have an entertainment news website or YouTube channel, and get the trailer or interview from a studio; can I use Google ads in the video if I don’t own the clips? Is it turned into a news type story with a review that I can put ads in? This is actually a question I think we will follow up with later, because I think that it corresponds, not necessarily exactly two attribution, but the overall set of using advertising on YouTube. Another question here: If a firm doesn’t have 1-k money to spend on attribution technology, like a small business, Is there really no solid means of multichannel attribution tracking? That’s a good question, in terms of like how you think about the spectrum of offering there. Today the main offering around that is to leverage the capabilities that are in Google Analytics, and there’s enhancements that are going to be coming to those set of capabilities in 2016.

 

You know the way that we look at it is that there’s some dramatically more sophisticated needs that come from the largest advertisers for whom providing a paid fully serviced solution is really the right answer. But, there’s many advertisers who absolutely need multi channel attribution, and we do provide and will be providing multi-touch attribution capabilities. When will the cross device insights be available in AdWords? As I mentioned, the last click cross-device capabilities are available today. However, the multi-touch attribution capabilities are going to be rolling out in the first half of 2016, and they’ll really be focused on understanding Google search at first. So, the next question here is: how does the measurement protocol know that a customer went to the website and then made a purchase? This is great question, and I think it really goes to the heart of the Galeries Lafayette example that we showed before. In this case, it’s essentially leveraging advertisers own identity for that user. So, advertisers that really want to provide a customer experience that encourages and provides users value, based on providing either a signin or part of a signup as part of their purchase process online. And, then have, for instance, a loyalty program for in-store purchases. That is essentially what allows those things to be tied together. So, you know on a technical level, Google Analytics, when you see when you are sending any information about a website visit from our tags, you’re using our User ID feature. You would be able to send in the user ID for that user. It’s anonymized, so it’s not using personally identifiable information, but it’s an anonymized identifier. Then, when that same user is in-store making a purchase and you have that same ID, you can use the measurement protocol to send that back in. It definitely relies on that, and you know completely acknowledged that that may not be the case for all retail customers. The alternative to that is really the solution that I was referring to Adometry provides today which would allow essentially through partnerships a set of capabilities to take that offline data that you might not have been identified for an onboard and there’s a lot of information on that. So, another question here around TV attribution: can that be a computer model be used on any other source of traffic. Today we can support who will search and we support Google Analytics visits. We are working on a capability that would allow you to essentially provide your own time series of data. Likely, the main thing would be uploads of site visit information from other types of tracking systems. If that’s something that’s of interest to you, we can definitely follow up and talk about what capability would be coming there. But, essentially you know any time series of data of sufficient granularity, could eventually be supported. So, the next couple of questions here are basically similar which is like “how is TV attribution packaged?” So TV attribution is essentially a standalone product that’s essentially packaged as a part of a Dollar Tree. So, you can use Adometry’s digital attribution capability, it’s TV attribution capability and its marketing mix modeling capability. You can either use them all together, and you can pick and choose the ones that you want and use them as a stand alone. Right? So, with TV attribution, you can actually use that capability, and if the data that you have to feed into it is Google Analytics, you can use that. If the data you have to add into it is Google search, you can use that. So, the way that the TV attribution pieces offering. So, just two more good questions here. I think they’re both related to POS point of sale collection. One of them was effectively what I already answered before in terms of either using your own user identifier or leveraging essentially the partnership driven solutions that we have. The last one here I’d love to show in-store purchases for your hardware store client from online/ offline. “You mentioned something about a rewards program as a way of tracking in your case study. Do you have suggestions on how to track that effectively?” That’s a great question. So, effectively that work in the current Google Analytics setup you would need to have both those users have an incentive to essentially login on that advertiser’s website on the hardware store website. Then, when they have a loyalty program to have some kind of identifier’s that’s related to that. You know very common that it’s something like email address, and then eventually what you would then have of saying, this same user who was on our website, was the same user who was in the store. Then, when you send that information back into Google Analytics, we used the anonymized identifier as a way to join those two events together. As I said earlier, acknowledging this can be a major effort in terms of not just the data collection, but also in terms of really having overall customer experience that is focused on that kind of multichannel interaction with customers. In the sense that the customers really interact with you deeply on your web or mobile app properties (as well as in your store. I think you know in some ways, the most successful most cutting-edge retailers are really heading in that direction, and they understand the value of forming those relationships with those customers across all those touch points. I don’t mean to make it sound easy, but at the same time, with the right data and the right strategy, it’s definitely possible. We’re seeing large retailers as they are able to put these programs together, we’ll begin to get a better understanding of how digital and offline are working together. I think that wraps up the Q&A portion as well as the presentation. I want to thank you for joining us for the last hanout of the year, and I wish you all a very happy and successful 2016. Thank you.