In his continuing regular series, programmatic expert Paul Gubbins asks what questions agencies and advertisers should be asking of their buy-side vendors, and examines the changing nature of transparency in the sector.
My column this week has been inspired by a conversation I had on Twitter with Paul Frampton, CEO and UK country manager of Havas Media Group. Paul was asking why so few ad tech vendors can differentiate and are often, at times, reluctant to lift the bonnet (or hood for my American friends) on their platforms to illustrate just how they work and make their money.
As we move full steam ahead into Ad Week Europe and the resulting social white noise it will generate, I wanted to have a focused look back before suggesting how we can look forward when it comes to buy-side vendor selection and adoption.
Many moons ago, I left Unanimis to follow a chap called Alex Rahaman, helping him launch a mobile DSP called StrikeAd. If we cast our minds back to late 2010/early 2011, mobile ad networks were still the default partners for many agencies and provided both great service and returns. However, in the same way that programmatic buying and selling had disrupted desktop digital several years prior, the same was about to happen in the mobile ecosystem.
My job at the time was to generate demand, either via IO’s or SaaS contracts with the agency trading desks. Although I was starting from a position of strength as I had good relationships with most of the agencies I was selling to, very few buyers trusted the concept of programmatic in mobile (again, this was a time when mobile SSP’s and exchanges were still getting ramped; think Nexage, MoPub and Smaato, etc.) and the narrative of migrating audiences had yet to gain real momentum.
This reluctance from the buy side to work with us via a managed or self-serve basis needed to be addressed, and, as a fledgling start-up, had to be done so quickly. We decided on a course of action that to many in the industry at the time was unheard of: we invited each of the senior heads of mobile into StrikeAd and ‘lifted the hood’.
We illustrated exactly where we bought the inventory from and explained that to run a managed buy, we would make a margin of ‘XYZ’, but provide a service layer that included optimization reporting and post campaign analysis etc. If they licensed our DSP via a SaaS contract, they would be charged ‘XYZ’ and expected to pull the levers at the ATD level themselves.
This exercise was the catalyst to both an increase in managed demand and ATD adoption almost overnight, as the value proposition of RTB over opaque ad network buying was clear, the programmatic technology explained and concerns around margins and non-disclosed fees mitigated by the transparency on business models we had offered.
Although the scenario of full disclosure above sounds like basics in today’s practices, according to many senior leads on the buy side, this it is still not happening.
So how did we end up here? At a time when venture capital firms were liquid and the appetite to win big on the next AdMeld or Invite Media was palpable, many backed ad tech companies that in hindsight transpired to be exit vehicles that could trade around the early confusion that surrounded programmatic, rather than being built to meet the needs of advertisers and their agencies.
During this time, there was also a perception that proprietary tech (UI/Bidder etc) was a vendor’s main USP, and if you were seen to be relying on any technology other than your own, this was perceived to be a weakness and prohibited many from lifting their hoods.
"There are no precedents for today’s data driven scenarios we are planning for, so collaboration and knowledge sharing are a must"
In hindsight, we should have recognised as an industry that a company with only several million in seed or series A funding was never going to have built a scaled programmatic infrastructure to support the querying and bidding across billions of impressions, and many were in fact leveraging third parties for help behind the scenes.
Fast forward to 2017 and there are many companies on the programmatic buy side that now recognise the value is not in trying to claim or even paying for ownership of back end infrastructure like bidders, but the intelligence you push into them to make informed decisions.
Open tech ecosystems such as AppNexus or development companies like the renowned IPONWEB have empowered vendors to realise that owning a blender (a crude analogy for two very sophisticated engineering companies) does not make their juice taste any better, but it is the ingredients that differentiate the quality and its health benefits. Today’s ingredients are data, bidding logic, cross-device intelligence and the application of machine learning (I won’t say AI), to name a few that enable vendors to create real value for their buy-side partners.
The good guys should always be quick to say they if they are built on top of companies such as AppNexus and IPONWEB, as both provide the educated procurement team with confidence around tech stability and scalability. No longer is third party support for commoditised functions such as bidding deemed a weakness, in the same way I don’t ask myself if the blending machines used at my local store to make my morning smoothie are leased or owned. I care about their ingredients and the output in the form of my drink, and therefore I choose them and not their competition next door.
Ok, but what else can be done to bring pricing transparency to agencies and their advertisers? The majority of OpenRTB vendors operate on a “second price” auction model, where the amount paid by the winning bidder will always be £0.01 more than the second highest bidder, but there any many senior industry figures in ad tech that are now talking about the benefits of moving from a second price to first price model. I am far from an auction dynamics expert but the three benefits I see for our industry moving to a first price standard are as follows:
Many vocal commentators still think consolidation is needed among buy-side tech vendors. Some are hoping this thinning may come post the GDPR, however agencies time and time again voice their frustrations around having to use multiple DSPs for a single campaign. I personally think we are going to see more DSPs, not fewer, as the race for ID management heats up and walled gardens close off access to third party bidders and keep their inventory or data held back for customers using their own data loaded DSP (think of YouTube’s supply for DBM customers only etc.).
With this in mind, for anybody in the process of trying to differentiate or evaluate new programmatic buy-side partners, I would recommend asking questions in the following areas:
It could be a good idea as an industry to open source a buy- and sell-side RFP template. Enabling agencies and publisher to really get under the skin of the correct questions to be asking. There are no precedents for today’s data driven scenarios we are planning for, so collaboration and knowledge sharing are a must in order to smoke out the bad actors and to navigate programmatic away from the negative headlines it is currently generating.
To summarise, we know there are a lot of VC-backed, privately-owned buy-side ad tech vendors in market that at times can struggle to differentiate themselves. We know that agencies are under immense pressure around pricing, and opting for the cheapest ad tech vendor (managed or via SaaS) is not always the best approach.
Yesteryear, pricing could be driven down when dealing with a publisher or media owner directly as the quality of the print or TV ad would not be compromised. However, with programmatic technology you do often get what you pay for and with increased scrutiny around inventory quality and margin transparency, it is wise to select a buy-side partner based upon their ability to differentiate and offer a long-term partnership that enables them to make money so they can continue to build products and features that meet the agencies and their advertisers’ requirements.