Good day, everyone, and welcome to the Rubicon Fourth Quarter and Fiscal Year 2016 Earnings Conference Call. [Operator Instructions] Please note that this event is being recorded. I would now like to turn the conference over to Bonnie McBride [ph]. Please go ahead.

Thank you, and good afternoon, everyone. Welcome to Rubicon Project's Fourth Quarter Fiscal Year 2016 Earnings Conference Call. As a reminder, this conference call is being recorded. Joining me today are Frank Addante, Founder and Chairman; Michael Barrett, the company's newly appointed President and CEO; and David Day, our CFO. Before we get started, I would like to remind you that our prepared remarks and answers to questions will include expectations, predictions, estimates and other information that might be considered to be forward-looking statements, including, but not limited to, guidance we are providing and other nonhistorical statements related to our anticipated financial performance, operating and strategic plans, expectations regarding new initiatives, our relationship and business with buyers and sellers using our platform, competitive differentiation, fees and take rate, capital investment and organizational development and competitive position, and market conditions and trends and growth expectations, including growth in header bidding, orders, mobile and video. Forward-looking statements involve risks, uncertainties and assumptions, and actual results may differ significantly from the results suggested by forward-looking statements for various reasons, including, without limitation, if such risks or uncertainties materialize or assumptions prove to be inaccurate. Further, we may adjust our plans and expectations in response to market conditions or other factors. Reported results should not be considered an indication of future performance. A discussion of these and other risks, uncertainties and assumptions is set forth in the company's periodic reports under the heading Risk Factors and Management Discussion of Analysis and Financial Conditions and Results of Operations. We undertake no obligation to update forward-looking statements or relevant risks. Our commentary today will include non-GAAP financial measures. Reconciliations between GAAP and non-GAAP metrics for our reported results can be found in our earnings press release and in the financial highlights deck which we have posted to the Investor Relations website at investor.rubiconproject.com. At times, in response to your questions, we may offer incremental metrics to provide greater insight into the dynamics of our business. Please be advised that this additional detail may be one-time in nature and we may or may not provide an update on the future of these metrics. I encourage you to visit our Investor Relations website to access our press release, financial highlights deck, periodic SEC reports and webcast replay of today's call to learn more about Rubicon Project. I'll now turn the call over to you, Frank.

Thank you, Bonnie, and good afternoon, everyone. I'm excited to welcome Michael Barrett joining David and me on today's call. Michael's joining us as our new CEO this week. I will speak more about the context that led to our decision to ask Michael to join the team. But before I talk about the state of the business, I would like to say that we are pleased that our financial results in Q4 came in within the range of outlook that we provided last quarter. Our fourth quarter wrapped up a difficult but highly productive year for Rubicon Projects, as the team faced multiple challenges throughout the year and met them head-on, enabling us to further expand our platform. In addition to the success we have seen in mobile, video and orders, we are now beginning to see very promising results in header bidding. The difficult decisions and changes we made in 2016 have resulted in the company reinvigorated with focus. We now have all of our resources, capital and attention committed to our most important business, the ad exchange. Over the past couple of quarters, we've gone through a lot of effort to restructure the company so we can successfully transition our platform to best serve the ad exchange. This includes our accelerated product introduction of FastLane, our organizational changes and our decision to divest the intent marketing business. To move forward in the right direction and return to revenue growth, we must continue to build strong relationships with publishers and application developers, operate at scale and continue our transition to build technologies required in a mobile-first world. With that, I'm excited to have Michael Barrett joining us as CEO. Michael and I are very aligned on vision, and he's a perfect cultural fit for the company, which is…

Thank you, Frank. I'm very excited to be joining Rubicon Project and to work alongside you and the terrific team that you have built. As Frank noted in his comments, I've had the opportunity to compete against Rubicon Project as well as be a customer, and have always been very impressed with the professionalism and industry leadership that you've shown over the years. So when Frank approached me, I was excited by the opportunity, and I'm delighted to be a member of this team. The advertising space, particularly the programmatic piece, continues to grow, both in size and importance to buyers and sellers. Rubicon Project is well-positioned to capitalize on these trends. Yes, the industry is dynamic, and this reality can lead to disruption. Frank has spoken in the past about the challenges and opportunities that innovations like header bidding brings for Rubicon Project. I look forward to working with our customers and team to make sure we are delivering the best possible product and monetization capabilities going forward. Rubicon Project has a strong leadership position, a strong balance sheet and an aggressive and innovative product strategy. We are well-positioned to become the leading independent global exchange for publishers, application developers and advertisers. I'll be heads-down for the next few months meeting with customers and our team. Soon after, I look forward to meeting with all of you. Again, thank you, Frank, and the rest of the board for this opportunity. David will now give an update on the company's financials.

Thanks, Michael, and welcome to the team. I'd like to start by saying that we're grateful for Frank's pioneering vision over the last 10 years, which has led Rubicon Project to the position in the industry that we enjoy today. I look forward to Frank's continuing involvement with the company, and I'm excited to work together with Michael. For the full year of 2016, we generated over $1 billion in advertising spend, $278.2 million in GAAP revenue, $256.1 million in non-GAAP net revenue, $70.9 million in adjusted EBITDA, and $1.07 per share in non-GAAP EPS. We executed well against our revised goals for the fourth quarter, which took into account the continued year-over-year decline in desktop advertising. Advertising spend for the fourth quarter of 2016 was $277.1 million. GAAP revenue for the fourth quarter was $72.7 million, and non-GAAP net revenue for the fourth quarter was $66.9 million. The decline in advertising spend of 18% year-over-year during the fourth quarter of 2016 was driven primarily by a continued decline in desktop, which declined 23% year-over-year to $178.2 million. We also experienced a decline of 5% year-over-year in total mobile advertising spend to $98.9 million. This is a result of declines in our mobile Web business, which still represents the majority of our total mobile business, in contrast to our mobile app business, where we expect more promising growth in the future. Advertising spend was composed of 36% mobile and 64% desktop for the fourth quarter of 2016, compared to 31% mobile and 69% desktop a year ago, reflecting our continued shift towards mobile. While we've historically shared additional advertising spend detail for real-time bidding and orders, we will no longer be sharing those details for a couple of reasons. From a product perspective, we are focused on developing additional products…

Thank you, David. I spoke earlier about why I feel -- why we feel positive about our product innovation and our execution in capturing market share of supply, both key drivers for our future success. While it's clear we still have some work to do to get this optimism reflected in the financials, we believe that by building on the strong positions we have in mobile, video and orders, and continuing the accelerated progress we're making in FastLane, we will emerge as the leading independent global exchange. Before opening up for questions, I wanted to reiterate my excitement that Michael is joining our team. He's a proven leader and has the operational experience and strong customer brand. We're lucky to have him onboard. With that, we'll now open up the call for questions.

[Operator Instructions] And our first questioner today is going to be Kerry Rice with Needham.

Maybe, first, a question on the header bidding or FastLane. What is the best way to best understand the penetration of FastLane? Is that the number of publishers, which, I think, you mentioned around 300 deployments out of, I think, roughly 500? Or is it maybe the percentage of impressions that you can sell? If you can give any more details around that, that would be great. And then, the second question is really around mobile. You mentioned that mobile came down, that was primarily the mobile Web, which I assume, again, is in relation to kind of header bidding. When do you think about mobile app revenue being the primary or the majority of mobile revenue?

Kerry, it's Frank here. So the best way to understand the penetration of header bidding, so first, just to kind of set the context of this, header bidding is not an entirely new platform for us. It's an API. It's another way to integrate into the existing customer base as well as new customers. So we have 1,300 customers overall. We have 300 deployments now with header bidding. The first stage is to sign the customer, then we've got to get it implemented. Then we've got to get it tuned and make sure that, that implementation integration is clean and operating efficiently. And then, from there, we've got to make sure that the algorithms are optimized for that particular site. And so, then, your revenue growth expands there. So if you look through the numbers, we've got 300 of our existing 1,300 customers, so there's a lot of room to grow from there. That's not including new customer opportunities, by the way. Even within that existing customer base, we estimate that about half of those or -- are tuned and fully efficient deployment, so even within those deployments that we have, there's some room for growth there as well. So we're pretty excited, obviously, about the prospects of it. And what really kind of drives that revenue growth, given that we're a marketplace business, is the access, as I mentioned, to supply. I mean, header bidding is ad requests, and those ad requests overall in our platform grew 50% on a year-over-year basis, even though we had a challenging 2016 year. Reconnecting that supply is critical for us in reconnecting that revenue growth, so we're feeling pretty good about that supply reconnection, even within the deployments that we have.

Frank, can I maybe just follow up there briefly. When I think about 300 versus 1,300 total customers, is it the intent, or is it the plan to all 1,300 should use the implementation of FastLane? And so, if so, how do we think about that progression of the other 1,000 deployments?

Yes. It's a great question. So look, our platform is designed to integrate with the publishers or app developers in any way they would like to integrate. So in some cases, we have multiple integrations. So sometimes, it's FastLane and our traditional Smart Tag business and our orders business. And sometimes, they'll also integrate our mobile business through SDKs. So sometimes, within that existing customer base, there's multiple integration points. We don't necessarily expect that every single one of them are going to implement header bidding. It's really the customer's choice. Yet, header bidding is -- it's grown more in North America, for example. It's grown more obviously in desktop. As you're moving to things like mobile app, yes, there's some nuances that changed there. And also, with our orders product, some customers just use our orders products, in which case, header bidding could be helpful, it certainly is, but not necessarily required. I mean, also, our xAPI products as well is a different type of integration point. It's basically server-to-server integration point. So while there's a lot of opportunity in that existing customer base, it doesn't -- we don't have to implement it with every single one of our customers to go reconnect that supply.

Okay. Then on the mobile Web versus mobile app inflection point?

Mobile Web does continue to constitute a majority of our spend, so we see that, I guess, a couple of different dynamics there. First, as we make better progress on our header bidding, that will stabilize the mobile Web side of the equation. On the mobile app side, we certainly continue to -- it's still a growth area for us. We have a product roadmap that will provide -- using our xAPI API, will provide a much greater access without having to have the SDK implementation. And so we think, from a product perspective, we're building that out. We've got significant momentum there. The timing of when that turns is just hard to say. As I mentioned on the call, it's just challenging right now with visibility, but we're certainly headed in the right direction.

Yes. We're making some good progress there already. One of the reasons that we're excited to bring Michael onboard is -- his most recent gig was CEO of Millennia Media, a mobile-only marketplace. So he's got a lot of experience in that area, and I think that's going to be really helpful to accelerating, hopefully, our efforts in that area.

The next questioner today is Andrew Bruckner with RBC Capital Markets.

Two, if I could. The first one, Frank, last quarter, you talked about the strategic market position as being a key question. I'm wondering if you've thought about that any differently over the past quarter, and if you think any sort of M&A or inorganic components are necessary. And then, secondarily, if you've disclosed any economics on the Tapad cross-device partnership? And if these are any capabilities that you could eventually have in-house versus having a partnership?

Yes, absolutely. So from a strategic market position, we're pretty bullish. I mean, if you look at header bidding, which is obviously an area that caused some challenges for us in the desktop display market last year, again, an area that we were less focused on because we had shifted our attention and resources to mobile, video and orders, which was where the market's heading. If you take that, right, so if you take something that was a challenge for us, and if you look at the rate that we've been able to turn that around from a customer implementation standpoint and a product implementation standpoint, you're pretty much coming from a cold start. We launched a product in the market, you get 300 deployments. 100 -- you go from 0 to $120 million of ad spend in just over 3 quarters. I think that illustrates the strength of our market position. It illustrates the strength of our customer relationships, and I think, it illustrates the -- not only the need, but the importance for a platform like ours to have strength of being desktop, mobile, video, orders, all those things in one place. Our global reach of demand is a critical factor there as well. If you look at our cloud computing infrastructure, our ability to bring on hundreds of bidders onto that, not only bring differentiated demand, but also creates more competition in the auction. So when we deploy something like header bidding, or mobile, or video or orders, we see that, that market share capture of supply rapidly increases because we're able to leverage that differentiated demand, that scale, our existing customer base. So we feel pretty good about that, even though with the header bidding piece, had created a little bit of hiccup for us. Our ability to kind of turn that around so quickly, I think, makes us more bullish about our ability to continue to deploy new products in the market in the future.

And with respect to Tapad, do you think there's any sort of capabilities you'll ever have in-house?

Yes. The Tapad partnership, it's too new to communicate any of the economics. For cross-device, we'll want to support any device graph that has buyer adoption and can grow ad spend for our customers, which could include proprietary and external solutions.

The next questioner today is Brian Fitzgerald with Jefferies.

This is John on for Brian. Just wanted to touch back on take rate for a little bit. In your as-adjusted numbers, you had take rate coming down to about 23.5% this quarter. Just curious what your thoughts are on kind of the long term take rate that you think this business can manage as you move to some of these other products such as FastLane? And then, just another follow-up on header bidding. Anything that you're seeing there from the client demand side on video and any implementation that you're looking to kind of roll out over '17 there, that would be great.

Great. On the take rate front, certainly, as we've mentioned before in the sort of short and medium-term, we do see continuing decreases. As mentioned, we expect that on a sequential basis, we should decline equal to, or perhaps greater than, what we've seen the last few quarters. From a long-term perspective, it's very challenging because there is just such a significant mix of different types of channels that we run through. So it will certainly decrease. And we don't really have the visibility at this point to know where that ends up, and so we'll play out this year and see where that leads us.

And on your header bidding for video question. So we're pretty excited about the opportunity there, particularly combined with our orders platform because, now, header bidding enables us to gain access to the most premium supplier impressions of the publisher. So if you connect that to our orders products where you've got higher CPM, more valuable deals occurring there, it opens up a new opportunity for us in video that didn't exist, just even a year ago.

The next questioner today is Jason Helfstein with Oppenheimer.

This is Alec filling in for Jason. You noted during your prepared remarks that you were going to look to leverage the $200-or-so million in cash on your balance sheet in your quest to reaccelerate growth. But the company was cash flow positive in 2016, so could you just maybe give us a little bit more color on how you're planning to use that cash?

Sure. Well, let me start with we're very committed to protecting our balance sheet. And so I'm not sure that we talked about significant initiatives using our cash, but we're very committed in protecting our balance sheet. At the same time, we do want to make sure that we make the appropriate investments in our products, our go-to-market strategies. We want to make the investments in our serving costs, for example, for ad requests, which we think can be a great competitive advantage over time. And so as we play out this year, one of the advantages that we do have is that strong balance sheet, and so we can continue to make those important investments even amidst some of this uncertainty and some of the challenging visibility. So we don't plan on expending significant amounts of cash but we have that in reserve in case that dynamic between our decreasing spend on our legacy desktop display doesn't stabilize as soon as we hope compared with the growth of our investment opportunities.

Andrew had a question before about M&A, and I think I might be getting, too. So we've grown this business organically and inorganically in the past. We've done 7 acquisitions now, both as a private and public company. We acquired a company a long time ago from Fox, which became the foundation of our cloud, which was pretty critical for RTB. We acquired a data and security company that was again very important for us to build a premium marketplace. We acquired 2 companies for orders. And clearly, we're seeing some great success with that. I mean, also, the foundation for mobile, while it was a small acquisition, it was a pretty significant acquisition, and we've built a tremendous business now in mobile from that. So the bar for us is really high, both from an engineering perspective and a people perspective, but if there are opportunities for us to advance our roadmap and better serve our customers, and if it makes financial and strategic sense, it's nice to have a strong balance sheet to evaluate those opportunities.

The next questioner today is Matthew Thornton with SunTrust.

Welcome aboard, Michael. I guess, maybe one for Frank, and one for David. Frank, you guys gave some good color on the take rate side of things. I guess, on the volume side of things, have you seen any precedent where, perhaps, your share was waning on a client, but as you've got FastLane and header bidding kind of implemented, and perhaps, maybe some of the video and mobile products kind of implemented where you've actually seen that share start to inflect and come back? I'm just curious if you've seen any precedent there? And then, similarly, when you think about clients, I guess, is client growth, still positive? Is the client count is still growing? And then, just secondly, for David, the cost rationalization program, I think, in total, was about $30 million annualized. Are we at that full run rate as we exit 1Q? Or is there still work to be done into 2Q? Any color there would be helpful.

Matthew, so look, I think, you nailed one of the most critical things that we saw in 2016 that we're seeing quickly turn around here from a supply perspective. So this is not necessarily the best analogy, but I'll give it a try. So pretend that there was a big storm in Manhattan, and let's say that Uber had all the Uber drivers in New Jersey, and let's say that, that storm cut off the bridge, so those Uber drivers couldn't get into Manhattan, right? So that's basically effectively cutting off the supply. Uber is not going to be able to generate revenue because there's a lack of supply of drivers that are in Manhattan, right? It doesn't mean that, that revenue is lost forever, but it can't monetize something if that supply access is not there, right? So what happened in North America desktop display with us in 2016 is that some of that supply was cut off and wasn't able to make its way to our platform. It's as though somebody poked a hole in a pipeline, and that oil wasn't actually able to get to our system. Once we deployed our own FastLane solution and as we're making those integrations, making sure that those integrations are clean and efficient and working well, we've reconnected that supply. Some of the stuff that we gave were -- we've seen now a 50%, 5-0, 50% increase in ad requests or impression access on a year-over-year basis, which has grown pretty rapidly. 300 new deployments, obviously, in header bidding that we just talked about, but 400 new deals overall. So to answer your question about client deals and client activity, new deals are being signed all the time. One critical point to make there is that because we're a marketplace business, as the supply in the past was -- or some of that supply was cut off, it took a while for that revenue decline to lag. And now, as it's being reconnected, it's got to, first, be made available in our marketplace, and then, it's available again for the bidders to bid on, and then, those marketplace effects start to reverse.

Great. Matthew, regarding the cost control initiatives, a vast majority of those initiatives and their impact will be complete by the end of this quarter. And so if you look at our Q1 expense run rate and normalize for the roughly $5 million in one-time or nonrecurring costs, that will give you a pretty good baseline for the rest of the year.

Sorry, 2 additional stats on your previous question. So in terms of client growth, the number of sellers, publishers and application developers grew by 172 on a year-over-year basis. And then, FastLane accounts who have header bidding installed, we're seeing 3x now the number of ad requests per account, per month, versus our traditional Smart Tag business, and that's why we're pretty excited about the growth prospects in the future, especially when you combine that with our orders platform where we're now accessing inventory, especially the most premium inventory that we weren't able to see before.

That's really helpful. Maybe just one quick follow-up, housekeeping for David if I could. I think, you've talked about tax rate a little bit. How should we be thinking about tax rates this year? And if you could just give us an update on the NOL balance, that would be very helpful.

Sure. We still have a significant portion of NOLs. And so from that perspective, we should have de minimis taxes, at least for the short and medium-term. The NOL balance for federal is, I think, $38 million.

Our next questioner is Sameet Sinha with B. Riley.

I want to start with a question for Mike. Mike, you've been at big companies, small companies, and you decided to join Rubicon. Are you -- the question is, in this ecosystem where we -- the larger companies are taking most of the share, what do you see in Rubicon that got you attracted to joining the company? And then, I have a couple of follow-ups for Frank and David.

Yes. Thanks, Sameet. Well, to me, I have been in this space for as long as I have, and being as close to the clients and customers, a continual refrain that you hear from them is a desired alternative to just the big guys who, in many times, are in varied [ph] competition with them, they're either outselling advertising. And if you're a buyer on our exchange, that's direct competition. And on the other hand, if they own inventory, they're outselling their own inventory and our publishers and application developers fight against that. And so the idea of this global, at scale, independent exchange resonates incredibly well with the customers and is more topical now than it ever has been. So that, to me, is the biggest driver. Obviously, I know the guys at Rubicon well. I've known Frank for years, respect the leadership, respect the team, really respect the product and the innovation that's coming downstream and just thought it was a perfect opportunity, both from a marketplace and from a company specifics to join Rubicon at this point.

Okay. Secondly, Frank, if you can just talk about header bidding. Obviously, you spoke about the deployments. And what do you still think from a technology perspective? Is that still a good technology? I mean, we hear about header bidding challenges, whether it's data center or server capacity that kind of soaks up all other challenges regarding fraud or even viruses. That's one question. And then, if you can also talk about just in terms of you've mentioned lack of visibility a couple of times, and that's going to be the reason for removing guidance going forward after the first quarter, but if you talk about the business about getting supply back, making more deeper penetration to current clients, would it not be helpful having broad guidance out there, so that we out here who have less visibility than you do can at least track the business as we go along?

Yes. Sameet, so header bidding from a tech perspective, so header bidding is a technology that clearly needs to operate at massive, massive scale. It's putting a lot of pressure on the DSPs as well. So making header bidding work and making it scale are 2 different things, and the latter is a much more complicated challenge than the former. So from a tech perspective, the client-side implementations, the technology that was initially used in the market, it was pretty primitive. There are some open-source technologies that we've embraced, and see us being able to take that and add onto it, a lot of the investments that have made us really successful in the past, things like our security technologies, our brand protection, safety. Things that have really enabled us to attract the most premium publishers in the world and create a safe marketplace and environment for the most important brands, advertiser brands in the world to buy. Those are the things that drove not only our success in the market to date, but also made programmatic and automation and real-time bidding, something that was not only a reality, but something that could be trusted within the entire industry. DSPs, agencies, advertisers, publishers themselves, et cetera. So us being able to bring those technologies into header bidding certainly make header bidding not only safer, more robust, but it gives us a differentiator, but hopefully will also solve a few bigger problems for the entire market, which will hopefully grow the market as a whole, which is something that we've always been very focused on. In terms of your data centers, serving capacities, I've talked about this in the past. We made investments from day 1 of this company to create a super highly efficient infrastructure. We knew that as an…

And Sameet, regarding visibility, certainly recognize what a challenge it is for analysts and investors to not have broader guidance from us. We have tried to share directionally some things that we think have some visibility into with regard to take rate declines. Certainly, from a cost perspective, holding those steady and CapEx levels. Beyond that, I guess, a couple of things, we certainly are committed to protecting our balance sheet, and so we will be mindful, as this year plays out, how that pacing and how that growth develops. And we'll be mindful of our cash burn. We are in the position to utilize some of that cash, as mentioned, to make the most important investments, which we will do. But we'll monitor that carefully, and we'll play out this year and try to directionally provide information where we can.

Yes. And some additional color to that as well. Look, as I said multiple times in my script, our #1 priority this year is capturing market share of supply and try to illustrate how that's been a successful strategy for us in the past. So if we need to balance or sacrifice some short-term revenue to bolster that strategic position based on history, that's a smart thing for us to do. And we just want to make sure that we leave ourselves some room and flexibility to make sure that we're truly optimizing for that, and then we're making sure that both our business model and pricing models are reflecting our ability to capture market share and make sure that this company is best positioned strategically. We think that's a smart thing to do, and we want to make sure that we are leaving ourselves some room to do that versus trying to tie ourselves to some kind of strict guidance, at least in the short-term.

The next questioner today is Jason Kreyer with Craig-Hallum.

Michael, welcome to the call. Frank, you mentioned several times, and just piggybacking on the last comment, the effort to capture additional supply as you move forward as a marketplace-focused business, how does that sale change trying to capture supply going forward from here versus your success capturing supply in the RTB business, capturing supply in the orders business? How does that shift from here on out?

The strategy is actually pretty similar. When we first launched our very first product, ad network optimization, similar strategy with real-time bidding, similar strategy with mobile, video, orders, and similar with the header bidding. We first go to our existing customer base. It -- I don't want to say it's an easy add-on, but it's an easier add-on because we say, "Look, you already have desktop and mobile. Hey, why don't you turn on this additional feature on as well?" We're also -- having the position to be able to leverage existing buyers, whether that be DSPs or advertisers and agencies that are using our orders product. So first, it's about going to our existing customer base. Sometimes, we're able to use new products to enter into new customers that aren't using existing products as an entry point, so for example, when we just had real-time bidding, we introduced orders into the market, some of the most premium brands in the world where they may be a little sensitive to auctioning their inventory, so we were able to enter with the orders product. They get comfortable with that, we establish a relationship with them, and then we're able to then go back and upsell them the real-time bidding product, or now the FastLane product, et cetera. So it's -- having multiple -- having a portfolio of multiple products helps us to either leverage existing customers or bring new products to new customers and then continue to upsell them.

Okay. And then, the Q1 guidance implies more significant deceleration. And just wondering if you can walk through the factors that are impacting Q1 more strongly than what has impacted Q4?

I think, it's just -- it's a continued deceleration of our legacy mobile desktop display. And the growth areas while we're encouraged by the momentum that we have there, they are not -- we haven't hit that tipping point quite yet, and that's also the foundation for some of the lack of visibility that we have going forward this year as it regards guidance. And Frank, I don't know if there's anything you want to add to that.

I think, just another comment on the guidance pieces. We have a new CEO that's joining here. So giving him the opportunity to come in, take a look around, really dig in, galvanize some of our strategies. I think, it's a fair thing to do to give him some space and time to do that.

The next questioner today is Mark Kelley with Citigroup.

Two quick ones. The first one, you talked about that 50% growth in ad requests in your prepared remarks. I just want to clarify, is that a 4Q '16 over '15 number, or is that full year? And then, second, any thoughts on the SoundCloud partnership? I know it's very, very early, but anything you could provide there would be very helpful.

Yes, the 50% increase in impressions was full year. For SoundCloud, so it's not launched yet, so too early to give any specific comments around that.

All right, fair. And then, one more if I could. Last quarter, you called out a lot of smaller solutions that maybe gained some traction, header bidding early on, and that caused some of the disruption. Has the shift to server-side header bidding happened? Obviously, scale will matter. Are you seeing some of the smaller guys already go away? Or do you think it'll take something like server-side to make that happen?

Yes. Look, go away is probably not a term that's used in ad tech very much. There's always ebbs and flows. Look, we've been at this now for 10 years. There's a reason that there's only a handful of exchanges and only a couple of really large ones, us being one of them. The shift to server-side, look, we have a server-side product, our xAPI, we've been in the market with this for, I think, it's 1.5 years now with our cloud competing infrastructure. We are certainly extremely well-positioned there. So I think, we're pretty bullish if that's the way the market goes. I mean, part of that is effectively what we've been doing for 10 years. So with that said, there are some challenges in server-to-server header bidding. Without getting into too much detail, there's some cookie translation issues where you might become difficult for some of the buyers to be able to do some of their retargeting or some of their just general targeting and things like that. So we're excited if the market goes there, we're prime positioned for it. Again, we've been in market with a server-to-server product, our xAPI. In terms of what that does to smaller solutions, it's hard to tell exactly without their smaller solutions, but we've clearly made a lot of progress with our own FastLane header bidding solution, just even on the client side. With that 50% increase in ad impressions here overall, the 300 deployments, you have to imagine that we are doing some damage to those smaller competitors, but it's a huge market, it's a $600 billion market, there's plenty of dollars to go around. I think, we're more focused on trying to grow the market for everybody versus trying to necessarily do away with any particular smaller competitor.

Looks like we have no further questions, so this will conclude our question-and-answer session and also today's conference call. Thank you, all, for attending today's presentation and you may now disconnect your lines.