Middleware, in the ad business context, stands for a layer of companies that came about and got sandwiched between the demand side and the supply side, broadly speaking. Suffice to say, this is a layer that came about opportunistically due to gaps in the yarn of the demand-supply fabric that, in an ideal world, should have had the buyers closely intercoupled with the sellers. Middleware had its humble origins in the 2000s, where ad-sales houses started packaging inventory from the large players such as Yahoo, MSN and a string of other larger publishers, with some value-adds like reporting dashboards and optimisation.
And then during the following decade, middleware exploded when programmatic widened the chasm between the buyer and seller worlds.
Middleware grew around display
In the 2000s, the buy-side and sell-side were tightly coupled through close human interactions, but the following decade saw this relationship fall away due to the emergence of the duopoly and programmatic buying. A big rift grew between the real sellers/sources of inventory and the buyers, with the creation of tech platforms that would make programmatic transactions possible, rather than human beings. These included demand-side platforms (DSPs) and supply-side platforms (SSPs) to begin with, followed by a long procession of adtech vendors: DMP, brand safety, IVT, rich media, dashboards, crowdsourcing, retargeting, tag management, privacy—the list is endless).
The rift became a chasm when Google brought together the trifecta of GAM (erstwhile DFP), AdX (dominant SSP) and Adsense (network buyer marketplace for dirt-cheap inventory) along with DV360 (dominant DSP), while Facebook created its own version via Facebook ads platform (and Facebook Audience Network). This separated the real sellers from the buyers forever.
While this grand domination drama was playing out, creating unfair advantage for the duopoly and chaos for web publishers, the middleware vendors multiplied exponentially, adding more chaos into the overall adtech ecosystem, primarily due to their own offerings trying to bridge different gaps in the ecosystem that had been created by the duopoly—but to their own benefit, of course. Dig up a Lumascape from that era, and you realise what I am alluding to. The industry was shapeshifting at lightspeed, such that even the best of the experts would get dizzy (below).
And remember, I've been talking only about display advertising.
Cut to the present decade, and the Lumascape isn't looking any better because of the way Covid has changed the entire landscape: the emergence of video as the new power, Google backtracking on FloC as expected, Apple's sledgehammer privacy policies, Google sun-setting third-party cookies, half-baked solutions from the industry in the ID space, SSP consolidation, crypto in adtech and more.
I would like to force myself to believe that these are the throes of just another evolution.
Where does this leave middleware vendors, and is there a structure to any of this?
Those that hitherto focused too much on issues that are not solvable will see what's coming and shapeshift or perish. Will there be any structure to the middleware? The answer is a resounding 'no', because middleware itself is the progeny of the polarisation of advertiser budgets to certain quarters (read: duopoly), and the lack of structure that emerged out of it.
Take any of the problems middleware has been professing to solve—privacy, invalid traffic, ad fraud, brand safety, creative solutions, reporting and analytics, dashboards—all of these emerged because of the tech that was created by the duopoly and the resulting independent programmatic landscape. And I say 'professing' because most of them fail even before they start out, given that there can’t be half-baked solutions to issues that are critical to the industry.
Take the case of multi-channel attribution. Google analytics has always been a clear winner here because of the stronghold it has over the data flows it has from paid, owned, and earned media (DV360, Google search, YouTube, Gmail, Google Display Network, which includes every web publisher that came to be). None of this data will ever see the light of the day outside the 'walled garden', and so how would a middleware vendor even attempt to offer an alternative? This is exactly the reason that 'data-based' middleware companies will continue to find themselves unable to scale. The most successful one in data-based middleware could potentially be a robust dashboard solution (if it can wrestle business away from Google Data Studio).
This has been and will always be the challenge for a lot of middleware vendors.
How will adtech middleware evolve?
The adtech middleware world may develop some semblance of a structure when (and if) independent buy-side tech gains ground and buyers start to funnel dollars via such tech. Let's take the case of The Trade Desk, which is trying to unify the ecosystem via UID 2.0. It's a gallant effort to get the open web to embrace a solution that can eventually provide the DSP with a unified deterministic ID. In this scenario, the actual sources of the inventory (web publishers) are able to pass their user data to an independent DSP, and in that process bring the original sellers closer to the buyers (although the world will never go back to the early 2000s, especially if the independent tech becomes powerful enough to erect new walls between the seller and buyer).
In this new world, where there is a successful independent DSP partnering with an equally powerful independent SSP that represents the open world, that is when the middleware tech could eventually start to provide solutions that cement and augment the DSP-SSP coupling to create a holistic ecosystem that is democratic and fair to both the buyers and sellers.
As this evolution transpires at a global level, how deep would this evolved offering go in APAC markets? This is the million-dollar question, purely due to the complex environments and varied maturity stages of APAC markets. Reflective of that, the middleware space in APAC has always been an intimidating panoply of vendors, mainly due to the proliferation of opportunistic scenarios created due to these variegated large markets (India, China, Japan, Southeast Asia, Australia).
What opportunities might middleware tech address?
The biggest opportunity that I see, situated right in the eye of the video storm, is that of ad monetisation of 24/7 livestreaming TV/VOD through ad-stitching solutions: bespoke players offering adtech capabilities around short-form and long-form content programming, creative, data, reporting, data and analytics. This would cement and augment the play of the evolving DSP-SSP independent adtech ecosystem. Of course, there are many other fringe solutions that will emerge, but the above would certainly be a space to watch.
And so continues the grand theatre of the big bad adtech world.
Hari Shankar, currently chief revenue officer with IVS (Intelligent Video Solutions) and a former executive with Performics, Singapore Media Exchange, Paypal, GroupM and Havas, has been interpreting the evolution of the digital landscape via Campaign Asia-Pacific for many years.