Shane Shevlin is Director of Business Development EMEA at IPONWEB. Here Shevlin discusses the IPONWEB solution in Europe, how the company works with ad tech vendors, and the critical role its technology plays in the real-time advertising eco-system.
For those unaware of the IPONWEB solution in Europe, can you give an overview of your current offering?
IPONWEB is essentially the tech provider behind many of the leading Ad tech platforms in the RTB space. We build and operate customised media trading systems that enable our clients to take advantage of RTB, Audience & Programmatic buying on both the buy-side and the sell-side of the equation, all of which can operate across display, mobile and video channels.
Our solutions leverage u-Platform™, our rapid development ecosystem that packages up the very complex component technologies required to operate in this space – Ad-serving at scale, real-time decisioning, machine learning, big data management, processing and analytics, all void of business logic. We then tailor the solution to specifically solve the business challenges that drive our clients to innovate and deliver the best results for their customers. So on top of this core u-Platform™ base, we then custom develop the algorithms & ad decisioning logic, together with any other unique feature-set requirements. We do this for each and every client.
Importantly our customers’ business models are always continually evolving and we provide our solutions in an agile-like framework that supports incredibly fast and flexible iterations of our clients’ ad decisioning and new feature expansion.
You recently left Google to join IPONWEB. Why did you make the move?
Google is a fantastic company and I learned a huge amount during my time there; an experience primarily centered on ad-serving, but which also included platform provision on the search side. When I began to discover the RTB ecosystem through my work at DoubleClick, I eventually found out about IPONWEB, as most people unavoidably do when they dig deep enough. What they were up to really ‘opened the komono’ for me on the potential of RTB. I was amazed at how prolific the company was across the entire space and as the display business at Google was quite compartmentalized, I saw working for IPONWEB as an opportunity to begin looking holistically again at the unique challenges of automated trading. With IPONWEB I have an opportunity to bring all the learnings of my ad-serving and search platform experiences to bear on a space that will have a profound affect on the future of media. It’s a really exciting thing to be a part of.
How does IPONWEB work with ad tech companies globally? Is it simply a spec-to-build proposition? How advanced do your clients have to be?
As a key enabler of the RTB eco-system globally, we have a incredibly diverse range of customers, operating our solutions across more than 20 countries. Our clients range from Global Ad Agencies and Fortune 500 companies to small start-ups. All of which give us some great insights on the key stages a market or business goes through in terms of RTB development, what is trending and how this differs and changes globally.
While each market is unique and at a different stage of development and complexity, the common delineating factor for our clients is having a very clear strategic opportunity, an existing asset-base and a deep understanding of their business model / application – this is quite universal. In many cases that may mean a client needs to build up some knowledge and understanding of the RTB space and opportunity first; either elsewhere via a 3rd party platform or through our new trial RTB product, u-Start. Then graduate to doing their own thing with IPONWEB, where we are more typically playing an ‘advanced solution role’. In some cases though, clients have been savvy enough to dive straight into working with us. We have had some great successes at all levels, so there is no fixed rule.
As a company however, we don’t consult on the product and business model. We think its important to be independent of this process and in so doing protect each clients own learnings, our version of a Chinese wall. So the single biggest dependency for working with us is that a client needs to own the product process and vision. This generally requires a small team of individuals with a highly sophisticated level of technical knowledge and industry experience.
To explain further, our engagements usually start with a business goal or a problem that our customer has found cannot be solved by existing market solutions or that to build in-house likely requires very high engineering risk & investment. The product owner role is therefore crucial in getting alignment on how exactly product features should work in the ‘real world’; establishing the trading and campaign objectives to be supported and understanding whether all business needs are being addressed for planners & sales teams etc.
Our project delivery team then literally works to custom-design u-Platform™ components, the business logic and the other functionality required through a serious of intense workshops & 101 question sessions. This is where we also design and map other key requirements, such as integrations with existing technology or data sources, inventory partners and other 3rd party solutions. A product person has to be able to understand all these questions, each required decision & possible customisations in order to take ownership of the solution and align it with business goals.
IPONWEB have been building for ad tech partners for the guts of ten years. Can you give some insight into the development capabilities and resource of the company?
Over the last 10 years or so, we have transitioned from building world class, bespoke ad-tech solutions such as ad-servers or ad exchanges, to more recently creating our own platforms and product solutions, but the focus is much the same. We simply love solving the complex, computational challenges specific to the ad tech space and we are very good at it!
I think the real thing to highlight here is that the vast majority of our developers and senior technologists are all still with the company after many years. Most companies today cannot keep a team together for more than a few years in this space, so the fact we have is something we are incredibly proud of. The teams’ continuity and vast experience has meant that we have designed u-Platform™ in such a way, that we can manage and support many totally unique, independent systems with a minimal resource footprint.
So we have a rapidly growing team of around 100 or so engineers, based largely in Moscow, from which we are able to support some 40 odd platforms today. Most serious companies in this space have that many engineers (if not many more) supporting just a single platform. Given most of our clients are flagship companies in this industry, it gives you some idea of the incredible flexibility and efficiency with which we operate. It has taken 10 years of experience and evolution to be able to do that, plus some incredibly smart people.
Right now, we are really focused on building out our in-market support teams at the account and sales engineer level. We currently have people on the ground in the US, London, Japan and Australia, but are looking to bring more people on in Germany, Southern Europe and the US in the immediate future.
How is IPONWEB positioned to help buy-side and sell-side vendors transition into automated buying?
Firstly, our platform solutions solve most of the complex technology pieces required to operate in this space (user modeling & prediction, algorithm design, real-time infrastructure, data processing & storage, analytics.) So companies are able to largely bypass the technology challenges and really focus on executing their business and learning how to make this new environment work for them.
In doing so, we have helped many companies improve campaign performance and uplift through automation, as well as realise incredible operational efficiencies for their organisations in the process. Our u-Predict user response probability modeling engine has repeatedly proven just how much more powerful machine learning is compared with manual techniques when it comes to campaign optimisation or maximizing inventory yields. We are quite confident that we really are the market leader in this space.
As we have been doing this for a while now, we can easily share with customers our experience of which data signals or factors work well and which less so when creating automated, decisioning models. It is entirely up to the customer however, to decide how they ultimately wish to optimise – how to bring their secret sauce to the boil, so to speak.
We are also starting to bring forward solutions to help mature companies new to the RTB ecosystem and struggling to decide how to proceed. We are in the process of launching a simple bidder connected to a single supply source (Google AdX) that allows these potential clients to ‘test the RTB water’ and sample how we work, then quickly move towards customising their own solution via u-Platform™ based on their exact needs. The product is called u-Start and basically represents an RTB bidder set of training wheels for platform operations and development.
How are you differentiating yourself from other “infrastructure” players in the market?
Most of the major infrastructure players in the market today operate a central platform solution and shared environment approach. This has some great advantages with ready to go inventory, integrated demand players and 3rd party partnerships, but it also has many limitations. You are effectively forced to fit your business into that eco-system and have very limited flexibility around feature sets or algorithm optimisation, as well as some significant data security concerns.
What we do is enable our clients to totally customize what they want to do, how they want to work and thereby take control. The key here is that our conversations start with a blank canvas. We have developed our stack in such a way that it requires something more than a signed contract to get up and going – it is built around our customers’ own ideas and vision. Our platform is cloud-based and the solutions we build are deployed on multiple installations with complete independence and isolation of data. We take user privacy very seriously and also understand that each customer has a unique set of requirements for how they a) use their own data and b) how they use the data made available through various bid streams. Both these considerations are central to our collaborative design process with customers and key to how our solutions leverage our core components. We don’t believe there is really anyone else out there that is currently able to offer that.
We are also quite independent of the media buying process, this means we are 100% focused on making the technology work for our clients and not driving volume or margin across an ecosystem – that is important to us and a key differentiator.
Where are the big growth areas for programmatic buying in Europe? It feels like the Netherlands and Germany are particularly hot at the minute. What trends are you seeing in the market?
We are seeing some very different market stages and approaches working across the European market right now. The Netherlands has been a very early leader in this space and has matured rapidly in terms of its size and sophistication over the last 2 years. We have excellent working relationships with our customers there and it has been a privilege to work with them to facilitate this growth and support their changing needs both domestically and as they expand across Europe. The culture of partnership we have experienced in the Dutch market really matches well with our approach to building solutions.
Germany is really still at that ‘building out an ecosystem’ stage, with many of the key domestic players only just starting to think about their strategy there. It is a very exciting time with lots of interesting plays and conversations shaping up. We have already had some great wins on the buy side, but see Germany as a key market for us this year on both Buy & Sell sides. In particular, we see a lot of synergies with our tailored, customized approach and the German Sales Houses’ need to integrate RTB and publisher solutions seamlessly within their existing business models and operations.
Video and Mobile are also key growth areas for us across Europe right now. We are really seeing a big increase in the number of mobile and video projects, either as stand-alone platform plays or multi-channel integrations. We expect to see both sectors grow healthily this year as more of this inventory is built out and the demand players come online.
There’s been recent trend around Buyers building their own bidders. How difficult is this process, and do you think that the “Bring Your Own Algorithm” strategy would work better for traders?
The beauty of working with IPONWEB is that you can start as big or as small as you wish. Many of our customers begin by building a simple bidder that has a slightly customised bid algorithm, which we pad with logic we know will allow for optimisation at a high level across the board. There is usually some flavour of prediction modeling baked in from the outset, as well as basic user retargeting.
The next stage integrates the more advanced thinking, experience and requirements from a customer; whether an e-commerce player, a pure performance agency or someone managing complex integrated brand campaigns that need more subtle optimisation and audience targeting. Each has unique data assets that they can leverage such as 1st party data, CRM, product taxonomy, user interactivity beyond impression and click, 3rd party data partnerships and more in order to better focus their goals.
A successful approach requires consideration and integration of all these factors in your ad decisioning and optimisation refinement, it is not a simple process. Every player in the industry is trying to establish a successful proposition based on the value that they uniquely bring to their customers or business, if you don’t have control of this piece, you can struggle to stay ahead. A ‘bring your own algorithm’ strategy ultimately becomes critical to your success very quickly.
At IPONWEB, we are uniquely positioned to enable this. Our ad decisioning, algorithms and prediction models are all custom designed for each client and their unique data sets, which typically represents their IP. This is also something we are constantly iterating, frequently as often as 5-6 times a month, as they look to evolve the complexity of their decision logic and try to stay ahead of the game.
We are now seeing a lot of interest from outside the ad-tech sphere, with Advertisers in particular interested in taking more control of this. For example, if you are an airline with real-time variations in flight seat availability and pricing, as well as a wealth of proprietary consumer information, you want to be able to adjust your budget spending, pricing, creative and targeting algorithmically in real-time, directly in tandem with your by-consumer, product sale opportunity. Just as importantly you also don’t want to share this valuable first party data across this ecosystem.
This is something that is quite difficult to do when there are many intermediaries between a buyer and the inventory and at present there are simply too many data hops and integrations in the value chain. With that said, we see a large role for DSP’s and Agencies in being able to provide these services and technology through to their end clients and cement longer term relationships through the technology.
We are all about trying to enable both of these scenarios, working with our DSP and Agency partners to work through to their advertisers. But also working with advertisers directly to seamlessly integrate solutions that work for their business.ExchangeWire