[MUSIC] Hi everyone, this is Vishal Sachdev here and we have Beth and Victoria from Grainger with us for taking some questions from you. We have some questions on the forums that you've already put up there and I've shared that with both of them. So what we'll do is we'll just chat a little bit about one interesting article we saw, rather I saw shared on Twitter by somebody about the trends in display versus search advertising. And then have Beth and Victoria share any of the trends they see in sort of their field of expertise. And then we'll move on to some of the questions, and I see some questions already popping up on the Q&A screen as well, on the Hangout, so feel free to use that as well as we go through the Hangout. So first off, Beth and Victoria welcome, and thank you for taking the time to chat with our students. Some of these students have, almost are close to finishing and we have a few, about ten, who have already finished, so we are seeing a lot of diversity in the sort of the campaign plans they're giving. So hopefully some of those give you interesting ideas as well. So over to you both, I wanted to just get your perspectives on this article that I shared with you which talks about the fact that display, sort of the spending on display seems to be on target. I think the time frame they had in the article is about five years to actually overtake search advertising. What are your thoughts on those kinds of numbers? >> Yeah, well good morning, Vishal, and to the rest of you in our listening audience, I feel like I'm WMAQ. This is Victoria last time was my first time doing a Google Hangout, So I decided to make it a little more visually interesting for you this time, so we have the Chicago skyline in our background, and a more attractive member of our digital marketing team- >> Thank you. >> With us this morning. But Vishal, the point that you raised actually was part of the reason why I made a change in my own career, moving from traditional media into digital, and specifically into display. And I believe e-marketer, they're quoting a similar article that I saw this week actually says that display advertising is poised to outpass search in 2016, in terms of total spending. So even sooner than maybe they predicted. But the reason being is because display has evolved far from the transactional retargeting bottom of the funnel that it originated to be, to really be a much more targeted and efficient way to reach mass audiences with branding messages. So we've really seen an explosion of video, and I think the marketplace is kind of poised and ready for, ultimately, what we hope to see soon, which is to have programmatic TV in more digital media available in programmatic marketplace. So it's really become a replacement for a lot of traditional TV, and I think we're really seeing a lot of brand marketers that are waking up to that. Thought leaders like the Procter & Gambles of this world that are really starting to put a lot of money into programmatic marketing. >> I always forget to un-mute myself, I know another sort of convergence we've seen happening with traditional video programming moving to the Internet. And I am one of the people that sort of cut the cord and moved all my viewing through the Internet. How does that, and I don't have a sense of what extent ad buying on that medium is programmatic. Could you give some insights into that? >> Yeah, well the really big media companies like Comcast and NBC, for instance, are recognizing that they need to embrace programmatic marketplace before they're forced to. And so they're starting to do larger, what we call private marketplace deals, PMDs, where they do upfront deals much the way that we would have with broadcast or network TV back in the day. You're now doing that with programmatic video. So advertisers are able to go in and they're able to cherrypick the big tentpoles of TV viewing for instance, the Academy Awards, Super Bowls, not Super Bowl but a lot of the big event TVs that are events that do very high ratings. They're able to purchase upfront now as part of a private marketplace deal. And soon more inventory will be available that way. I can't tell you exactly what percent of the inventory is traded that way because a lot of that is proprietary. A lot of those deals you're trying to keep quiet right now. But we see the industry quickly moving in that direction. >> Always good to be looking out at trends that are going to shape the market and I think programmatic video buying on these channels would probably make us retool on the skills that are required to effectively compete. But I'm guessing the platforms and the mechanisms might be similar, so most of the professionals might be able to extend their expertise in programmatic buying on the Internet to buying in traditional video content platforms as well. >> Yes that's true. There is, aside from the strategy and knowing what audiences you want to reach at what time and what requirements for the right price, right. Right people, right place, right times. Same concepts apply for the digital environment as they do for and traditional media, but what I've learned in my own career transition is there's this whole other thing called data, which traditional projects- >> [LAUGH] >> Marketers and buyers wouldn't be as well versed in. And that's the ability to capture all those signals. It's all first party, second and third party data, that is the whole part of the audience definition that differs so dramatically from what we were used to doing. So we used to buy demographics, we used to buy the type of content that we would thought would attract a certain type of audience. And now we're able to find the audience wherever they are. We can also make it context relevant. We can marketing and private market based deals, as I said. But we're truly starting with the audience [INAUDIBLE]. But an understanding of that data and how it's collected, and what data can be transferred to what platforms is really important. And that's where Victoria, with her background, she has a degree in mathematics, [INAUDIBLE] has a penchant for data, it's where her background has really be an asset to the team. >> So Victoria? >> [CROSSTALK], not to put you on the hot seat here. So yeah. >> No, it's fine. I actually think that kind of the key underscore there is, getting the data to talk to all of the platforms, right? That's really Where kind of like a lot of issues come in, because we're not yeah, we have IT talent here new in marketing, but a lot of marketers are not usually technology first. So that's definitely been interesting to watch. Kind of the industry acquire more statisticians, acquire more IT people to be able to build out all that technology, to adopt all of that marketing, pairing all of the marketing ideology and kind of see them come together. >> I think you had mentioned something with reference to the data integration between platforms. Then you'll share the data or sort of the simulated data about your campaign performance for the Bosch campaign. As perhaps it's a good time to move on to that data as well. So that was really useful for generating a lot of discussion among learners and I thought it would be good to have you just walk us through the data. And clarify the segments that you're tracking and how we differentiate between the segments that are there in the data file. And there are a few questions pertaining to that and a couple that are coming up on the chat here. So let me try and share that file on the screen and let's see if we can just go through that, and have you explain for let's say about five minutes, what that data is. >> Yeah, sure. >> So does that show up for you now? Do I need to make it a little bit bigger? >> Okay, so over to you. >> Great, so there seem to be a lot of questions about what exactly are these segments like where and how are they defined? So I just kind of want to clear up, these are not necessarily totally exclusive segments, because of the platform issues that we know that we have, right? So there's between all these various problems every time it picks up buyers it has to be able to talk to the server. If that server doesn't get that signal on time, if there's a delay, if something gets lost then whatever that pixel was sending is gone forever. It happens a lot less often with pixels, and it happens more often with the communication between the DMP and the media platforms. So that's kind of why, this is kind of an experiment in how big does the audience like a the targeting audience have to be in order for that drop off to no longer matter. So we used a combination of both site pixels and DMT acceptance. So the www.boschtools.com only segment, that is based completely on the media partner site pixels on www.boschtools.com. That was really just an all encompassing, we're not excluding anybody in here, it doesn't matter how irrelevant you may they think they are, they came to boschtools.com, so we got them covered. The other segment that seemed to be in question was the boschtools.com, but no Grainger visit. So this one was initially based on the DMP segment, but we add a technological error. And the time it would've taken to fix that error, it would have been outside of the scope of the campaign. So we kind of. Like simulated that segment. It's the same segment but instead of using DMP, we used a combination of segment pixels. So we use the site pixels that's on Bosch.com and the site pixels that's on Grainger.com. So we wanted to see how specific of a target did we leave, could we track the user's experience through these combinations of segments Because although we do have some path to convergent analysis, that's actually fairly laborious and it's easier to track check path to convergent through cascaded bodies. Hopefully that helps to clear up that definition. So I think that probably of the rest of the data should be a little bit more straightforward. The only other nuance, I guess about how this campaign is set up is we really wanted to optimize two people who were actually purchasing Bosch products. Because we use these site pixels for so very many things, they're not set up upfront to collect or to focus on certain data. So for example although we can see everybody's skews in our carts. All the items that they're ordering we don't usually segment out based on that. Like we don't usually optimize people who have bought certain things. But in this campaign we decided to because of the specific brand focus, we filtered out all of the orders that only had some Bosch item in them. So we know actually that this company drove a lot of other orders or it competitive products and also completely irrelevant products. There were people who received at least one impression from this campaign who ended up converting on grainger.com who bought batteries, for example. So we just decided to exclude those orders from this campaign because they were so much less relevant. >> And if I could just [COUGH] tag onto that, the type of targeting that we're doing here is really super targeted which is great, and it has made the campaign really effective. The ability to do that is I think, contingent on how broadly you're targeting. For instance, are we targeting campaigns with Grainger.com who sell over a million products, I don't know if it's 1.2 or 1.3 million. So to try to limit it to a specific product or a specific product category when you're dealing with that kind of type of offering, can be incredibly difficult. So with a campaign like this, it's much more focused. It's much more easy to do >> We also have more of a reason to do it, right? If it wasn't Bosch we probably wouldn't care what they were buying as long as they were purchasing from Grainger. [LAUGH] >> Right, exactly, so the partnership. >> Yeah, absolutely. >> Yes, thank you for that so it seems like that even calculating the ROI on these campaigns is not. It's sort of half-science, half-art as well, because you've gotta decide what data is not relevant to. Some would be orders that are not relevant to include as part of the ROI. So I think this requires a lot of iteration, and I guess getting into the campaigns and then Becoming familiar with the campaign on the objectives and this kind of data that you're tracking. But I think that the great part is now that Granger is doing this, as we know with these special partners, that provides much better guidance on where these campaigns are going. So I think that will also help in framing the question for our learners, and so that is kept them focus on the problem at hand. So that's been brave that we haven't had sort of a scope happening on what problem the learners need to solve. So while we're on this data set, I'll keep it on. So there's on question from Ranjit, actually two on this data set. Are you able to see that Victoria? >> I don't see the questions. I just see the data. >> If you see the Q and A button on the left panel or if you just mouse over it might pop up On the hangout screen. >> Is this the chat? >> No, it's about the fourth link from the top Q and A. >> [INAUDIBLE] >> The first one I've selected, that's the first one from Ranjit, he's from India, he's pretty active on our forums. >> Okay. So the question is that- >> Column B. >> [INAUDIBLE] the number [INAUDIBLE]. We actually, currently don't have a lot of visibility into the number of actual people in each of the segments. That's a custom poll that I'm waiting to get back. We know that there are about half the people who were in the Bosch segment actually made it to Grainger.com, and then it's likely that more people were just found on Grainger.com. I would say that the Bosch segment is probably bigger. The power tools category, segment kind of throw things off a little bit just because we had that on. So initially, we thought that it would be interesting to know who was going from Bosch to Grainger, and also who was going from Grainger to Bosch, so that's why we turned that on. But we realized that the politics category was so large, that we couldn't do that without completely skewing. So in terms of size, the power tools category visitor is going to be the largest, BoschTools.com is slightly smaller. And then about half of the people who were on Bosch came to Grainger Hope that answers the question sufficiently. >> On a question you might have already answered in your initial explanation of the data, I just want to reconfirm. This was in the email that I sent you. There's one question about what's the real difference between segment one and segment two. So you're saying does segment one, which is Bosch Tools.com mean that the user visited the web site afterwards which is not part correlated to MRO. I'm guessing website might mean, category and- >> Yeah, so, sorry. >> I mean, you have the second part as well. Does the segment two mean that the user was still a competitor site of Grainger.com after Bosch? >> Okay. So yeah, I will have that email pulled up. There's one [INAUDIBLE] website after this not correlated to MRO. Frankly, that is most likely the case. We don't have a ton of information about, there was some very broad contextual information about the site's that our ads were shown on. And so some of them are business related sites, like Business Insider, Wall Street Journal or the Practical Machinist is more popular one that we see a lot. So there are a lot of sites like that, that we do serve media on. But there are also sites like ign.com, that we serve media on because we are trying to be user focused. We want to be wherever the people are. So unfortunately, we don't have visibility into any site that we haven't served media on. So it's very possible that and this gets into question 2. It's very possible that these people went to homedepot.com or to any other competitor website. But we wouldn't know that unless we had served an impression there and those sites don't serve media impressions directly on their site typically. >> Thank you for that. I'm going to stop the screen share and I might just be able to pull up the other questions on the screen after this, let me just try that as well. So let me bring up, just hope I don’t bring up the wrong screen. There are too many windows open. So that's the right one. Okay, so that'll help everybody see the questions also that we asked. Okay, so that is, so we'll start with question 1 and 2. These are sort of not related to the data. So over to either Beth or you, Victoria. >> want to take the first one? >> Does Grainger do any social media display ads? If so, what social media websites do they use? Are these ads targeted differently channel that goes down? So Grainger does, we started a social media practice about five years ago. And we think of social much more broadly than just social media. We think about our social business, that's a little bit. Deviating from the heart of the question here, but we do social media. Certainly Facebook is one of the great, I think, websites visited today. So, we do have a very dominant presence on Facebook. We also use it as an inventory, as well as for our program display ads that we do actively. We have our own Facebook page, and we do place ads on Facebook. We've also done a little bit of LinkedIn, a lot of recruiting work on LinkedIn. We've what other social platforms, we've done a little bit of Twitter but very little, it doesn't really resonate as much with our target, which tends to be slightly older. Anything that you want to add here? For Bosch specifically however, we’re not doing social media campaigns. It’s more broadly for Grainger. >> So product affinity between Bosch power tools and other categories of products. So this is something that I actually had to throw out to our product team and I haven't heard back from them. So I don't know if there's product affinity in a very broad sense between Bosch Power Tools and other categories. But in this campaign specifically, the strongest are going to be that those are completely what made sense and it's totally intuitive. It was between power tools and the fasteners category, and fasteners are things like screws and bolts and things like that. And then. >> Does Grainger monitor visitors in real time? We do so from a, UX, user experience standpoint. So when users are coming to the site, we're looking at their browsing and buying behavior, we can engage them and look to chat, we have site personalizations, the experience on the site is going to be relevant. To the data that we have collected from them, so to get real time. That doesn't dictate what we do off site, but it does guide what we do with on site. we can move on to number four. >> Well I just wanted to add monitoring in real time. So from a display perspective. We could monitor visitors in real time, but we're not really so worried about what they do when they're on the site. We care what happens when they leave the site. So we monitor them in real, I would call it near real time, where we need to be less engaged with them in a very detailed way. So we do know when they have left in real time. >> Which is the that's one of the major signals in terms of how we optimize our retargeting campaigns in sports, because we can optimize based on delivered sites in. Asked how are those two hours, three hours passed some days, passed a week. We look at all kind of people's campaign optimization organically looking to drive purchase or re-engagement. So that's a very important segment of a plot. So thanks for taking that Victoria. Could we assume that Grainger can display ads on GCom users real time trying to get them to purchase? So we do have a team of people, our merchandising team that does exactly that. Once people are on our site, we're really not advertising to them, we're really customizing the experience to make their experience better and more relevant, to get them the information that they need. So in a sense, I guess it sounds like merchandising not so much advertising, advertising is more off-site but we're trying to attract people to the site. But yes, we’re very active in making our onsite experience relevant and hopefully leading to purchasing. And then you have a question here about BoschTools.com only. >> I think that's already been answered, yeah. We've already answered that. So I have another question on the Q&A. I'm going to select that, so we can have a look at that and answer that. >> Yeah, so on Grainger.com, Bosch products didn't appear above the cut. So I'm curious about the experience that Arlotta had. Just was it a search that she was doing or was it a product recommendation? Because all of those different tools that we have on the site, at this time they all kind of function a little bit differently. So for example, our search function uses a different algorithm than our product function. And the search functionality in particular is something that is always being updated. So it could have been that you just happened to be searching at just the wrong time. And maybe something was in the middle of an update, you unfortunately had a bad experience or it could have been that you didn't, if it was a product recommendation space that you were looking at. It's possible that you didn't engage enough with the Bosch products, so we didn't interpret you as being interested. So unfortunately, being registered on Grainger and being registered on Bosh, we aren't so tightly integrated with them that those two things would communicate to each other, right? So all that we know is that you've been to Bosh. And unfortunately, that targeting doesn't have at this time it's not a primary signal that we're taking in when we deliver the on-site experience. That's something will come around probably in the phase two of this program. >> Thank you for that, we have another question or two. So there's a clarification from Charlotte on that, highlighting that. Signed up as a new client looking for power tools. >> [INAUDIBLE] >> Okay, I'll just read that says, she says she signed up pretending to be a new client looking for power tools. Saw a banner from Milwaukee but none for Bosch and even searching for Grainger choice and she was on the site and clicked around for quite a bit. >> So I don't know how much time she spent off site after that, but that would be did she go to her Facebook page? Did she go to cnn.com? Did she go to marthastewartliving.com? It would depend on where she went and we might not have bidded on her. I don't know if she has a edu address, if that would in fact have something. >> Which to me it sounds like she's talking about her onsite experience. And so as I was trying to say before, and I apologize if this isn't clear like that's something where we're working to integrate more into the on-site experience, it's not as accurate as off-site at this time. >> Yeah, I think so and also you mentioned that's handled by the merchandising team as a separate user experience component as well. >> Correct, yep. >> So. We don't have any more questions right now. I have one, sort of getting a sense of. What is the time frame for this campaign that you're doing with Bosch, and can you give us a sense of where this partnership is going if you can? >> So this campaign in particular concluded just before Christmas. So we're working on analyzing all those results now, and kind of putting together some wrap up docs, we're going to be talking about how, the lift that we saw in Bosch sales, year over year. I would say that traditionally we had a very close partnership with Bosch, and the success of this campaign, and the willingness between the teams to work together, has definitely strengthened that partnership. >> Yeah, I would just add to that, that thanks to all the really great work that Victoria has done managing this relationship, is that this has really caught the attention of our leadership. And this is being held up as kind of a case study of what we would like to do with many more of our strategic supplier programs. So it could be a Honeywell, or GE, or many other partners that we would be working with in the future. So, some pretty, I think, groundbreaking work. We're pretty excited about it >> Great, so I know we're also asking our learners to give recommendations for email campaigns. Have you started those yet? Or do you have some sense of how that will fit into your partnership with Bosch? >> We have not started those yet, that's something that will be brought up in probably, like I alluded to, there will be a phase two of this campaign. And we would like to start using more digital channels, so that's just what will need to be strategized a little bit, like, do we have a more specific audience? Do we want to target some of the other power tools? What kinds of power tools most of this has resonated with? But you know that audience in the past hasn't done. Definitely will be included in the future, just don't know about [CROSSTALK]. >> [LAUGH] I'm sure some of our learners' recommendations might be useful when you start planning the other platforms. We have one question from Joshua, he's also a very active user in our community. Just selecting that, I wonder if you can see that on screen for you or should I read it out? >> I can see that one. So, do you have specific and what would you consider is a microconversion? So since Bosch and Grainger both have sales goals to meet, we really optimize to ROI, or return on assets So that would be our main specific KPI. Other microconversions might be, so for this campaign it was registrations on grainger.com. That's something that we're looking at as well. But in the past we've also considered a click a microconversion. Any sign that someone has actually engaged with the ad or site landers, actually has been another microconversion we've used in the past for that. That we might re-visit in the future and see how many people who received an impression actually landed on grainger.com, regardless of whether or not they bought anything. >> Thank you for that, and Joshua thank you for your questions and for your work in the community. I guess we don't have any more questions so we'll close the session here. Thank you Beth and Victoria. And I hope the learners, particularly in this data set that you gave us really helped them sort of put a concrete thought behind what they're recommending. So thank you for that and we look forward to sharing with you some of the submissions or sort of the top submissions that the learners gave us over the next few weeks. And we finish by Feb 7, so we'll try and collate these submissions probably by the middle of Feb. So thank you, both of you. >> Thank you. >> Yeah, thank you. >> Thank you for having us. >> We look forward to hearing some fresh thinking from the learners. >> Thank you.