Valnet’s Ji Heon Kim on how the publisher encouraged users to authenticate themselves
With websites covering topics like entertainment (ScreenRant), gaming (Polygon) and automotive (CarBuzz), Valnet caters to users across a wide array of interests.But according to Ji Heon Kim, Valnet’s head of monetization, Valnet realized it could create more value for its users by encouraging them to subscribe or authenticate themselves.Maybe a “mass scale” of users wouldn’t sign up for their websites, but perhaps 10% would. And, as Kim puts it, that “10% would still be valuable, and we can do a lot with that 10%.”“We created more value to [those] users, more exclusive content and high-quality content,” Kim says. “All of that became an initiative on the content side for us to deliver a premium model and give users an incentive to sign up.”Kim further talked with The Current Podcast about balancing advertiser value, user experience and performance, which he says are “always affecting each other.” Episode TranscriptPlease note, this transcript may contain minor inconsistencies compared to the episode audio.Damian Fowler (00:00):I'm Damian Fowler, and welcome to The Current Podcast. Today we're talking to one of the biggest digital publishers. You might not know by name, but you've definitely read their stuff. I'm talking about Net. The company behind Screen Rant, the Gamer, Kaleida make use of, and a bunch of other sites that rack up hundreds of millions of sessions every month. Joining me today is Ji Kim Valnet's, head of monetization. Ji'S been leading the charge on everything from supply path optimization to first party data to figuring out how to drive real revenue without compromising the reader experience. We'll get into some of the big shifts they've made in their tech stack and how they're bringing newly acquired brands like Polygon into their ecosystem and what other publishers can learn from their approach.Ji Kim (00:52):At Valnet, I'd like to think of us as a publishing powerhouse. We started very small. Our motto is humble and hungry. We like to remind ourselves that it's always good to keep a humble mindset. I've been at NET for 10 years and we've grown tremendously. We've went through a lot ups and downs, but even as we grow, we like to think that we're small and agile and the publications we range from automotive, gaming, technology, entertainment, but entertainment has always been our flagship, but we've been kind of branching outside of that and trying to expand more and more. And then we have some lifestyle brands as well as sports.Damian Fowler (01:35):Let's talk about a moment that changed the game for Net. Can you walk us through your, I guess we're going to talk about supply path optimization at first anyway, which is a hot topic around these parts and what work you did around supply path optimization, like cutting resellers and boosting direct inventory. Could you talk us through that a little?Ji Kim (01:57):It's an ongoing process. It's certainly, I think most people agree that SPO is not an easy thing to achieve. You can commit to it one shot, but that's much harder to do considering that there will be a revenue impact. So for us, we tried both ways. We took a few sites and we took the direct approach and we saw a pretty decent stability, and then some other sites did not, and then we have to kind of revert back to it. SPO, it was always a topic that was talked about but not well enforced. And tradedesk took a big initiative to push publishers towards it. And then we started working closely with Jounce Media as well, with Chris Kane started kind of talking through some of the ideas, how should we go about it? How do we retain the value and still achieve removing the resale alliance and keep our inventory as clean as possible?(02:51):But initially our outlook of SPO was about making our inventory as clean and transparent as possible. Net considers ourselves as a premium publisher and we want to make sure that the advertisers see that as well. So we were heading in that direction. But ultimately, I think the biggest challenge with SPO was it's impossible to do an AB test because you have one A TXT file and you can't test one setup with the resell alliance, one setup without. So that's been pretty challenging to understand where's the value going, where is it coming from? And even with the Resell Alliance, when you talk to the SSPs with Resell Alliance, they'll go, oh, these are PP deals. These are not just rebroadcasting and all this stuff. So trying to understand the granularity and all that details of what each resale align means was very difficult. But ultimately we know we have to go in that direction, but we know it's not going to happen overnight, so we're kind of just taking a step at a time.Damian Fowler (03:51):That's great. What would you say was the kind of catalyst or moment that sparked that shift?Ji Kim (03:57):We always talked about advertiser value. It is important to yield as much value as possible and get the performance that we need. We always think that advertiser value is important, and when we think about that, it's like you go through stages. You go, okay, viewability needs to be important. Let's get viewability up to above standard, above average, make sure our CTR is good, but it's high quality clicks. It's not just users just clicking on stuff. Then you go through the lines and eventually you get to SPOs. Make sure that advertisers know what inventory they're getting access to, what they're buying, and make sure that they're getting insights. The transparency is there. Then we've increased the value of our inventory.Damian Fowler (04:46):Yeah, I mean that's the key, right, obviously. And speaking of that, having made these changes, are you in a position to be able to see the kind of impact that they've had from a revenueJi Kim (04:58):Perspective? Honestly, I don't think I can everything, especially with these kinds of stuff, what I've learned is it doesn't change overnight. Let's say we remove all the reseller lines yesterday. Today, likely the performance is going to drop initially and maybe things recover over time, but there's so many moving parts that it's hard to associate the value towards SPO, and that's a lot of things that we do in this industry. But I think that's when we like to look at it as, you know what? Ultimately we are improving the quality of our inventory, so we will get rewarded at some point. And that's how you move forward. But with SPO, I think the other side is that it's not just about removing reseller lines. You also have to market yourself and tell the advertisers that, Hey, we have gone in this direction. We have removed the reseller lines. All of our inventory is direct. It's clean. And that part is also hard to do. We haven't spent a lot of time or resources into marketing ourselves, and that's why we talked about, people may not know net, but they know our brands. It's the same thing. It's like we are now making a big push to let people know who Val net is, and that's going to go in hand in hand with this stuff.Damian Fowler (06:21):In terms of that messaging around the surgery as it were you're doing on the supply path, does that land well with advertisers?Ji Kim (06:32):I think it's always positively looked at when you tell them, it's like everybody, it is never negative, but I don't know if actually if it's meaningful for them because at scale, they're buying at scale. So yeah, we're a big publisher, but they're also buying at multiple publishers. Maybe only small portion of their budgets come to us. So it's positive, but I don't know if it's all that meaningful to them. At least that's what I've felt.Damian Fowler (07:04):So in addition to the SPO, what other tweaks or changes are you as head of monetization looking at to basically bring in those ad dollars and keep readers satisfied, I suppose?Ji Kim (07:17):Yeah, so there's three things. So we looked at the advertiser value, but then there's the user experience and then the performance side. So always those three things, there's constantly affecting each other. Ad density is probably one of the biggest part of advertiser value and performance and user experience. So we are constantly trying to reduce our density, and we look at this metric impressions per session and request per session. So we look at that and injections our injections based on content length, a paragraph breaks and all that stuff. So we'll try to work with the content team to create optimal breaks. I'll have a little sit down session with the content team. The leads say, okay, this is how the admin injection works, and how you break out your content really does impact, because we won't break a paragraph in half to inject an ad. So there needs to be natural breaks for the ads to inject. So if you have massive paragraphs, we're going to have less ad injections, which is fine if the content works like that, but they also need to think about how all this stuff works.Damian Fowler (08:26):That's really interesting. I mean, I think that sweet spot between not being the Vegas strip, but also ads have to populate at the right time to have value.Ji Kim (08:35):For net, we've focused mostly on open market programmatic spend. We have a small direct initiative. This is something that we've been trying to grow, but when you don't have huge direct sales initiative and direct spend coming in, you kind of need the density because the CPMs that you get from open market is much lower. So we want to try to move away from that as much as possible. I don't think found that will ever be a publisher where we drive like 50% of the revenue from direct sales, but we want to grow it to maybe 15, 20%. And once we do that, we can yield higher CPMs, which allows us to reduce the density, which would be better for advertiser value, better for user experience, and we'll still get the performance that we need to kind of go forward.Damian Fowler (09:24):So it's a balance.Ji Kim (09:25):Yeah. Yeah. I think if we can drive higher CPMs, we would love to reduce density, but it's always the constant battle between the two of, okay, well we reduced density. Oh, we went too far. Okay, we got to bring it back a little bit.Damian Fowler (09:38):How difficult is it to kind of innovate in ad tech? This is a broader question, I guess given how fast things are changing, especially on the programmatic front,Ji Kim (09:47):It's been very, very difficult. Rapidly changing environment is definitely one of them, and you have to adapt quickly. For example, the video definition of having instream outstream, and then now there's a third definition of accompanying that stuff. When it happened, the enforcement happened quickly, so we had to adapt quickly, and that's difficult. But innovating is, I think, much more difficult than just adapting to the new policies and new rules. So many different ways to innovate pre, for example, you have the open source code, you build that, but there's so many customizations that you can do and even a single customization, you interpret how you should approach that topic and how you should build your tech. So you kind of have to talk to your developers and walk through. And our biggest challenge I would say was bridging the gap between developers and ad ops. I was like, because I am an ad ops guy, I understand programmatic landscape very well, but our developers do not. And I'm not a developer, I'm not a technical guy. Obviously through 10 years I've learned a lot of stuff, but still, if I needed to build something, I'm not going to be able to tell them exactly how to build it. So you need somebody in the middle that understands both sides,(11:03):And that was the most difficult part. And eventually we did find resources that they were able to bridge that gap and were able to build stuff. But ultimately, there's just so many different ways to build your product and you want to make sure that product that you build or tech stack that you build is going to keep that balance that you need between the user experience, the performance, and the density, everything that pertains to page speed as well. If you build it to be too slow, everything gets affected as well, and that's harder to tell. So yeah.Damian Fowler (11:37):So how have some of these technical changes influenced your broad and monetization philosophy?Ji Kim (11:43):Yeah, so I guess one of the things, if we talk about authentication, we talk about cookie deprecation and why authentication became so important to majority of the publishers. And I remember our thought process around authentication was pretty pessimistic, I would say. But eventually we said know what? We can create content or value for the users that's going to want them to sign up and want them to get authenticated. And we said we got to start somewhere. Ultimately, maybe we've become a little bit more realistic about what critical mass of a value would be if we're at, if we're expecting 50% of users will log in, that's not going to happen, but 10% is still very meaningful. So it was about our philosophy was changing, about our expectations changing and still understanding that 10% could be very valuable and we can do a lot with that 10%. So we created more value to the users are more exclusive content, high quality content, high quality videos. All of that stuff became an initiative on the content side for us to deliver the premium model and to give users the incentive to authenticate a sign up on.Damian Fowler (13:03):That's really interesting. I think one of the things that also I'm hearing is that you kind of have different audiences, but you're getting to understand your audiences. I mean, this strategy gives you more insight into who's coming.Ji Kim (13:15):Yeah. We also created what we call threads. They can talk about the article, talk about topics that we're discussing, and that really improved our engagement.Damian Fowler (13:30):As you look to the future, how do you think about, as it were, locking in some of these changes and this value that you see from this audience?Ji Kim (13:40):So I want to go back a little bit about innovating and how difficult it is. So I went through the stages of, okay, what am I focusing on to optimize to yield more value? And initially it was demand. Okay, we want to work with as many high quality as P as possible, but then you do work with all of them. There are going to be going to be one or two that come here and there, but generally speaking, they're not going to create incremental value. They'll just take a piece of pie that was taken by somebody else, not meaningful value. Then you work on ad tech innovation, all that stuff, and that we'll continuously work on that, but that also has lots of limitations, and you eventually reach a plateau point of say, you're not going to find a lot low hanging fruits. So now we come to premium inventory, which we need to learn our users, we need to learn who they are so we can offer these users to our advertisers to grow our PMP programmatic direct, as well as your conventional IO based direct deals that's going to yield as higher CPMs.Damian Fowler (14:53):Yeah, I mean, talk of premium inventories is characteristic of the moment we are in when it comes to programmatic sales for publishers.Ji Kim (15:02):Yeah.Damian Fowler (15:04):Let's draw back and look at the big picture and some of the kind of industry context. I guess think I'm correct in saying Valnet reach has more than 400 million sessions a month across its network. That's correct. And how do you think about that, that kind of scale when every property has its own audience profile and publishing rhythm?Ji Kim (15:30):Yeah, it's sometimes a bit overwhelming how much reach our sites have, but I always try to look at it as our advantage, and this is the opportunity that hasn't been tapped into, is that okay, we're 95% of our inventory is sold in the open market, and we have so much data that we could collect and leverage in order to drive higher value. And it's just looking at it, it's overwhelming, but you start to see the real value that hasn't been tapped into, and that's exciting, but it's also very, very difficult to manage all that information, manage that data, and use it properly. So yeah, I mean it excites me, but also I know how challenging it can be to create value through that. So we're taking one step at a time, even first party data collection. I wouldn't say we're crazy sophisticated, but we're keeping it a level that we know how to manage and understanding it well first and then starting to kind of grow a step-by-step.Damian Fowler (16:45):Yeah, I mean, I suppose the whole back and forth about third party cookies may have provided a spark. I know it lit a fire under the industry. Speaking of first party data, so that is a focus for you?Ji Kim (16:56):Yes, yes. But I believe when it was really a huge focus for the industry was when Google had first announced that they're going to deprecate third party cookies, and we had the initial moment of, oh, you know what? We also need to look into this, but we didn't want to panic. Our outlook was, I'm sure everybody went through the initial panic. We did too, but we didn't want to stay in that moment. And we said, okay, what's realistically going to happen for publishers like us? How much first party data can we collect and really sell because we don't have a huge direct sales initiative? And at that point we had none. And you can't grow direct sales overnight. It's a highly competitive environment, and you're entering that new market. You have to build relationships, you have to have crazy amount of salespeople that are constantly going out there representing balance inventory.(17:55):And we weren't set up for that, and we weren't willing to just fully invest everything into growing that at the time. So we said, well, maybe first party data isn't as important. Collecting first part data isn't as important as just understanding how to go about direct sales. So that's what we worked on. We've hired salespeople, we enter that space. I was very naive about how direct sales worked, and now we have a better understanding. We have good salespeople that understand our values as well. We don't want to just go out and sell anything and everything. We want to understand the creative types that we're also selling isn't going to impact user experience horribly and negatively. The high impact guys, the site scans when they're done, right, it's great user experience, but it could also go the other way. So we wanted to build a baseline first, and that's what we did the last few years. And now we can go after the first party data in a more sustainable way for us.Damian Fowler (18:56):Let's talk about your acquisition of Polygon from Vox Media. Speaking of inventory that expands the real estate, how does that property fit into what you're doing?Ji Kim (19:07):So Polygon, obviously, we go through a lot of due diligences. We look at different opportunities, and Polygon was an easy one to go through because we knew Polygon has great content, it has a great foundation of creating high quality content. But the difference was that Fox has a lot of direct sales. I can't remember the exact number, but it could have been 75%, 80% of their revenue was generated, direct sold inventory, and then 20% was open market. And for us, it would've been the other way around, flipped around even less. Maybe 95% open market, 5% directive. Initially when we acquired it would've been a hundred percent open market, but that's also why it excite us because it's a premium inventory that doesn't get seen in the open market. Open market buyers don't see the bid requests coming from that website as much. So we're super happy, but we knew this was a high quality inventory, high quality website, and we knew that there was a very small chance that it was going to go poorly.Damian Fowler (20:20):Interesting. When you buy a property like that, you're actually buying an audience to a certain extent.Ji Kim (20:25):Yeah, absolutely.Damian Fowler (20:27):Do you think about audiences as discreet to the publications or do you see crossover?Ji Kim (20:34):Crossover? Yeah, lots of crossover.Damian Fowler (20:37):Yeah. Alright. So I guess the big question here is for other publishers looking to upgrade this strategy that we're talking about, especially in this very complex environment, which is something you clearly understand very deeply, what's one piece of advice that you might offer?Ji Kim (20:54):I think you have to think about realistically what you should go after, what opportunities you should go after. So many things that come up right now, I think the big thing is curated media. And on our end, a lot of the SSPs and DSPs are doing the work for us. They going out and curating our inventory for us, and that's fine. But if you were to go after that and trying to grow it, but you don't really have the resources, it's easy to just kind of see everybody, what everyone else is doing, like, oh, I want a piece of that too, but it's not going to yield the value. Same value if you don't have the right resources in place if you're not focused on that opportunity. So my advice would be to understand which opportunities realistically are you able to get and have the right resources who are going to be passionate about that. Take accountability. That's huge, the accountability part. And that's not something you can just kind of force people. You have to believe that this person that's taking on this project can be really passionate and sink their teeth into it. If you got that, then go after those things. But it's too hard to go after every single opportunity there is. Even if seemingly it seems like a low hanging fruit. Nothing is really that simple in this industry.Damian Fowler (22:15):That's for sure. So finally, we're going to wrap this up with some what we call hot seat questions. So what's one thing you're obsessed with figuring out right now?Ji Kim (22:27):How to yield more value? No, no, no. I'll give a better answer than that right now. For me, it's how to grow direct sales sustainably and scale it in a way that we don't get too bloated. Because through acquisitions, one of the most valuable things that I get is insight. I get to see under the hood of a lot of publishers, small to medium to large, how they operate, what is their strategy and direct sales. I've learned some of the big publishers do it extremely well. It's a well-oiled machine, it's not bloated. They generate a ton of revenue, but some have a huge cost, and that's what we were afraid of. And right now it's very hard to do. So you need the right sales team, you need the right operational guys, you need account representation, you need reporting guide and all this stuff. And right now I am trying to find a way to scale it, but without having massive costs, just kind of take over and then expect this to yield value in the next year or two. I want that line to kind of grow together. And that's not an easy thing to do, obviously. And I'm looking for the right resources. I'm looking to build relationships with agencies with limited guys, just hustle through it and offer them our inventory, charm them, whatever it may take. But yeah, that's what I'm currently obsessed.Damian Fowler (24:01):Okay. What's still missing in the ad tech stack that you wish someone would build?Ji Kim (24:07):I don't know if this would fall under their ad tech stack, but I think we could really benefit from a bit more standardization around, it could be reporting and creatives. Maybe I'm speaking out of line because I'm on the inventory side, so I don't know everything that goes on the buy side and the creative side. But what I see is that there's so many different creatives that just either break the page, the creative's broken, it's too heavy, it slows down the page, and it's hard to target those and remove those. It can come through so many different channels. So if there is a bit more standardization around what kind of creatives are acceptable, I'm sure there is some or a standard already, but it needs to be honed in a bit more maybe.Damian Fowler (25:00):What's one thing advertisers misunderstand about monetizing Publish it inventory today?Ji Kim (25:08):So I thought about this and something that it's more of my frustration around advertisers perspective. I understand it, but a bit more frustration because it's hard to create context around it, which is brand safety. I understand the brand side. I advertise side on why they wouldn't want to associate their brand with certain content, but brand safety is police by keyword list and it's very restrictive. And some of the,Damian Fowler (25:37):It's one toolJi Kim (25:38):And it's like, okay, and we have gaming sites that will, a lot of gaming, natural will talk about shooting, but some of the game developers won't want to associate with those articles. And it's like, hang on, hang on. Now you bet you guys also have games that are first person shooter or whatnot. You don't want to associate with those type of articles. There's a bit of a mismatch, and I think it's just hard to manage that. So they go with a broader approach and I get it, but I think it's just there needs to be more about understanding the context of certain articles. And it's like the word shooting can be anything, everything. Right?Damian Fowler (26:22):Yeah, I like that. I've been hearing more about a shift from brand safety to brand suitability, which brings in the concept of context. What's something unexpected you've learned from reader data or behavior recently?Ji Kim (26:39):So I wouldn't say it's recent, but it's something that's surprises me how the smallest change that I, from my perspective is like, is that really going to do anything? But at our scale, the numbers changed so drastically. Recently we were playing around with the video size because our outstream unit will float once the user are scrolling and the size of that unit. Obviously we want to give advertiser value, so we want to make it as big as possible. But then user experience wise, it could be very bothersome because as they're trying to read, there's a video playing. So we want to keep mindful of that. And we're constantly testing the size of that unit and we decreased by 10% and 10%. While it's significant, if you look at the actual size of the unit to the naked eye, you really wouldn't be able to tell what the difference is. But the CTR of that video unit changed drastically. It was cut in half, actually. And that's the thing is like, okay, users are really sensitive to these things. And to me it's not, maybe I'm looking at it too often, but that's always, that boggles my mind and it always catches me by surprise when I see the numbers is like, wow, I did not expect that. I did not expect users to behave this way.Damian Fowler (28:00):That's amazing. The details really matter.Ji Kim (28:02):Yeah, Big time. Damian Fowler (28:03):And that's it for this edition of The Current Podcast. We'll be back next week. The Current Podcast is produced by Molten Hart. A theme is by Love and Caliber, and our associate producer is Sydney Cairns. And remember,Ji Kim (28:21):I like to think of us as a publishing powerhouse. We started very small. Our motto is humble and hungry. We like to remind ourselves that it's always good to keep a humble mindset.Damian Fowler (28:34):I'm Damian, and we'll see you next time. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
