Not Taking Programmatic In-House Is Short-sighted

By
  • Facebook
  • Google Plus
  • Twitter
  • LinkedIn

tom-triscani“Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is by Tom Triscari, CEO at Yieldr.

When an advertiser’s CEO, CFO and CMO meet to discuss their programmatic marketing strategy and budget, they increasingly talk about whether to take an in-house approach. Some argue such an approach is short-sighted.

I believe the opposite. The answer is often simple. Just do the math.

If you buy a lot of full-funnel display and pay 50% margins today (note: ad tech vendors and trading desks cannot disclose margins) but could pay 15% or less for in-house ad tech and pay the going market rate for a small team of campaign DMP engineers and optimizers, you end up generating massive savings. The savings can be used to augment or even create a big competitive advantage.

Some early adopter advertisers have already moved past talking and are actually doing it. When we see companies like Accenture move into the business, we know something big is changing. Early adoption is very smart; when done correctly, under a C-level mandate like P&G’s, the competitive edge gained can be an enduring one.

In-House Advantages

The simplest way to describe taking programmatic ad tech in house is to say that in the future, most advertisers will run their own in-house trading desk or put it on the consideration table. Those that go in-house have the potential to reap extra advantages and gain a market edge by virtue of owning a new competency and absorbing it into their DNA.

The argument to bring ad tech is increasingly validated. This conversation takes place because of six fundamental reasons:

  1. The sooner trading takes place in-house, the sooner the marketer can lock in better supply to find increasingly better audiences.
  2. Taking programmatic display in-house guarantees the advertiser gets 100% total and absolute cost and media transparency.
  3. The sooner trading takes place in-house, the sooner the marketer can activate more first-party data and experiment with strategic data sets.
  4. Taking programmatic display in-house allows for much higher RTB bidding, leading to much higher win rates for users the advertiser wants to reach.
  5. Taking programmatic display in-house guarantees the advertiser gets 100% total and absolute control over striking a balance between data privacy and meeting customer needs with respective product and service.
  6. The financial payoff of investing in in-house programmatic is massive, while talent acquisition, competency development and goal implementation can be easier than people realize.

Get The Job Done

When advertisers take control of programmatic display, they want to get the job done. As long as they have a sensible plan with a passionate and transparent partner, the job to be done becomes relatively easy.

One of the more important questions to think about: Why do so many advertisers feel so compelled to pursue in-house ad tech? Looking at this modified display landscape, it's becomes easy to see the big unmet need. For every dollar spent on getting an impression in front of a user, cumulative fees or “taxes” account for 75 cents while the actual media only costs 25 cents. While some of the fees are necessary pass-through cost, such as auction fees and ad serving, other fees, including DSP vendor or trading desk arbitrage, account for the largest tax.

It's important to note that the pass-through costs have favorable marginal cost aspects as the advertiser display spending grows in tandem with overall demand for display executions. These costs quickly become inconsequential as the amount of purchased impressions grows. The in-house staff are also very scalable because an advertiser with two people running 200 million impressions per month today does not need extra staff as they double or even triple display spending over time.

News reports, research and blog posts over the past few months further illustrate the unmet need and advertiser dissatisfaction. Since April, when there little few in-house discussion, we’ve seen the pace pick up to the point where big media outlets like The Wall Street Journal are following the story and equity analysts are factoring this threat and opportunity into valuation models.

The dissatisfaction has two facets. Advertisers want more cost and media transparency, but their ad tech and trading desk partners are reluctant to give it. Advertisers also want to activate more first-party data, but they are reluctant to let it outside the brand walls.

Follow Yieldr (@Yieldr) and AdExchanger (@adexchanger) on Twitter.

  • Facebook
  • Google Plus
  • Twitter
  • LinkedIn

Email This Post Email This Post

By on at

5 Responses to “Not Taking Programmatic In-House Is Short-sighted”


  1. Seb says:

    Hi Tom,
    interesting POV, though i disagree on many points.

    First, your arguments are based on an "if" assumption.."If you buy a lot of full-funnel display and pay 50% margins today"..
    Honestly, out of rocketfuel, criteo and google, which are publicly recognizing they do high level of margins (60 % for rocketfuel as declared pre IPO) i don't see any agency making a 50% margin on these campaigns.

    Now, even if they were doing margins above 15% and below 50%, you should consider what's integrated in the price paid by the advertiser to the agency (and that you do not list in your article), and that ultimately will have to be paid by the advertiser itself :
    - the team..i don't see any client running 200M impressions per month seeing an interest in integrating its trading activity in house..so rather consider bigger advertisers..with multi billions impressions campaigns per month..in that case you can indeed have one or 2 persons in house setuping the campaigns, that could be enough..Now you don't have the maths guys, the creative guys, the accountant, the marketing guys considered in your proposition..and this makes a lot more fees to pay by the advertiser by himself.. then consider all the scope before making any decision, not a small slice of it,
    - other tech fees : you know what ?..data's not free of charge, neither are viewability and safety solutions..ah and oh, by the way neither are semantic targeting or dynamic creatives and adserving ..all these are extra cost that an agency will always have better rates for, for the simple reason that its size is bigger than the size of one unique client by himself.. (any doubt on this ? try negotiating adserving deal with google..)
    - Cost transparency : so you think that a client who goes alone will have more transparency ? i m sorry, they will not be disclosed the terms between a publisher and the ssp or the exchange..they will not have access to the terms of the deals between the 3rd party data providers and the DSP, they won't have the transparency onthe relations between a dsp and an SSP.. so, if you're an advertiser..think about it, do you really need to send to the dustbin what your agency's doing for you (all the relations and hard work the agency's has done since years) to be alone by yourself, spend thousand or millions of dollars and in the end realize that, you haven't had more transparency by yourself (if this should happen, remember what your agency of record was telling you about this..)...
    - transparency on the delivery : so do you think that by yourself, you'll be able to have more informations on which publisher is hidden behind the "anonymous.google.com" domain ? ..think about it twice, it's the publishers decision to be hidden, not yours, not google's decision..that's the way it is..(for now)

    now back to you 6 arguments :
    1/ The sooner trading takes place in-house, the sooner the marketer can lock in better supply to find increasingly better audiences. ==> please explain..remember, it's not an upfront world..you do not need to be there 1st, to take the best placements..it's an auction world (think audience (bidded) planning rather than media planning
    2/ Taking programmatic display in-house guarantees the advertiser gets 100% total and absolute cost and media transparency. ==> not true , see above..
    3/ The sooner trading takes place in-house, the sooner the marketer can activate more first-party data and experiment with strategic data sets.==> still have to be done with the CRM & IT teams..i could agree with you on this, but that's also 100% possible when working with an ATD
    4/ Taking programmatic display in-house allows for much higher RTB bidding, leading to much higher win rates for users the advertiser wants to reach. ==> hmm, how many dsp's disclose the win rate ? remember that your article is based on an "if" assumption and does not include all "humans" and "other tech fees"..consider all of these before knowing what bid you can really assume
    5/ Taking programmatic display in-house guarantees the advertiser gets 100% total and absolute control over striking a balance between data privacy and meeting customer needs with respective product and service. ==> true, though it's possible with an ATD if the client wants to..
    6/ The financial payoff of investing in in-house programmatic is massive, while talent acquisition, competency development and goal implementation can be easier than people realize.==> that's a calculation that no one has already testified about.. talent acquisition and developement is not so affordable..especially as this is a new area of business and talents are rare..so expensive.

    So, as a conclusion, honestly i don't see real strong arguments unless the advertiser decide to be "half blind" when taking its decision.
    Any mature CMO/CEO deciding to take these operations in house, should consider the entire scope of charges that it includes unless they'll have surprises end of year.

    now consider this : programmatic is not RTB for display or mobile..programmatic (will) include(s) TV, radio, DOOH and potentially newspaper as well..all areas where the landscape is not made the same as with display, so eventually if you want to consider investing into an in house TD, consider what will happen in the next months/years and ask yourself if your team of 2 traders will be able to manage your programmatic display+ TV/RADIO/DOOH deals with what it implicates (deals, analytics, studies, internal education..) or if an agency has better options.

    • Tom says:

      Thanks Seb. Thoughtful response. It does leans toward protecting the status quo that is leaving many advertisers dissatisfied today, but we appreciate all points of view in the ever interesting programmatic space.

      The market is really big and in constant change. The opportunity for advertisers to go in house is also really big. Some will implement correctly and overcome all the fears you mention while others will not. Some advertisers will stick with more traditional strategies and either gain incremental satisfaction or remain indifferent. It all comes down to implementation and long view strategic alignment. The final decision is up to advertisers so I guess we'll all find out how things shake out one way or the other.

      Water always finds it's own way. It's the way all markets mature. With all the dissatisfaction around programmatic today, new models of efficiency will emerge. It seems that more in house control over decision making and asset creation with respect to data and talent must happen one way or the other as programmatic winds it way into other mediums. Display just happens to be a low risk / high reward starting point.

  2. Dorothy Higgins says:

    This is a very siloed POV. Two (at least) arguments against this are creative optimization with dynamic development and reading and acting on interaction across media platforms. In essence this devolves digital display (or other digital programmatic efforts) into latter day efficient spots and dots without opportunity for reflection and consumer media interaction.

    • Tom says:

      Thanks Dorothy. We have our own dynamic creative tech, algo and team in Amsterdam and Barcelona. On your second point, we never said there are not trade-offs. We do believe the trade-offs of going in house fall in favor of many advertisers, particularly those taking a long view and who also seek to develop privileged capabilities that turn in privileged assets.

  3. Brian says:

    Hi Tom,

    Thanks for the article. I (and my peers) agree with taking this direction. The value of bringing the data in-house, along with the inherent value in being able to efficiently activate media buys off of that first party data, adds to the overall package, even if the resourcing vs agency fees are a wash.

    Granted, we are one of the lucky groups that have already internalized SME's for many of the aspects that might be a barrier to "just doing it", and we have seen our costs drop and buying efficacy increase with our digital media campaigns. As programmatic brings in other media, we only see more benefits in following this course!

Leave a Reply