Has anyone heard of Blackwood Seven? They’re a media agency – no, check that – they’re a technology company – that has recently launched out of Denmark. And they are already posing a threat of sorts to the traditional media agency model. Having worked in and around the media agency business for a while, I get used to predictions about ‘the death of the media agency as we know it’.
This message has been playing with increasing intensity over the last year or so, as we all revisit the question of agency transparency – which, because of the efforts of the ANA, is being scrutinised harder than ever before.
Automation is happening…but is it happening right?
A debate partially linked to agency transparency is automation of the media trading process, which, it is argued, will ultimately create more transparency, as well as signalling the death of the media agency as we know it (not that the current automation efforts of media agencies are particularly transparent – at least, not according to the ANA).
I tend to be a bit cynical about proclamations of doom, or wholesale change. But I must admit, Blackwood Seven’s proposition has given me pause for thought. In a nutshell, the company offers a piece of media trading software, which couples algorithmic mathematics with an AI or self-learning capability, both of which operate together to generate the most efficient, and most effective, media plans. Essentially, it’s billed as an attribution model, a scenario planner, a channel strategy/channel learning tool and a media booking and evaluation tool, all rolled into one self-serve platform.
Don’t ask me how the machine actually works. But the financial model – self serve platform at a fixed cost, no matter what media an advertiser buys – could, if successful, have an ultimately profound impact on the industry.
- For a start, it removes at a stroke any implication of non-transparency on the part of the agency. Self serve platform with no clunky dashboards or multiple interfaces, no commissions, no deals as such given that all inventory is traded via an automated and biddable platform – and a machine that self learns, and is essentially predictive in nature.
- Second, it is explained as a ‘real time predictive modelling’ process, which removes a bugbear of more traditional approaches to modelling – namely, that the model is based on historical data.
- Third, it is designed to work across all media channels, something that media agencies are starting to achieve with automated buying, but relatively slowly.
- Fourth, it wraps all of the above into one platform. Media agencies have been working with multi-touch attribution models and automated trading platforms for a while – but in many cases, the process can be clunky/across multiple systems, and true all-channel automation continues to elude.
Blackwood Seven has recently raised €13.5m of backing, has wooed clients such as Volkswagen Germany from established names like Mediacom, claims a 25-50% uplift in effectiveness, has launched in the US, is planning entry to the UK market and has an ambition to be a global player.Also, given it is not actually a media agency as such, it is actively encouraging clients to talk directly to media owners, thus further cutting out a traditional agency.
Traditional media agencies have of course been talking about, and to some extent achieving, automation for some time. But from necessity, they’ve been trying to adapt it to their own commercial model, rather than starting from scratch – which has in part led to the allegations of murkiness that we’ve all been reading about.
Whether or not Blackwood Seven goes on to be a global game changer is anyone’s guess – perhaps it should ask its own model for a prediction. But it’s in the market and it’s showing no signs of going away.
The threat to the traditional model.
Traditional agencies are certainly challenged by such a model, in a variety of ways:
- Blackwood Seven, or companies like them, are tech companies, implying from-scratch specialism, core-business focus and no legacy systems to overcome.
- Whilst one could argue that an AI-guided trading platform is just that – a trading platform, without the strategic nuances of a media planner – it’s not a huge leap to imagine the AI developing along more pure-play strategy lines, or Blackwood Seven developing a human team of strategists to complement the service, or both.
- Media agencies don’t make a huge bunch of money from strategy. They make money predominantly from trading media, and the various fees and diversified revenue streams associated with it.
- ‘Strategy’ would also be one of the easier skillsets for clients to bring in house, as it is more reliant on human IP and would have less overhead for a client to manage than a trading operation.
- Many media sales networks would claim that they, in fact, are the ones coming up with all the good ideas that media agencies pass on to their clients. This, of course, is a subjective and debatable viewpoint, especially in context of multi-touchpoint campaigns that need more than one channel to work together. If Blackwood Seven is encouraging advertisers to get together with media sales networks more closely…well, you can guess the rest.
- A large proportion of media agency IP, across all disciplines, is wrapped up in human resource, and also local office overheads. Which costs money. Remaining competitive against a self-serve, non-localised platform, is going to be very hard.
Not all plain sailing.
Having said all this, Blackwood Seven will of course have its own challenges:
- Clients (not to mention media networks) are often naturally distrustful of media attribution models being used to try and calculate media channel or trading strategy – not least of which, clients that are entrenched in views about ‘what works’. Blackwood Seven will face the same challenge as any traditional media agency in this space – what’s in the black box, and does it actually deliver?
- The quality and breadth of inventory released to Blackwood Seven, particularly given the myriad of legacy relationships and agreements between media agencies and media owners, may be called into question, particularly in the early days. They’ll be using ad-exchanges for digital media, but what about other media not yet universally traded in this way?
- The universal nature of inventory access – i.e. will they be able to sign up ‘all’ media networks as, as a media agency does – will also be very interesting to understand. How would Facebook, for example, react to such a proposition?
- However we look at it, a large part of the success of media agencies has been in driving human relationships with their clients. Clients would have to seriously ask themselves if a self-serve option would really work for their business, in reality.
- The nuances involved in executing something more than a spot-buy (e.g. a sports sponsorship package) would likely not be something that such a model could grasp. Yes, clients could negotiate directly with media sales networks – but isn’t this something that a media agency could do better?
- Self serve platforms, or technology stacks which claim to have ‘all the answers’ for the client in other areas are not new – but they are often harder to actually operate than clients initially realise.
- Finally – whether the model is historical or predictive, the quality of results from a Blackwood Seven model will largely be defined by the quality of data input from client sources – and strong data is something that continues to plague many, many organisations.
No Crystal Ball, but worth watching.
To quote Bertram Wooster, I’m certainly going to be watching Blackwood Seven’s future career with considerable interest.
Whether this kind of model rips traditional media agencies apart or not, it’s clear that the company is gaining some traction.
And furthermore, that the evolution of AI capability could represent a huge difference, plugging a big human gap that allows a self-serve model to become viable.