Robert Webster, Founder, Canton Marketing Solutions and Wayne Blodwell, Founder & CEO at The Programmatic Advisory, are two of the most knowledgable people in the digital marketing industry. Both have firm, informed, views on industry development. While normally aligned in outlook, the two have very different views on the benefits or otherwise of walled gardens. In a special series of deep-dive articles, the two defend their positions.
Rob goes first:
Why the Privacy First Era will be defined by next generation walled gardens
To set the scene, I have been a huge fanboy of open programmatic for much of my career. My background of over twenty years in adtech meant that I had the opportunity to lead the programmatic rollout at both Unique Digital and MediaCom, as well as actually designing and building programmatic technology for the open ecosystem at Crimtan. Yet, this background means I have had the same experience in paid search and paid social being at Yahoo in the early days of paid search and MediaCom when Facebook Audience Network launched.
This experience and focus on cross-channel skills and the technology is something we have instilled at Canton where we advise tier 1 clients on execution in all digital channels, as well as the data and technology required. It is with this fairly unique experience, and with a heavy heart, I have come to realise that the future of all advertising is (mostly) on buying from platforms protected by walls. That in fact there is a natural progression from the open RTB era to the modern walled garden – just as there was from the ad network to the programmatic ecosystem an advertising era ago.
There’s a certain romance in the open ecosystem, the little guy next to the big. The analogy is that of an open street market vs a brand name owned megamart. The open market gives opportunity for smaller players to succeed and for more competition – it is an attractive notion and one that in the right circumstances I believe in. Yet the street market can also rob the unwary buyer of their shirt.
As Derek Trotter said “no income tax, no VAT, no money back no guarantee”. In an open market where no one takes responsibility and regulation is hard, inexperienced buyers can struggle, bad actors can succeed. Even sellers of quality products can be hurt as the neighbouring stall offers seemingly identical knock off products for a fraction of the price. Despite the romance of the open market stall only a small fraction of goods are bought that way. As it is for the sale of goods, so it goes for advertising.
It’s at this point we need to define what we mean by a walled garden. A general definition is one where the access to such a garden is in some way restricted. As such this means restricted supply (only buying from approved vendors) or restricting the right to buy media to approved advertisers (where that approval can be rescinded upon poor behavior). Another common trait I want to highlight is that walled gardens restrict data leakage on behalf of advertisers and publishers alike and don’t allow access to consumers’ IDs unless those users choose to actively opt in or engage with that actor (normally an advertiser). As well as the GA giants, this behaviour describes many of the new big players in connected TV (notably the broadcasters), emerging publisher co-ops and vertical ad networks.
Do not mistake this piece for an endorsement of the dominance of Google and Facebook particularly. I hope that other owners of walled gardens (and not Apple or Amazon either) can challenge the hegemony of GAFA and lead to a more equitable ecosystem. Instead, with a heavy heart, I am asking the market to learn the lessons of the programmatic era to better challenge GAFA in the privacy-first era and for almost all quality publishers to move to a set of closed ecosystems that can challenge GAFA. For publishers to remain fragmented and with all the disadvantages of an open ecosystem is to surrender the web to the giants.
*Now taxation may be an issue the leading walled gardens address I confess.
Walled gardens provide better returns for advertisers
High walls protect an advertiser’s spend from bad actors, give quality publishers the confidence they need to offer the media advertisers want and have clear rules on data usage that protects the brand’s customers’ confidence. These key pillars are why advertisers prefer buying media from walled gardens (over 80% of online media spend is with walled gardens).
Not only that but the walled gardens make it easy to buy media. Compare buying media from a walled garden, which almost anyone can do to some degree, with the difficulty in setting up a programmatic campaign – giving walled gardens a huge SME advantage. A walled garden is only as good as the media opportunities that lie within, however where publishers with sufficient advertiser value are concerned, a walled garden is infinitely preferable as it allows the advertiser to easily gain that value without having to navigate the open ecosystem’s many issues.
The open ecosystem has sadly failed to deal with some crucial issues. The PwC transparency service showed that only 51% of fees was reaching the desired publishers – this from campaigns run by the biggest companies on the best inventory – many will have experienced far worse. As Traffic Guard says; “with complex and opaque supply chains between buyer and seller. It is estimated that programmatic ad fraud levels stand at 17% in the US.”
These figures taken together are truly terrifying. Whilst it is true that good buying practice and technology can protect buyers from this (trust me I know) the point is this should not be necessary. The unwary buyer should not be punished in this way and it’s a business risk of funding hate and terrorism that advertisers simply do not want or need to take.
This is not to say that the walled gardens are perfect here. However, there is one crucial difference, they are responsible for the inventory that they sell. Advertisers can choose to ignore walled gardens that do not look after their needs and can seek money back where they can show problems – something that has been almost impossible with the opaque open ecosystem (so opaque that experienced auditors at PWC struggled to audit it). Any attempts at recompense in the open world usually fail due to the myriad spider’s web of actors who all seek to blame others in the first instance.
Results are of course the biggest factor in advertisers’ chosen media. Until relatively recent times most big brands were built on the back of TV advertising. In the digital era, businesses have been built on the back of search and paid social. There are many stories of elections being won on the use (or misuse) of walled garden advertising. Yet similar stories do not exist for open RTB.
I can name many brands that focus on search, shopping, affiliates or social, yet I know none whose main channel is open RTB based. As connected TV develops into a walled garden-led sell, this really matters. If open RTB were the dominant opportunity it could afford a few missteps. It is not. Indeed as open RTB is disrupted by the identity revolution I know many brands will simply look elsewhere until the disruption dies down.
That is not to say that the open ecosystem can’t produce results as I have seen many times in my career. However, doing so and mitigating the inherent problems makes an already tricky space require expensive expertise and technology to be able to operate effectively. DSPs have high minimum spends followed by the need for fraud detection, content protection, viewability control, supply path optimisation, optimisation algorithms. All of this is much easier in the strong walled environments because they have the ability to remove the worst media at source and control the experience.
Fans of the open market will point to frequency management as their killer point. Now frequency is important and effectively impossible to control across walled gardens, it is true. Yet when you consider that 80% of all media is spent on said walled gardens (and rising) the argument becomes mute. Instead, buyers need to control frequency individually on each ecosystem. Had the open web been the 80& and the walled gardens the 20 there might have been a strong argument here for the open web.
Furthermore the idea of cross domain frequency caps relies on cookies and consent in a way incompatible with privacy legislation. Much as we would love a world where frequency can be applied on an individual level in all channels, that ship has sailed and any benefits then of the open ecosystem are purely theoretical – sadly the world exists as it is and not how idealists would choose it to be.
The experience for advertisers will tilt even more in favour of walled gardens with the movement to the privacy-first era. Open RTB thrived for a time on retargeting and personal data buying through the broadcast of consumers’ IDs to myriad other parties (both benign and malevolent) to facilitate that. This practice will rightly be massively curtailed in the privacy-first era and whilst new ID solutions may replace some of the cookie practices of old it will be at far less scale. Walled gardens as we shall see in the consumer privacy section are far better environments for data buying to operate in.
Ultimately advertisers also go where the quality publishers live and that is increasingly walled gardens. Walled gardens dominate overall spends as we have seen and also the high quality emerging areas of connected tv and digital out of home. We will expand on this in our publisher section. If we all lived in utopia we would all be shopping at local stores and buying our media from the open ecosystem. The reality is, for the vast majority of the time we choose to buy from protected environments. For advertisers that means advertisers need to move on from the open ecosystem to a focus on walled gardens.
Walled gardens are a better way for advertisers to measure media performance
I feel it’s important to call out attribution specifically. In a previous time this would have been a big win for the open web. Cookies allowed for attribution to run cross site and over at least 30-day periods (if not much longer) to see the impact that media had on sales. However now in the privacy first era this ship has sailed. Without a single ID running through all media touchpoints (which we haven’t really ever had anyway) different models are required to run cross channel, site (and garden) attribution. Thus at a stroke destroying one of the key benefits of the open ecosystem.
In this new era there may be less personal data but actually opportunities for more precise measurement due to the benefits walled gardens provide. Many of them offer seamless opportunities for marketing experiments, without having to buy the control group media as is the norm in the open ecosystem (the infamous charity ad in many cases). Walled gardens also have the opportunity to use their own first party data (i.e. the logged in users to the garden) to facilitate attribution studies.
Panels can be seamlessly integrated to walled garden media experiences for a human led touch on insights and attribution. The development of ever more sophisticated techniques with data clean rooms – where data can be used but privacy protected – allow for more advanced attribution tests to be carried out between a series of publishers’ gardens. Indeed the great data each garden can give can then be compared to real world advertiser sales figures.
The real opportunity here is for a new measurement paradigm to emerge based on the techniques above and combined with turbocharged media mix modelling to deliver on the promise of true incrementality of media spend.