A version of this article by Tim Chen originally appeared on The Harbinger China, a community that leverages a network of China’s top VCs and tech companies to generate exclusive, informative deep dive interviews for a global audience.
Friend of TechNode Tim Chen recently interviewed Dongtu co-founder Grant Long about the Chinese GIF company’s monetization model. Akin to the US platform Giphy and Tenor (part of Google since an acquisition last year), Dongtu plugs into existing messaging apps—including all of China’s major chat platforms—to deliver a searchable database of moving images. Long describes how the company makes money off GIFs, widely used to distribute rough-hewn meme humor. TechNode’s version of the interview has been edited for length.
Harbinger: In the US it feels like advertisers don’t understand GIFs. Is it easier in China to sell to advertisers, and do they immediately understand the value of GIFs as an ad format?
Long: I would say that it’s definitely not easier here, and for the entire industry spanning from some of the Western companies to ourselves, we’re still in the educational phase with brands and advertisers of why this is really going to help drive business. But the usage of stickers and GIFs is a universal experience in China versus the rest of the world where it may appeal more to the younger demographics. I think that ultimately a lot of business decision makers here recognize that the universal aspect of GIFs here is valuable, and can see the value in connecting their brand to the experience.
I think when you talk about ad adoption, one important factor is creating a kind of standard that people can use for everything from creative to measurement of success, and one thing that I’ll point to is early last year when Google acquired Tenor, a large global GIF company. And I think when you talk about driving standards and things into the digital media environment, Google is the largest digital advertising company in the world, so they have a significant ability to push things forward. We are in a great position here in China where because of localization challenges as well as things like government regulation of the internet, we feel very isolated from competition here, and we think actions happening in the West will help the overall industry.
Harbinger:Chinese internet ecosystems definitely feel like a different world from the West. So you’re still in the educational phase with branded ads and you’re looking to set up standards. You recently partnered with NHL (National Hockey League) and PSG (Paris Saint Germain Football Club), both of which are foreign brands, to help them target their Chinese fans. What were some considerations in terms of the partnership? How are they able to measure the effectiveness of this? What was the sell to bring them on board?
Long: China, despite the fact of the brewing tensions of the trade war, is still recognized by companies around the world as a key growth market for them, and that’s particularly prevalent in media, entertainment, and sports, where what they are selling is not a physical product, but is digital and cultural. When they look at our platform, they see a huge benefit in partnering with Dongtu, because although China is a big opportunity, it’s also very challenging to have a presence across the entirety of the Chinese digital landscape. So many brands may have a presence on Weibo or Wechat with an official account, or more recently with the Bytedance companies and products like Douyin. But with Dongtu, they see an ability to work with one channel to reach users across thousands of apps and to start seeding their content organically into those apps, and especially if you’re some kind of brand that looks to be culturally relevant, such as someone in sports, media, entertainment, but also big Fortune 500 consumer brands that want to have a share of voice and have their brands be recognized. It starts with just impressions and driving engagements across these different channels that we work with, but down the road, things that we’re working on that are already live on a number of our partner apps are the ability to click through to different experiences, so if there’s a sporting event or a music event then maybe clicking through to a ticketing page or a more information page or different parts of the app.
Harbinger: Help us understand the GIF business model. For Dongtu, can you share more about how you currently monetize, how you measure engagement as well as what matters in terms of engagement?
Long: For Dongtu, there’s definitely a lot of overlap in some of the ways we look at monetization and how Giphy and Tenor do. So, at its core, it’s driving engagement for our clients on our platform, the way we measure that would be through how many of the pieces of content they’re actually serving into user experiences.
I think in the West, Giphy and Tenor have primarily API driven businesses, meaning they don’t necessarily have a native SDK in their partner apps but it’s a communication between the two apps. The easiest information to gather is from their CDN—how many of these images they’re serving into a user’s app. And there’s certain signals that can be captured through those integrations, but it’s not necessarily rich with data. For us, because the vast majority of our integrations are through our SDK, we capture a lot more data both around the user information as well as was a GIF not only viewed but was it sent in a conversation. So we know who was the person that received that content. Additionally, we can see all these interactions and who’s involved and therefore we build this web of social networks and social interactions across thousands of apps in China. We have a pretty compelling data set from these device IDs and IP addresses too, and with pure data and insights we can try to leverage this into generating revenue.
At its most basic form and in the earliest days in this industry, a GIF is a file, and has been around for decades. It’s ultimately not that great of a file type: it is large, has low resolution, but it does have inherent properties of autoplay and supportive environments that made it appealing, so we’re actually pushing through our platform because of our SDK and the fact that we can control rendering in many of our apps, more towards a GIF as actually rich media. So in a second generation a GIF may actually be an mp4 or a piece of video content, that can be longer, better resolution and smaller file size. But then, the later iteration of that would be to interact with content—can a user click on a DONGTU and open up a webview that’s triggered from the experience? That’s where we see a lot of interesting overlaps and synergies between one of the huge trends in China with mini programs.
Harbinger: You mentioned engagement quite a bit, for brands not familiar with a GIF ad format, if they’re thinking about advertising through mini-programs, when you say rich format do they think of this as you guys are similar to a TV-channel where within the third party app, when you key up GIFs to send, they have the real estate up top where it’s relevant to users? So they can have their stickers “air” at a time that match their tv show airs. Is that how brands think about GIF ad formats within your ecosystem?
Long: So, earlier in our conversation I talked about how we represent two very important things inside communication, one is search and one is media. When we talk to brands, we actually have commercial models and pricing models that relate to both of those more traditional digital business areas, and so if a user opens up a Dongtu experience in Taobao for example, they’ll see trending content immediately that is most relevant for them based off of algorithms and editorial efforts from our creative team, and because of the convenience factor, many of those GIFs are actually the ones that are selected to be sent.
So there’s a lot more to it than just seeing content, since there’s that interactivity of sharing that feels like brand endorsement.
There’s another model that’s more similar to the search business: for example if McDonald’s wanted to target a term like “happiness” or “breakfast,” we can take their content and place it to target different search terms for contextual targeting. So this McDonald’s example can be used to show how powerful GIF advertising is: if someone is texting their friend in the morning telling them their hungry, and we place the McDonald’s GIFs in that texting environment, we’re relevant in the right time and context – maybe to even plant a seed in the user’s head to possibly go to McDonald’s.
Harbinger: As you grow, what do you look for in a strategic partner or investor?
Long: In terms of strategy investment, in China, many large companies chose to fundraise from BAT (Baidu, Alibaba, Tencent). But that’s something we need to tread carefully. We work with many of them already and it’s a risk to us to show favoritism because that could exclude us from the other set of services or ecosystems. If we raise from Alibaba then Tencent could say you’re an enemy now. That’s important to consider in China. If we can raise from multiple players that would be ideal. In terms of other types of strategic investment, we have raised from Cheetah Mobile in our Series A – they’re relatively neutral when it comes to big players. We’ve considered investments from media entities in China. Another factor in China is the government, like many startups in tech we do have lots of early fundraising from government funds and we’re doing some exciting things with state-owned media on how we can help them better communicate with their readers and audiences using Dongtu and rich media. So that’s maybe more unique to a Chinese company than internationally. In terms of international strategic partners, we’re always aware of how closely guarded the government is with any large scale media entity on the internet in China. There’s a reason why Facebook and Google aren’t operating here so that’s something we’ll have to weigh if they’re interested in investing in us.