This commentary is my own and is not necessarily reflective of the views of the WFA or its members.
First I want to extend my gratitude to Marco who has taken the time to comment on these important points in such a structured manner (source: WSJ). It is very nice to see that a major ATD agrees with so many of the points made in the paper, right out of the gates.
I do have with some of the comments, which I’m afraid, could be misleading to those not entirely clear on the workings of the eco-system.
To provide further personal background on myself for the evaluation of the relevance of my commentary:
- started to work with digital media in the mid-nineties and have a proven track-record in innovation and thought leadership
- founded two companies that provided technology and consulting solutions for ATDs and ITD
- serviced and advised various ATD/ITD between 2012 and now
- was part of various ATD vendor selection processes as a vendor
- advised brands on vendor selection for their in-house programmatic needs
- was the first to provide whitelists / blacklists for ATD/ITD in the APAC region
- had 1st-hand exposure to most ATD/ITD being setup in APAC during 2012 and 2013
Here is what the GM of Omnicom’s Accuen for APAC at the time, Matt Harty had to say about my relevance to their work:
“Mikko is a very clever man. He has the ability to take some thing from a brainstorming session and have it in prototype in a matter of days. It is truly amazing. I expect to be working on things with Mikko for many years to come.”
Here is another comment, from James Welch, at the time MD of WPP’s “private” trading desk theMIG:
“Mikko is bright, driven and what’s more he’s well known and revered among the digital data & programmatic players in APAC. He sticks his neck out for the sake of clarity, transparency and keeping the other buggers honest.”
Coming from that kind of background, the main problem I saw was that the trading desks did not seem ideally positioned to have enough time for the level of focus an individual client would need. I found that there is always an element of generalisation in that situation. It’s like having a football manager managing a few teams at one go. I never thought that ATD would not be or should not be a partner for the advertiser, I just never believed the initial model was the right one. If the initial model was always what we stuck with emerging things, then there would be no progress or it would very slow.
Basically my argument is that it’s better to have a manager that is 100% focused on the success of your team, even though it was not the best manager of them all.
MARCO: Don’t just ask these questions of your agency trading desk, however; Ask them of every programmatic provider you might be spending with today.
MIKKO: This is exactly the point I would as an advisor to advertisers emphasise. You have to treat all of your vendors equally and the same kind of deliberation (and procurement) should go to evaluating the ATD as any other significant purchase the advertiser is making.
Media investment is one of the most significant functions of any major advertiser. Now with programmatic we will automatically have the opportunity of consolidating that investment more or less under one central point of control. I think this is the kind of point we should be discussing; where is that central point of control located.
MARCO:. An in-house operation is not going to resolve all transparency issues. It might give marketers complete control, but it also results in limited visibility once the campaigns go out the door, and you are only as good as the technology you tie yourself to.
MIKKO: It will completely solve the transparency issue with the advertiser has with the ATD (at least). Please note from the WFA survey that this seems to be a concern, so as an industry we should discuss that particular point.
Also there is no less visibility for BTD than there is for ATD. On the contrary, the BTD has better incentives towards not buying masked inventory for example, among other things.
Basically my argument is that a) the brand itself will always have clearer incentives do what is best for it as compared to ATD and b) media investment is way too important to be controlled by a party with mixed-incentives. If we can agree to this, the whole conversation becomes much simpler.
MARCO: Meanwhile, if marketers outsource to an ad network, managed service DSP or non-disclosed trading desk, you have little control, less visibility and no ownership.
MIKKO: Yes, this is why the paper makes a specific point about being flexible and about the ease-of-movement between partners. Also it makes the case of ATD being better than ad network, and about programmatic giving an opportunity for the advertiser to gain better control over how commissions are being spend. I’m not making any comment about if the ATD should be or shouldn’t be one of the partners. That depends entirely on the case.
Again I want to emphasise my personal view, that investing money in to something (value-adds, audience data) is not the problem, but it should not be done on the advertisers expense without pro-actively explaining to the advertiser their options.
This Includes the option where they eventually end up controlling everything. That is a discussion where the ATD is under pressure to clearly articulate it’s role as a partner in the short and long-term. Just like any other vendor would have to.
I understand this is less than desirable from the ATD stand point, especially in those cases where the ATD does all the vetting and makes many of the decisions for a particular client.
MARCO: It should also be tireless in pursuit of viewability and quality, and it should show you how it is trying to protect your ads from fraud.
MIKKO: If the Advertiser chooses one of the accredited viewability solutions, they will be within +/-10% of the result the ATD will get from using one of the other accredit solutions. This we can clearly understand from the viewability standard created and published by IAB a few months back.
In terms of quality controls, these are also widely available solutions that a BTD will have very much the same access as an ATD will.
MARCO: I submit that an in-house option or managed service demand-side platform that buys on a marketer’s behalf will provide less brand safety than an agency trading desk.
MIKKO: First it’s important to understand that brands are in a position to go far more custom in the longer term. Even completely custom bidding algorithms. This seems to be the main issue of the ATD, there is a propensity towards generalisation at all level. Which is understandable from the profitability stand-point.
Second, there is absolutely nothing that suggests brands would not have the access to same tools as ATD in this category of solutions as well. Peer39, Integral and all other solutions that tackle these issues are more or less equally available for brands and ATDs alike. They work the same way for both and are typically accessed through the DSP. Many DSPs support these same solutions. Baking them in to your own stack is also available for BTDs as much as it is for ATDs, with very few exceptions.
As before, we find the advertiser far better (compared to an ATD) incentivised to deal with issues like brand safety well.
MARCO: It simply costs too much to deliver extensive black and white lists
MIKKO: This takes us back to the key point i.e. are you likely to get a better result when one player does things based on their own interest, or when a 3rd-party does it for them (and a lot of others at the same) based on mixed interests?
When it comes to building lists, there is no way the ATD can ever be as competitive as the advertiser itself, when it comes to building an advertiser level whitelist or blacklist. In the shortest of terms maybe yes, but that’s only because now the ATD holds the lists the Advertiser’s campaign is ran against.
Also I think manually building these lists is very much an out-dated practice. I would use a small seed list in english, extrapolate to a much larger look-a-like list, use language-agnostic NLP to match sites from all other languages and use one of the crowd platforms for quality control. Is just a small example of how we can solve some of these problems for all of us with a little ingenuity.
MARCO: tech vetting and human vetting at a client level.
MIKKO: I hate to repeat myself, but this is again the same question if it is better to look after your own interests with something as important as media investment, or out-source it to someone who has their own bottom-line to think about and is bundling decisions made for you with decisions made for others.
As an advertiser, you would not have to over-engineer this to improve your position significantly. I think a lot of people agree that companies like The Trade Desk, Krux and Integral are all very competitive products. Putting those three together and working out of that initially will not put you behind from where you are now. It’s not the technology and data that wins games, it’s the way you put it to work for you.
In regards to the talent aspect, there are guys like Dean from Accuen or Kajal from Xaxis who are like these great scholarly geniuses, and you’d definitely want them somehow contributing to what you do. But I think a BTD could be an interesting place to work for some of the future Deans and Kajals.
Also what comes to buying services in general, advertisers (procurement) are experts with that already. Nobody knows that better than media agencies ;)
MARCO: It might give marketers complete control, but it also results in limited visibility once the campaigns go out the door
MIKKO: The advertiser already has that limited visibility now, there is no additional limitation to the visibility resulting from moving to BTD.
MARCO: you are only as good as the technology you tie yourself to.
MIKKO: The technologies have much less variety between them than the possible ways of using them. Or the people that are using the technologies. Or the motivations that drive those people to behave in certain ways. So outcome wise, technology seems the least significant of all the factors at the moment. Technology is the car, but it needs a driver, fuel and a destination.
I’ve said this for years, programmatic is not about data and technology, it’s about culture change. And culture change is always a result of behaviour change. Behaviour change is very much connected with motivations.
That’s the kind of things we should be focusing on in my opinion. How can we all change our behaviour in a way that leads to more efficient markets.
MARCO: I would love to talk to authors of this report about their findings, where the insights were obtained and how the calculations were done. So much of the report is spot on in terms of what questions to ask, but the bias and inaccuracies need to be corrected.
MIKKO: I’m completely open to a factual discussion about these matters with you and the rest of the VivaKi leadership team. Perhaps a direct peer dialogue between our parties would be a meaningful way to untie some of knots that seem to burden the conversation at the moment.