In this interview with ExchangeWire following the recent webcast discussing the future infrastructure of the ad tech industry, Alkimi Exchange's Neil Bruce (CCO); Adam Chorley (COO); and Ben Putley (CEO) discuss the benefits of pursuing a decentralised approach and how this addresses cornerstone issues within the industry, such as trust and publisher revenue bleed.

What does a decentralised approach entail in the context of digital advertising and how does this differ from the current conventional structure?

When we talk about decentralisation, we mean the transfer of control and decision-making from a centralised entity to a distributed network - which reduces the level of trust that participants must place in one another, and removes the ability to exert authority or control over one another in ways that compromise the functionality of the network.

The concept of decentralisation is not massively new, but has been gaining momentum with the emergence and evolution of blockchain technology - particularly since decentralised finance (DeFi) solutions emerged in 2019.

DeFi refers to a system by which software written on blockchains makes it possible for buyers, sellers, lenders, and borrowers to interact peer to peer or with a strictly software-based middleman rather than a company or institution facilitating a transaction.

Our story here is that we're applying the principles of DeFi exchanges to the advertising auction.

Neil Bruce, CCO, Alkimi Exchange

There can be no doubt that the growth of programmatic has been revolutionary for the advertising industry. But it’s rapid growth has been unchecked - meaning that the Lumascape has become extremely complicated. A brilliant analogy we’ve been told recently likens the evolution of programmatic to the Weasley’s house in Harry Potter. Bear with us here! Open RTB would be the blueprint for the house. Well-intentioned and well-planned. But as the Weasley family grew they needed to add more rooms to the extent it ended up higgledy-piggledy.

That's the way that we see the ad tech landscape at the moment. You've got this tangled and complicated web of businesses sitting between publishers and advertisers. These businesses rely on centralised infrastructure - taking server space from Amazon Web Services, Oracle, IBM, or any of these massive blue-chip companies - which has to a large degree led to the “ad tech tax” which now siphons off 49% of an advertiser’s dollar before it reaches a publisher.

The advertising industry has known for some time that blockchain technology can solve a lot of problems that it currently faces, but to date, it has been impossible to host an advertising auction on the blockchain. The first problem has been that the benefits of decentralisation come at a cost - trade-offs such as lower transaction throughput. A transaction time taking minutes might work for financial services, but is clearly not fast enough for real time bidding. The second problem is that legacy blockchain solutions like Ethereum charge relatively high transaction (“gas”) fees. If we were to apply these fees to every impression delivered via RTB, even a well funded business would go bankrupt in minutes.

We’re hugely lucky to be the first business incubated by Constellation Network, who have built a revolutionary blockchain technology called the Hypergraph. The Hypergraph is fee-less, infinitely scalable and for the first time, able to process transactions at the speeds required to host an RTB auction on the blockchain. Our thinking is that by taking some of the computing to the edge and decentralising, it means that we can create a system that puts a lot of that revenue back to publishers and creates transparency in terms of the fees and the overall model.

The headline is that we can come in at a 10x lower cost than the centralised exchanges are charging, but the other benefits of decentralisation stem from the fact that data that has been passed through a distributed ledger, has been validated through that ledger as being a genuine dataset, and is immutable (cannot be changed). We can then pass that validated data to publishers, advertisers, and people that work in the exchange. There's no running away and there's no hiding. Let’s take margins as an example: they will be there for everyone to see - a fully transparent view of what's happened in any advertising auction.

The entities that operate on Constellation’s Hypergraph are called State Channels and each State Channel is able to work to incredibly sophisticated logic sets. Within the realm of advertising, this logic would allow us, for example: to preemptively filter out fraud before it goes into the auction. Similarly, we can create complex rules for audiences, contextual targeting, etc. It has a huge benefit in terms of transparency, but also has additional functionality that centralised exchanges can't bring to the table today.

Decentralisation can actually go a level above this and provide instant access to log-level data because of the nature of the blockchain. You have some current exchanges making a big song and dance about the fact that they will give you log-level data, but that data could be 24 hours old by the time that they give it to you, while other exchanges will not give you log-level data at all because they say it's too expensive for them.

So, to cut a long story short, the benefits of a decentralised approach are many, but the key here is that we are the first company that's actually able to host the advertising auction on the blockchain. There are blockchain companies out there that are doing great work in the advertising space, but because they're trying to use Ethereum, or build their own blockchains, they are unable to host the auction itself. The fact that we can offer this in the real-time bidding environment is completely ground-breaking and it's never been done, and that's why we're ridiculously excited about what we're bringing to market.

What are the main challenges for publishers and brands within the current advertising ecosystem? How does pursuing a decentralised approach solve these issues, along with restoring trust from the user?

Adam Chorley, COO, Alkimi Exchange

If we look at this from the supply side first, we feel like at the moment publishers are at the behest of big tech. The recent set of financial results show record highs for Facebook, Google, Microsoft etc. Taking Facebook as an example - they've created an ecosystem where they don't create any content themselves, but they have a platform where people will be sharing content with their own echo chambers. That means that publishers are increasingly beholden to their algorithms, so we get to the point where publishers actually write content based on the people that are consuming it in social media channels. In turn their content becomes increasingly click-baity and polarised.

We've got to an environment where even the most respectable news platforms have started to dumb down their content in order to appeal to these audiences. We want to rectify that.

Add to this the 49% ad tech tax we’ve already mentioned and you have publishers resorting to creating unpleasant environments with 56 ad slots on the homepage. You also get the really annoying and poorly-targeted ads that follow you around.

The way that we position Alkimi is that the intended value exchange of the open web has been broken. You have this tacit agreement that if you're not going to pay for content on the internet, which most people don't want to do, then you get free content via advertising. It's as old as consumerism itself.

We shouldn’t be upset with publishers letting tech steal a march here. They’re not tech companies. The internet as a technology has existed for 30 years, give or take. CNN and The Times were pushing out news before Mark Zuckerberg's parents existed. Zuckerberg grew up at the time when the internet was proliferating rapidly, so he was able to steal a march on the world's publishers because they were still making magazines and creating print. They were still using old technology, then new technology came in alongside Facebook and Google telling them "this is how you should do it." Publishers were so busy trying to get a double-page spread sold, so they went with it and didn't realise that they were giving away the keys to the kingdom.

As the internet became ubiquitous, another thousand companies sprung up in the void between publisher and advertisers. Publishers, while trying to gain more eyeballs, push out good content, and continue to attract expertise into the business, also have also to monetise it with all these different technologies, RFP all those technologies, manage those technologies, get legal to sign off on the SLA, etc. It just becomes unwieldy, and that confusion has allowed people to take this 49% tax.

We want to create an environment where we move away from an exploitative economic model where a few people in the middle take out the resources to a generative economic model, where the resources flow more freely. The reason that we're able to do that is because of Constellation’s groundbreaking technology. On the Hypergraph, any data validation emits a reward. By inputting data, Publishers will immediately see actual value, which in turn can be used to collect more data. This becomes much more circular as they start to tokenise data at scale.

The tokenisation of bandwidth means that publishers won’t have to pay their AWS bill or cloud infrastructure bill every month. It’s not like you stop paying and you lose it - a token equates to a certain amount of bandwidth that you can use that forever as yours. You can lease it to other publishers, you can lease it to other parts of your business, whether you are in the US, or Europe, or anywhere else. It's as though you're buying your part of the telephone wire: your telephone pole is fixed to the other telephone poles and you can monetise it as such. Whereas at the moment, you just get to rent a bit of the pole rather than own the whole thing. That's how big this is going to be in terms of its ability to tokenise data, notarise data, and use that to create brand new economies that don't exist right now. It can usher in the era of the data-preneur rather than the entrepreneur.

[Bruce] I can bookend this from the buy side perspective. I spent five years at Mindshare running the mobile and digital innovation teams, and my first and last months there saw two very similar experiences. During my first month there was an ad for an 18+ product that was displayed in an app for children. Obviously the client went insane and we had to unpick that. In my last week, one of our brand's ads showed up in Breitbart and the client was equally unimpressed.

In both instances, you get the horrible email from the client with the screenshot asking "What the hell is going on?" Everyone in the business has to start working out where this came from. Once you start chasing down that rabbit hole of your supply chain, you start realising how mixed up and ridiculous it is. For the latter campaign, we realised we had a partner on plan which was buying from another partner, which was itself buying from a network which was also on plan. We had the original supply source twice essentially on the plan, but one of them was that original supply source with three layers of arbitrage and margin put on it, thus being XYZ more expensive.

Supply path optimisation is one of the buzzwords of the moment - with good reason. It’s hugely important because, even with the best will in the world, ads.txt compliance doesn't fully let you know what is going on. There's bid shading, there's undisclosed margins, and listed companies putting out results showing that they actually make more margin than their websites say that are charging clients. And we have ISBA saying that 15% of supply chain costs are “completely untraceable.”

There's the famous quote of "I know half of my budget is wasted, but I don't know which half", and it's literally never been more true. We were talking to an agency working on supply path optimisation initiative for a global bank. They could only directly attribute 30% of the spend within the supply path. So 70% of that was lost. They had ad tech lawyers turning up to say which part of the data they could, or couldn't, see. This is your ad spend and they're sending lawyers around to tell you that you're not allowed to see where it's gone. It's a bonkers relationship.

How can conventional advertising data streams be incorporated within the decentralised platforms and how would this work on a cross-channel basis?

Ben Putley, CEO, Alkimi Exchange

About six months ago, Benjamin Diggles, who is the co-founder and CSO of Constellation, was talking about the work that they did with MOBI, in authoring a standard for self-driving cars. This meant working with the likes of Ford and General Motors, as well as the people that made the LiDAR for the cars and all these different stakeholders. Diggles was saying how ridiculously difficult it was for Constellation to try and juggle to organise that relationship, to the extent that they've actually walked away from MOBI because it was too hard for them to manage.

We had this light bulb moment when we were having this conversation, because the standards in digital advertising already exist. You have OpenRTB and Prebid. These cover off around 97% of the programmatic market and are fit for purpose. Constellation's Hypergraph is great because it's built to be interoperable with any other programming language. It's very easy for us to create endpoints, essentially connecting OpenRTB logic and Prebid logic to the Hypergraph. That means that although we are decentralising the infrastructure in the middle, we're using the existing standards that the advertising industry already uses, so there is no need for publishers or advertisers to adopt a new standard. They don't even have to hold crypto. They can work via a SAAS model exactly how they've already worked. They don't have to know the technical details, they're just going to get this solution that costs them a lot less and provides them with a lot more transparency.

In the future, referencing our earlier points around datapreneurs and data being the new oil, we envisage that the end user can actually decide to give their attention in return for monetary reward. Look at Tim Berners-Lee's new project, Solid, focusing on data sovereignty and the fact that you have your own "pod" of data. You have permissions and you can decide to give your data and information about yourself to your bank, to the government, to XYZ, to make your life easier. There's no reason why you couldn't then give permission to an advertiser in return for financial reward. You would have that interoperable, as a state channel, with Alkimi Exchange. Similarly you can port in other data schemas, as many as you want, in order to build the best picture of the audience you want to reach.

In terms of cross-channel, it works in a similar way. All of the exchanges are making a big play for connected television and DOOH. There's no reason why we can't (and aren’t!) bake(ing) that into what's being delivered on Alkimi Exchange. We would certainly be going live straight away with the ability to VPAID and VAST in order to unlock the desktop and mobile worlds. Down the road, we would envisage that anything that is tradable programmatically is able to be surfaced by Alkimi.