Wollo’s Experts Interview: Technology Trends and Insights with Nick Morris
Third-party cookie depreciation, reliable tracking setup, AI and machine learning, automation - Topics that concern the performance marketing industry.
Through the "Wollo's Experts Interviews", our Director Communication DACH, Wolfgang Scherer (aka Wollo), spoke to Nick Morris, VP Product Management at Tradedoubler, about these topics. Among other things, the two clarify which tracking setup is safest for advertisers, what role probabilistic tracking will play in the future, and what is already possible today in our Tradedoubler platforms with the help of APIs.
About
Could you please briefly introduce yourself?
Hi, I’m Nick Morris, and I’m VP Technical Product Management at Tradedoubler. I’m responsible for Tradedoubler’s backend systems; ad-serving and tracking, our core platforms, partner management, commission attribution and billing systems, and the APIs that we use to power the frontend – so, essentially everything that is not external facing.
APIs
You already mentioned our platforms. Tradedoubler has been working with an open platform concept for a few years now, and in 2022, we migrated all our platforms to a globally distributed cloud architecture. Can you tell us more about the open platform concept and its benefits?
Our frontends are fully running on APIs. We built the whole Tradedoubler platform API-first. This means anything in the interfaces could also be managed via API, which gives our partners the possibility to automate these tasks. When we started building the open platform concept, our vision was to naturally sit as a middle point between publishers and advertisers, as a reliable, neutral party.
A great benefit of the new platform architecture is that we can roll up updates and new features faster. So, if you add a service for the publisher APIs, this service also becomes available for our private network (PVN). Likewise, for changes in the advertiser interface. Before we had the open platform concept, we did development work multiple times for multiple places. Now we are much faster to market with new features.
When we talk about the APIs our partners can use, do you have an estimate of how many of our functionalities they can already access via an API?
It’s 70 to 80 percent. Clients can already do a lot of tasks via an API: On the publisher side, for example, they can automate applying to new programs, getting a list of new programs, creating new sites, getting ads and vouchers, and creating claims – all of that is already possible. Likewise for advertisers: They can use our APIs to put ads in the system, do all the reporting, validate sales, manage claims, approve publisher applications, invite publishers to the program, and see all publishers in the network. These are only a few examples of what is possible; interested advertisers and publishers can find further information on our APIs in our Developer Documentation Portal.
Is there any development in new APIs or functionalities planned right now?
We are constantly improving on the API standards we have. Besides planned internal projects, we are looking at increasing the ability of advertisers and publishers to communicate with each other, for example, by making it easier for advertisers to send newsletters.
Tracking
From a technical or product development perspective, what are the biggest challenges in performance marketing in the next 24 months?
There are short-term and long-term challenges: One short-term topic is obviously Chrome’s depreciation of third-party cookies. Although most of our tracking integrations are first-party, it is important for us to keep an eye on that and develop our tracking proactively.
On the longer-term side, historically, tracking has been very deterministic, but we are going into a phase where we will have to invest more in probabilistic approaches. At Tradedoubler, we already have our own, inhouse developed, probabilistic tracking solution, so we don’t rely on other services. The benefit of this is we know exactly how it works and can develop it in the direction we need – with a focus on affiliate marketing. In addition, we have a lot of experience in tracking, and we have a vast number of data points generated in our network that we match in-house and use to test and validate our tracking models. This immense database also helps us in training our own machine learning models to improve tracking.
Reliable tracking is crucial for performance marketing. Do you have any recommendations on what tracking setup is currently the best and safest for advertisers?
The most robust tracking is a server-to-server integration. You are not beholden to any changes that Apple or Google make since your connection is from your server to our server, and there is no third-party in-between. But it is also the most technically involved setup because you need someone on your side who can manage a connection from your server to ours, thus you need a developer or engineer who does that. It needs proactive development to keep aligned with the affiliate industry. It also means there are no probabilistic/cookie-less fallbacks available.
This is why we’re mainly pushing our Tradedoubler Tracking Library as the most reliable option. The Tracking Library is a script integrated into the advertiser’s website. We can push updates automatically through the Tracking Library to quickly adapt to new tracking and privacy regulations. Furthermore, using the Tracking Library, we can better support product-level tracking, manage cookie and tracking consent, and activate fallback solutions if our standard tracking is not firing, e.g., due to ad blockers. Adding the Tracking Library is a one-time integration that allows us to manage first-party tracking automatically and switch on container tags for accepted tech partners if the advertiser permits us to. The advantage of the Tracking Library is that advertisers always have the latest version of our tracking, and tracking audits or code changes from advertiser side are not needed anymore.
Product Development
Can you give us a sneak peek at what product development projects are currently planned for the future?
There are many more ideas and concepts we are working on right now but only to name a few: We are heavily invested in AI and machine learning. We already use it for several internal processes, but we also want to adapt it for external features, particularly for recommendations, e.g., on publisher recruitment or publisher mix. Furthermore, we want to focus more on automated insights, analysis, and reporting.
Thanks a lot for your time and the insights. We will repeat this interview, hopefully soon again!
Do you have some questions? Please get in touch with us now >>