Google: Selling Display Advertising

Brian, it's a pleasure to have you with us. Could you share some context to when you first joined Google?

I joined Google a little more than 13 years ago. I actually joined Google at the time they were acquiring YouTube, so my job was actually to help launch a distinct sales team in the Google world to sell advertising on YouTube, and it was called a Display Team at the time. As folks may or may not recall, back then, Google was selling almost exclusively search advertising, performance-based advertising, and this was the time that Google was diversifying its advertising platforms and products to include video and more display opportunities. So the team that I was involved with starting there was the team focused on video and display.

That was the time that analysts and folks in general were trying to figure out how could Google have spent so much money for YouTube? Because it really hadn't been a well monetized platform at all. It really wasn't well organized, and now in hindsight, it was a rounding error in terms of Google's fiscal position and the revenue that it's generated has been amazing. It's also been a platform that's become, I don't want to call it a standard setter, but has really created the environment for so many new ways of content being created, distributed. So that's when I joined. Very shortly thereafter, Google decided that it did not want to have a distinct display team so we became integrated into the larger sales teams. Google tends to rethink and reorganize and re-imagine how sales teams should be organized every couple of years. So eventually, we essentially became mainstreamed within the organization and not a YouTube standalone organization.

How has the thinking at Google evolved in structuring sales?

Yes, so that could be a whole hour. But one thing has remained true always, and that is that the customer relationship is the primary focal point of any sales, of any organization and it's organized around industry segments or sectors, so that a seller has a set list of clients and they are within a specific industry, so that that seller learns the details and the nuance of an industry. Whether it's financial services, travel, CPG, healthcare – where I spent a lot of time – you learn that. What has gone through different iterations as Google as expanded its offering, is what do partner teams look like? Are they built out around different ad products?

So while I mentioned I started at Google to help launch a YouTube sales team, there wasn't a YouTube sales team, but now there's a person who represents Google at a client and has to be able to talk about search and talk about display and talk about programmatic and talk about video and talk about data analytics. How that gets organized, in terms of partner teams, that's what changes a fair amount, in terms of what's the relationship between that person and the sales team, that person and the client. Are they attached more closely to the product team? Are they attached more to the sales team? That's gone through iterations, but the one thing that always remains constant is this core, what I call pods, this concept of a pod, that is an expert in an industry and has a finite set of accounts that it knows. That's always been true in the 13 years that I spent there. It was the surrounding support, partner, product oriented or service-oriented roles that kind of moved around during that time.

And so can we run through a 101 of display advertising, in terms of the core functions and different stages of the process, between demand side platforms and supply side platforms in display advertising?

If you're a customer and you're purchasing display ads, you are largely going to purchase it through probably one of two entities, but you can't dismiss the role of the agency in this conversation. Because it could be that the end client actually has no idea how their display ads are purchased at all. In some cases, they don't and I believe that that is to their detriment. The ads purchased directly by the end client or the agency can be purchased in aggregate across, a very large network like a Google, who has such enormous amount of inventory across such a wide array of sites, that it’s a simple, almost one stop shopping opportunity. There's also the opportunity to purchase via point players, who are display networks or display experts, who will bring a perception of specialization in display to what they offer to the agency or the client.

In some cases, that expertise is very real where they've built proprietary technology to essentially provide a better mouse trap. In some cases, it's really just human capital applied to the problem to dedicate people to analysis and optimization. I'm going to try to portray a balanced point of view as an ex-Googler, or also known as a Xoogler I think.

Are you familiar with the concept of the LUMAscape and LUMA Partners, the industry private bankers? People who watch this could look up LUMAscape if they don't know it already. Essentially, this investment bank built a map of the industry. When I started in 2007 there was no LUMAscape because things were really simple. Then LUMAscape created their first LUMAscape map of the ad tech world in 2010, I believe it was. It was on a single page. I just checked recently and I think that LUMA now creates and publishes 18 different maps to cover the industry.

So back to the conversation. You've got display buying highly fragmented, because you have all these point players who are representing probably segments of the available inventory on the universe and may or may not bring a proprietary component to it, either technology or high level human capital applied to the problem, which is making the display investment work as best as it can. The end client, in partnership with their agency, has to figure out, where do they get the best performance?

For a long time, display was not held to the same performance standards that search would have been. Search was a performance-oriented ad buy. We're going to go from a click to some transaction. We're going to measure that and we're going to know what our return on investment was. Display was not held to that same standard, for the most part. It may have been held to a higher standard than other forms of media but certainly not the same standard of search. You could get away with buying across several networks, having disjointed reporting functions, disjointed dashboards and hoping for the fact that you get some tonnage and you could get your brand out in front of a bunch of consumers and that, eventually, it would contribute to conversion through some other platform or advertising or marketing activity.

The technology has increased its capabilities and either the owners of the content or the managers of the ad tech can now apply many of the same performance measurements to display that could be applied to search; there's that accountability that didn't exist before. With that accountability the investment gets a larger amount of scrutiny, so now the selection of a partner, either a network or a Google or whoever it is that you're partnering with, becomes much more critical and the investment gets a lot more scrutiny.

Honestly, I've seen that with many end clients who are working with multiple display partners or multiple points between the purchase and the actual serving of an ad, where more than 50% of their investment and their display dollars was not in working media. It was in commissions to various platforms along the path. We've had a highly fragmented world on both the demand side and supply side which has created confusion for the buyer who is aware, and probably inefficient investment for the buyer who is unaware. I'm seeing a path forward, where consistency consolidation seems to be what is required. That's what a Google offers to a certain extent – this consistency and consolidation by having enormous scale – but then being able to tie measurement of your search investment to measurement of your display investment and see how they work together. It's the same with video and even going down the road into offline and how offline contributes to an eventual conversion.

What is the rationale for the advertiser or the client, whether it's the agency or the end client, to use a different network? Is it kind of just, not boycotting Google, but is it that they don't want to put all their eggs in one basket, effectively?

That legitimately exists and I know boycott is not exactly the word you wanted to use, but it's a concept of diversifying one's portfolio. There's a business rationale that says, I should not put all my eggs in one basket; I want to diversify my portfolio. Part of it is, people still buy from people at the end of the day. So if Google comes in pitching math and science, which in large part it will – and I don't mean that in a derogatory way, because the math and science works – and someone else comes in and says, yes, Google is going to pitch you with that math and science, but we've got humans who are going to manage your campaign every day and we're going to be attached to you.

This is grossly oversimplified, but Google could say, with all their automation, it's a set it and forget it world. Our offering is going to be a dedicated account manager who works with you to understand your business and optimize on a daily basis; that will work for a lot of people because they feel they want that attention. That will be part of it as well. The third piece is when an agency is involved, and I don't want to suggest collusion or anything inappropriate here, but they may be aligned with certain networks. There are alignments that happen for business reasons. Maybe they've negotiated some sort of volume deal with the network that they can pass along to their clients. There can be a number of reasons, which in the eyes of the end client, at face value, could all seem very legitimate. They're not experts on the technology. They don't know every logo on the LUMAscape. They're either going to make a decision based on a relationship that they've built or on a third party, an agency that they've hired, to make that decision for them.

How did the acquisition of Double Click change that industry structure in the ad tech world, specifically for Google?

It's really interesting. We talk about portfolio development and it gave Google capabilities in every place in the pipe. Over time, almost every key point in the pipe has been either built or acquired by Google. I mentioned earlier, every point between when an ad gets purchased to when an ad gets delivered to a consumer and its impact is measured; Google has technology for that now. It was one of the single largest pieces in creating a fully threaded pipe for Google.

It obviously created an enormous amount of pressure on other competitors to get smarter faster. It also created enough pressure so a lot of competitors said, this is borderline monopolistic behavior. What is Google's end game here and how do you as an advertiser, want to be treated in that world? Not just diversify your portfolio but do you know that if they own everything are you actually getting the best efficiencies that you can in terms of what you pay in that scenario? It created a couple of different scenarios that were really interesting. It also created some internal issues where we created firewalls between the DoubleClick sellers and the end ad sellers. There was a sense that, in the market place, if those two bodies were connected to each other, the ad sellers – the teams that I worked with – would have too much insight into the back office that would give us an unfair ad selling advantage.

So there very much was a firewall created between the two organizations that I'm assuming still exists. I haven't been there for a few months, but it's very important in order to have confidence at the agencies in particular, that we were not rigging the deck on our behalf to make sure that ads performed better for our clients or that we knew more than we should know, in terms of performance. That's an interesting dynamic.

So DoubleClick actually operates on both sides, as an advertiser service and publisher, ad server as well?

Yes, absolutely. Again, that's the full pipe because the first stop in the consumer experience on a site, other than Google, is at a publisher site, where they're engaged in content and they're seeing that ad, so there's absolutely that publisher relationship as well.

Can you provide some context around exactly what responsibilities you had as healthcare industry director?

I was specifically the industry director for what we called health services for Google in the United States. That meant I managed a national team who represented our ads products, primarily for very large hospital systems; several of them Fortune 200 type organizations. Large health insurers, payers, one of them is a Fortune 5 company in the United States; and then a select group of consumer health products.

So I managed a team of sellers who went into them and brought the entire array of Google Ad products, and then a service team, account services or account management teams, primarily. Earlier I described these partner teams; I had relationships to data analytics people and to the product experts who supported us. For every member of my team, there was another person. If I managed a team of 40 nationally, on a day to day basis, I actually worked with 80 people. We really had a well-funded support operation. At the most basic level, I was responsible for driving year over year growth of ad investments by our clients in Google. Google has very ambitious growth goals. That was the first piece.

The second piece was that I was responsible for the personal and professional growth of my team. Google takes that very seriously. While there's a lot of opportunity presented at scale by Google, as a corporation, to support the growth of employees, part of my every day role, how I was measured for success, was what I was doing to help create an environment in which people could feel safe, they could succeed and grow.

The third piece was becoming an industry leader. There was an expectation that, as an industry director, you would become an industry leader. You would have a point of view that would be sought out by your customers. My job was not to go in and sell the next quarter's search program. My goal was to go in and to articulate, what are the deep insights that we can observe and quantify and then put that into a narrative to show how consumer behavior might be changing, as it relates to interaction with your own healthcare. What does that mean to a provider of healthcare, one year, two years, three years from now? I don't think it's very reasonable to think any further than that. Three years is probably even a stretch with the rapid evolution of technology and how people manage their lives every day.

Take that story to the executive teams of customer partner organizations and really create a relationship where they said, they are challenging me to think differently about what the consumer experience is like and how people are taking care of their health and what that means for our business at a larger business level. Over the last couple of years, in particular, we had a really focused rigorous effort on trying to move ourselves out of the category of marketing our advertising vendors with our customers and into business partners with our customers. Everyone wants to do that but our path to trying to do that was by moving to a conversation where we worked with our direct contact.

Everybody says that's what they want to be but we really gave this concerted effort to meet with the C-Suite – that was my job – and to understand what is their business objective? Historically, if you think about digital advertising, the vendor talks to the customer about click rates and impressions. Those are not business measures or metrics; those are marketing metrics. What we wanted to do was go to the C-Suite and say, what is your business objective for 2019/2020? Let's state it in very specific numbers and with specific time frames in exec language. Generally, it's going to be, we want to increase profit margin by X in 2020, or new patients by Y in 2020.

Once we did that, we could then establish how the marketing activity laddered up to the achievement of that business objective. My job was to tie all that together. It was to say, now that we've got those objectives, what kinds of strategies are we going to build to help them reach their objectives? We had such a portfolio to draw from. If you think about healthcare in particular – if someone in the industry sees this and if someone I know in the industry sees this I say this with love – they're behind from a digital adoption perspective. It's always been ironic to me because if you walk into a hospital, the world's greatest state of the art technology is often times being deployed to treat us. Yet if you walk down the hall or really into the basement, to the marketing department, it's 2010 in many cases from a marketing perspective.

They're not thinking the same from an innovation and technology standpoint to help their business. I started at the C-Suite to try to have this top down pressure mount that says, Google has insights into consumers that we should be better deploying in the way that we market. Not just are they behind in terms of deployment but healthcare, historically, hasn't thought about marketing as a profitable growth driver. They've thought about it as a necessary evil in their budgetary process. If I could then show, through measurability, that ads and marketing, particularly digital ads and marketing, are actually profitable growth drivers, then that's a C-Suite. The CFO wants to talk to me about that. The COO wants to talk to me about that. The director of marketing gets promoted if we succeed at that, but the CFO wants to hear about that. My job was to elevate the conversation and it was to play the narrative.

Earlier I said the role was to be an industry leader. Every industry has their conventions, their events. We started hosting healthcare events at Google. We would bring them in and try to provide, at larger scale, introduction to all these thoughts, provided by our either third party partners or the early adopters, to try to bring all of them along. I will say that in the area of healthcare, a lot of it is not-for-profit. A lot of these large hospital systems are not-for-profit and they have a more collegial relationship with each other. They're willing to share best practices, whereas in some categories they won't share a best practice because it's a secret sauce. In this category, the mission is much different. The mission is to keep people healthy.

Let's walk through an example. Let's say that I am a for-profit hospital or a device manufacturer. You come to me; I tell you I want to increase my revenue or patients by X amount. How do we get into the strategic discussion around which products to use?

We'll make up an example; say we're a hospital system and what we want to do is, in 2020, we want to take our top five service lines – a service line is pediatrics, neurosurgery, whatever it is – we want to take our top five service lines and we want to go from 80% of bed usage to 90% of bed usage. This means we'll need this many more patients in each of those top five service lines. If we understand what those numbers are out of the blocks, we actually work with them to create first just an understanding of the math. That means that given current benchmarks and current success metrics, to get your next 2,000 orthopedic procedures in the next year, that's going to require a budgetary increase of X.

We've been able to build all of that using, not just Google data, but third-party data. We'd use Google query data. How many queries are there around hip replacements and how do they correspond to areas where that enterprise has hospitals? That's one set of data. We'll also look at the hospital or the system's own data about bed usage and some other marketing conversion metrics. We'd also look at American Hospital Association data to look at competitive data against other systems because there's a competitive nature. Even though, as I said, they're collegial, the reality is if you're in a place like Houston, Texas, there are at least five different hospital systems there all hoping to get that hip replacement.

We do the math for them because again, interestingly enough, when I talked about healthcare being behind, they weren't doing this math for themselves. When we came in and did this math for them, they thought we were wizard ninjas from a future world. They hadn't done it for a number of reasons. So we do the math. Then once you do the math we look at the competitive and consumer landscape where they are. What is the role of television advertising and how will we then position digital advertising? What is the competitive spend for the competitors in this marketplace? What are the conversion rates for the different platforms that will be used and what would it cost to play in each place? So we would build a plan from the bottom up, based on their numbers, based on that. I will say, quite frankly, because they were less sophisticated, as a generalization, we would start with search. We would start with search because it's the most measurable, it's the most obvious, it's the most urgent and timely. You can see the result sooner.

When we started to layer in video, when we started to layer in display, the results were going to be further down the line and if they weren't believers or adopters yet, there's a higher likelihood they would want to see something in a month; not see it and say it doesn't work, rather than make an investment to learn and iterate.

What's the barrier to the adoption of video or display versus search for advertisers?

Part of it is what I just said. It's that instant gratification component. You may or may not see the results right away, is number one. Number two is the power of some existing media platforms. If they're already a television advertiser or a radio advertiser or in hospitals, as an out of home billboard advertiser, it's hard to move away from something because any move might be seen as a gamble or a risk. Even though we would be able to say what we're going to do with you is, we're going to set some goals and partner with you to see if we reach these goals, if they don't reach them or worse yet, if their same store or same facility results go down for that period of time, they immediately put the blame on that. There's risk aversion and some industries are more risk averse than others. Healthcare is highly risk adverse.

The fact of the matter is everyone has always had to go to hospitals anyway. People get sick. They'll be going less and less in the future, based on our recent experience and the growth of telehealth, but people have always had to go to the doctor. We know people are going to come to us no matter what, so why rock the boat? Why do anything experimental, when they always have to come to us anyway, in this category? There was a piece in healthcare that's particularly prevalent which is this perception that most healthcare providers are more concerned with privacy than any other single industry segment. Everyone is concerned with it but when you think about a person's health records, the care and the seriousness with which they place privacy and brand image, as it relates to advertising, is paramount to healthcare. So if they're going to develop a partnership with Google or Facebook or go down the list of other possible providers, there's this fear that, oh my goodness, you're going to have data about my patients that you shouldn't have and God forbid there's a security breach and I show up on the front page in the headlines.

How do you get them over that?

You hope you find somebody in the organization who actually has the faith. It's not only just the faith, it's the understanding that the protection of that data is just as important to our survival as it is to yours. If the data that Google has on consumers is abused, we can't go sell to the next person either. That is Google's whole business, knowing an enormous amount about a consumer behavior, but also protecting it with great care and great security because if that leaks out or that's abused, then Google's business or a large provider is fully compromised. You bring them into that and then you have to walk before you run. Then also assure them that – and there have been some major breakthroughs in some areas that have helped us with this – we don't want to see their data. Maybe with a travel client, we'll connect to their CRM so you can look at first party customer data and use that for better targeting. You'd never do that with a hospital, if in fact they're even using a CRM; let's stay away from that because only bad things can happen. Even the idea that we could do that is a bad idea.

It seems like it's still pretty early for Google in shifting these, what is it, 50% of total advertising is digital, so I think that's kind of the long-term case for Google?

I think that eMarketer has said that next year is the first year that more than 50% of all ad dollars will be invested in digital. This industry still has significant growth upside for all of the platforms.

Let's say I'm going to advertise then and you convince me, I agree to move part of my ad spend over to Google; what products do I actually use, is it the marketing platform, is it AdWords? What exact new product do I use?

The products, essentially, over time are being consolidated. The ad platforms themselves which are not all consolidated yet, but your view of your search campaign and your display campaigns and your video campaigns, you're able to view those all on a single dashboard right now.

Is that Google 360?

Yes. Then you start to layer on if you become a Google Analytics customer, to further enrich your ability to understand the data that's generated and apply the insights from that to ongoing campaigns. That starts to get to the full array of products that you're going to use. That's where different agencies are going to have a point of view of that because some of them are going to offer their own solutions; there are competitive solutions out there. The team that I would have represented wouldn't have been selling the platforms. We would have been selling the execution. Again I mentioned that 2X number of partners that we would have who would come in and say, we think you should be adopting Google platforms on which to manage and serve your ads.

That's part of that collaborative sales environment and in some of those with the DoubleClick, I don't want to call them the legacy DoubleClick people, but that's the platform side of the house as it's evolved over time. Often, we would sell in tandem but not complete partnership because we would have different end goals and we were not necessarily even talking to the same person at the client. We might be talking to the agency, we might be talking to an IT-ish type person who serves marketing, whereas we would have been talking primarily to the marketer.

So you actually have multiple sales people with big enterprise clients selling?

With the largest enterprise client exactly, and that creates opportunity and it creates challenge. You have to be crystal clear with the client so that they can understand what does Brian do versus what does Susan do versus what does Jake do.

Would you do that as well, I assume?

Depending on the client I might because I was a leader of an industry and because, in almost all cases ads, is the number one buy already for most; they've all been doing ads at some level, whether it's lots or little. I or a member of my team would play the general contractor role. We would help facilitate meetings for other groups. We would then help bring it all back to the Google office and say, where are we all each with what we're trying to do? In healthcare it grew to ever greater dimensions because, for instance, Google Cloud created a HIPAA compliant cloud solution for healthcare organizations. Now the earlier challenges of data access, where marketing could look at other data in the organization to help them make critical business decisions, could be overcome if the healthcare organization was a Google Cloud client and put all of their information into a data lake.

You almost cross-sell from cloud?

There was definitely mutual self-interest in making sure that the cloud team would get an introduction from us. Google also introduced a health product area.

It's almost like cloud is acting as a lead generator for advertising?

It's usually the other way around. It could create greater opportunity if that deal is made. The ability to query the data, in a secure environment, would create more actionable strategic data for the marketers and for us to work with the marketers on.

How did you find the end client look at YouTube and the positioning of the brand advertising opportunities or display on YouTube?

If you look at the ad investment segmented by industry, healthcare would be very, very low. There are several reasons for that and maybe I'll start with healthcare.

Let's start with generalities for this conversation. Most marketers think of YouTube as a platform where younger consumers engage the content. It's all relative but we'll say younger, under 34, under 40. Most healthcare organizations are looking for older consumers, because older consumers are less healthy. Older consumers are buying insurance. So there was automatically, whether it's right or wrong, this perception of, the audience is not there. So that was the first thing we had to overcome. The second thing we would have to overcome is related to healthcare and every one of the industry segments that Google serves has had to overcome this at some level.

It could be that our audience is there but there aren't as many there as television or you name the other platform. So first it was that. Second thing was privacy, for data and for healthcare providers. Third thing their brands. Content on YouTube is the Wild West and we know there were stories over the last two years about ads on YouTube appearing next to less than favorable content, and we don't have to rehash that. But there's a concept of, I'm the Mayo Clinic – and I'm not speaking on behalf of the Mayo Clinic now, I'm just using them as an example – I'm the Mayo Clinic, I'm possibly the preeminent hospital brand in the United States and many parts of the world. My brand should only appear adjacent to the highest quality content and when I think of YouTube that is not what I think of.

Why is that an issue though? What is the issue on the data structure or how can YouTube improve that?

YouTube can deliver the answer to that; but your question was what's the advertiser perception? My job to drive the relationship at the C-Suite, at some point, an executive is going to hear, you're investing marketing dollars on YouTube? Isn't that rock videos? Isn't that individual extreme sports fanatics? Isn't that cat videos? That's the advertiser perception. So a couple of things, like I said brand safety, audience quality, audience segmentation I would say are the primary objections. But you're exactly right, when you get to the actual delivery of the ad capability, it can be targeted to only the highest quality content.

So for instance, the Mayo Clinic. They're not just about ads to get people to come for a hip replacement, and then frankly they're far more specialized than that, globally. We could say that healthcare organizations like Mayo are also trying to become lifestyle managers. Teach you how to eat better, teach you how to sleep better, teach you how to exercise better. We could get a Mayo diet ad or diet content in front of household managers, on high quality sites where they're seeing Better Homes and Gardens, Martha Stewart Real Simple, the types of content delivered here in the United States, that are high quality audience, high quality content. That can all be managed, but first we would have had to overcome that initial perception of quality of audience, size of audience and quality of content.

Have you seen that change in your time, over the last few years, in the perception from advertisers?

It definitely started to change and it gets back to something I mentioned earlier. This is kind of a sales tactic concept which is, you've got to work your way around an organization to find a champion. You've got to find someone who gets it. They're either just someone who gets it or, in many cases, they're the target audience. They live the behavior we're trying to show.

As new, younger people come into work at the client and they're that target market and the old guys retire,

That's what I used to say, when I started 13 years ago and I started working on YouTube. I said, we need a lot of people to retire for this to work. It's happened faster than that because that whole generation hasn't retired yet, but that's why you have to change a couple of things. If you can understand the business objective and then you can actually show them a series of events where a YouTube ad, in the right place, leads to a click on your site, leads to a return visit on your site, leads to a branded search that arrives at the doctor and they become your patient. If you can show them that and show that the margin for that cardiology patient on your end is so high that the investment that you made for that is just such a wildly efficient return on investment, you can win. But you need to be able to get to that level of detail with that senior executive to make that happen.

I can tell you a quick anecdote. One of the larger for-profit systems in the United States, I was meeting with their executive team, and I don't have it sitting with me now, but I held up my mobile device and I said, do you know that when your patients have a perceived health issue and they want to research it, they're searching for it on this device? They're clicking on something on this device, potentially, in best case; in another case they might be going eventually over to their laptop or to their tablet or whatever it is, and they're showing up at your site. You have to have a presence here., Chief medical officer. I had data to prove it. I brought all the data, much of the data I described to you and I showed them the data.

He's the chief medical officer. He's a scientist; data should work. Afterwards he said to me, I know I should believe you, you showed me data, but since I don't behave that way, I can't support an investment in that. I mentioned the Mayo Clinic, they were coming around to this. I'll tell you a phenomenon I was seeing, I'm not sure if this is what you're looking for. In healthcare, the most progressive organizations right now, at hospitals and insurance providers that I described, to lead marketing, they're hiring people from out of category so that people, who have seen this succeed in other places and have enough self-confidence and enough stature, come into a healthcare ;organization and say, no we have to move to this; I've seen how it works, it will be a growth driver for us. They don't have the biases and they haven't been damped down by the system for 20 years.

How would you look at the targeting versus Facebook for YouTube? When I'm on YouTube I seem like I don't get as great ads, to the level of targeting, I'm curious to why that would be, given that Google is Google?

I can't speak to your experience but the same targeting that you see across most of the rest of Google's platforms should exist for you, on YouTube. I see it; I'm a pretty active YouTube user for my own entertainment and information gathering purposes. I do see things following me around. I shouldn't use that word, but I do see things targeting me that are very appropriate to my recent online behaviors and to my demographic. I think Facebook, and if you have a different experience with Facebook, Facebook is layering on some other, I want to call them user ID logged in data points, that might start to get to the edge of where Google is comfortable going. By saying this, I'm not trying to say Facebook is doing anything illicit, but the types of things that you're doing on Facebook are different. Through Google, it's an engagement with content and it's a behavior. In Facebook, you're actually giving them more about you that they actually know. They actually know your age, they actually know your address. They actually know who your friends are. There are certain other things that they can layer on. I think that that can be problematic in some cases when you get to healthcare marketing. It's less a concern but still a concern for certain less regulated industries, CPG, travel etc., retail.

Would Facebook be more powerful for healthcare customers?

I think in some ways it can, in terms of creating communities for individuals who have similar conditions, absolutely. Google doesn't really have an offering for that. There are certain capabilities that, absolutely, I believe Facebook is much more suited for. That would be one of them. I also think however Facebook could potentially get closer to the third rail of that privacy risk or that brand perception risk, by taking some of that information and making it available to advertisers.

How how have you seen the agency's role in digital advertising?

I'm going to articulate this from Brian's standpoint and not Google's standpoint. As a person who worked for 13 years at Google it is hard to separate the two, but I'm going to come at it from Brian's standpoint. I always thought that agencies were in a really hard position. Even 13 years ago, when I started at Google, and I'd been at AOL before that, so I'd been in the digital ad space for a while and others before that, the agencies were a third party that was just getting pinched. As companies like Google developed more analytic capabilities, more understanding of consumer trends, agencies were losing some of their unique selling proposition. It used to be that the agency did the consumer research. It used to be that the agency created the consumer insight. It used to be that the agency knew everything about that and therefore they drove the creative.

Agencies were slow to adopt the technology, and as the technology companies like Google came along, they started to become more of the strategic partner than the agency, and the agency started to become more of the execution activation part of the equation. Now some large holding companies started to build portfolios including technology, creative, so they could offer within their portfolio all of those things over time. Many started from scratch or maybe started late. Where I see the most successful agency relationships right now, where I saw them with our clients, was where the agency was fully willing to sit with us as an equal partner and we wanted them to succeed.

We knew if they had been selected by the client to be that person who makes decisions for them, if we could feed them information, to be able to now bring insights they might not have been bringing in the past, while not cutting us out of the insight conversation, that's a win-win-win. The agency is then seen as strategic. We're seen as strategic as the provider of that data and as the party able to crunch the large data, and we both can drive to better results. That's where I think it got really hard for them.

Google is slightly disintermediating them, in the long run?

I don't know if that's really fair. It could be perceived that way.

What do you actually create then, let's say five, seven years out, what value do you see the agencies actually providing?

I had the good fortune to sit in some small meetings with Eric Schmidt, who always said, none of us are capable of thinking more than two years out; don't try. I think that the agencies that survive and the agencies that thrive will invite the Googles in to be their tech partner. They'll actually say, you're not a vendor – and I think this is true with their relationship with Facebook, I think this is true with their relationship with Amazon – we invite you in to be our tech partner and we invite you to be an equal partner to the relationship with the client and not someone that we are vetting for the client and making media decisions for or against.

That does not happen, at scale, across the board, for a lot of clients. There is not this invitation for technology. I think that the agencies have to be open to that and then transparent with their client that that's the role they're now playing, that they are not vetting media, they are vetting technology. Marketing and advertising is becoming a tech game.

Yes exactly. Let's take my position, so if I'm a manufacturer or a hospital, I bought TV ads through the agency; I now want to experiment with Google and I'm moving 50% to 60% of my dollars over to Google. Do I need the agency to operate my Google marketing platform for me or what role does the agency play in actually me allocating my dollars?

It's an excellent question. It's a large general question; but here's the deal. The idea of bringing things in house is very, very real. It will happen because, you're right, you may not need the agency. We talked about this with many of our clients. We said, you might want to consider going in house, maybe when they were in an agency review process. It was like okay, review different agencies but you might want to think about bringing this in house. But you don't have the expertise to bring it in house. Maybe what you need to do is think about a bridge solution. You hire an agency for a year and in that time, we'll work with you to spec out what are the roles and responsibilities and skills necessary to bring it in house and what will that department then do?

Many don't want to employ that. Depending on the margin situation of the enterprise and the head count view and budgeting process of the enterprise, they may not want to do that. However, to have an agency that they can, essentially, negotiate down to a very skin in the game relationship, and that is disposable after a year, they might like that game instead. They have their own business concern. Is there a case to say bring it in house? Hell, yes.

They might be happy to use the agency to use the Google platform for them?

That's the thing, the agency becomes their department; that becomes their marketing department.

That’s non-working dollars though, right?

You're right about that, but this is why the margins are getting crushed for the agencies because the agency desperately wants that client and they're willing to take next to nothing to keep the client.

So it could be interesting then, let's say when all the old guys retire, young guys come in who are more savvy around this stuff, it could work out cheaper for the advertiser to in house that and then have actually more working dollars to allocate via Google or digital?

Absolutely; there's a case to be made for that. Like I said we asked people to consider that when they were going on agency reviews.

Is there anything else that we, maybe just looking out for the next few years or anything that you would say in terms of how the industry is going to evolve in Google offering their services to larger clients?

Large clients are still not taking full advantage of the machine. Everything about what Google has always done has always been machine learning. Let's really get on board and understand that machine learning works. You can create other efficiencies, you're right. Get out of the optimization game and get into the strategy game, client. Let Google's machines do the work for you and you or agencies spend more time on the strategy. Spend more time on a couple of years out. So adopt automation. Learn that the machine is your friend. Everybody, the machine is not going away, the machine is your friend.

I have no clear picture of what this looks like, but ad formats have to evolve to this cliché internet of things. The fact that we're starting to gather data from everything we touch in our lives right now, and the fact that that creates more opportunities to understand consumer behavior and deliver messages to them, creates far more opportunities for marketers than I think is being fully taken advantage of. I think that advertising as we know it, five years down the road, 10 years down the road, changes a lot or crazy unimaginable formats come into existence, around everything we touch every day. Some of those things are already starting to become real, and I don't know what they all look like, but I just think that all of those data points can be used and deployed and that’s not happening. How do we do that in more places, with new formats, or new modes of interaction?

Brian, thank you very much for your time it was great.