How to reduce fragmentations in AdTech?

The current AdTech industry looks a lot like a "flea market".

Many of the adtech vendors such as the buying platforms and the data platforms are offering very similar products and services. The decision to differentiate often falls on pricing and salesmanship. On top of that, due to a highly fragmented technology stack, advertisers often have to deal with a conglomerate of various technologies just to get something started.   Simply look at number of account managers (AM) advertisers have to work with on a daily basis. There is an AM to talk to about serving the ad, another AM for buying data, another for attribution, and the list goes on. The promise of programmatic marketing is great, but in today’s implementation, it is very complex, inconsistent, human driven and disconnected; as the result, many campaigns end with poor experiences and inefficient results.

A root cause for this fragmentation comes from the fundamental operation model of demand platforms.   While on the surface, it appears that the DSPs give the flexibility to allow media buyers to choose from various vendors such as data, inventory sources, creative, and so forth, but at its core, because of the lack of quality control and transparency, the flexibility actually translates to fragmentation and inefficiencies, and makes it really costly to determine the optimal stack for each campaign execution. As an interesting trend, a leading DSP recently announce a new marketplace for campaign models and algorithms. Campaign model building is an iterative process, building models without control over data sources is like building transmissions without knowing what the engine is. As the industry is getting more disconnected, the client spends a lot testing through various permutations of technology in the stack, but don’t see any increment results. The media budget has been increasing YOY, but there is much less or none to brag about on incremental conversions.

In terms of vendor transparency and quality control, let’s start with data vendors. Many of the data segments in the marketplace sound very much like a new flavored drink. They all sound very good, but you have no idea how it’s made. Clients have no idea what count as an in-market audience, and whether the definition of in market applies to all advertisers and industries. To test against them is daunting and costly, and eventually retargeting takes over.

So how do we start tackling this fragmentation problem?

1. Inventory transparency. Domain masking simply hurts the programmatic buying environment and creates inventory frauds.   Get rid of domain masking, and pass more attributes about the placements such as slot positioning and number of ad units on the page.
2. Data transparency. Data vendor has to reveal how data are collected, the brand associated with the data and how it gets packaged into and taken out of segment. Without those pre-conditions, unknown data could be bad data. And bad data is worse than no data.
3. Measuring ad effectiveness. Ad effectiveness can be measured instead of attributed, and that is shockingly missing in many platforms today. Run a control test to measure whether people seeing the ad buys more than people not seeing the ad. That is the incrementality to measure your marketing and how you should price.
4. Actionable analytics and insights. Running media without analytics is like sailing a ship without a compass. Good analytics, especially industry specific analytics can facilitate true data driven marketing. And those insights have to be actionable by instantly creating targeting around them.
5. Integrating data with creative. Leveraging data in creative allows clients to tailored the creative to the interest of the user. The offer could be the same, but how to display the creative leveraging the specific interest of the user will drive superior results.

Interested in being part of the evolution to re-invent programmatic marketing? We are working on that at ADARA, one industry at a time, starting with travel.