Customized Ad Tech Can Give Brands A Competitive Edge

wayneblodwell"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 written by Wayne Blodwell, founder and CEO at The Programmatic Advisory.

One of the interesting trends we will see continue over the coming years is how more advertisers will take ownership of technology contracts rather than using a third party. The use cases are clear: better understanding from a privacy compliance perspective, particularly with the impending EU regulations taking force in 2018, and a need for increased transparency in tech decision-making as brands start to double down on the number of technologies used.

Brands should aspire to having technologies customized to their needs, which allows for a competitive edge. Every brand has unique business challenges yet the tools used to overcome these challenges are likely the same or similar to those of their competitors. The differentiator today is the way in which people operate the technology, which is critical, but there are significant potential gains from customizing technologies around specific business use cases. It’s surprising there are few examples of brands doing this.

Those specific use cases that would benefit from customized tech include real-time budget management across product lines, complex dynamic creative recommendation rule sets across products, audience taxonomy rules across multiple brands and channels, multi-KPI optimization frameworks and creative asset management across channels.

More often than not these kinds of edge cases cannot be solved by existing solutions or off-the-shelf technologies. As a result, brands often need to operate a best-fit mentality with the technologies they work with.

I think many advertisers would be surprised at how sophisticated their programmatic setup is via their third parties. When considering customization, they should start by auditing their existing setup to see if customized technology is commercially viable. If it is, brands should modify existing contracts to assert ownership of needed technology and restructure resources with vendors where customization isn’t viable.

Designing A Custom Solution

Nowadays many technologies work in an open manner, which means brands can create custom solutions and tools on top of theirs.

An example here would be Wayfair and AppNexus. Wayfair wanted to create custom bidding logic based on user intent and its own intelligence on user behavior. Wayfair worked closely with AppNexus to create these custom algorithms on top of the AppNexus infrastructure and has seen positive performance improvements as a result.

If the infrastructure can support customization, brands can then start building their requirements similar to what Wayfair did. The requirements should be split into critical functions and optional functions. Brands should never lose sight of what they are trying to build and what is critical to delivering that, which is why these two lists are important.

Building A Custom Solution

Once a solution is designed around a brand’s needs and the right partners are in place as the foundation, it can be built in many ways.

For example, say Brand X is a digital-only ecommerce brand that specializes in selling DVDs and console games. Its product portfolio has more than 100,000 products and is incredibly data-rich. The brand struggles with product upselling, such as getting a buyer of one game to also purchase another. This can get quite complex from the creative and bidding logic perspectives.

Brand X may have scoped a solution, such as a web-based piece of software, which will allow it to build creative recommendation logic at a product level based on five of a user’s previous purchases. Brand X also scoped the solution so that the bidding is dynamic and based on 10 variables associated with each impression, such as geolocation or recency.

There are multiple ways Brand X can take this solution to a minimal viable product. An internal tech team may build it, for example, or Brand X may work closely with a tech provider to leverage its resources. Or it might hire a third party to build it.

Making that decision is a relatively simple process. Brand X needs to understand the expected value delivered by this solution. Brand X would then collect proposals from potential partners to ascertain cost and expected time for build-out. This would quickly help Brand X ascertain whether this customized tech is commercially viable or if certain changes must be made to reduce the costs of getting it to the minimal viable product stage. Timing may also change.

There are some potential hurdles ahead for Brand X. The time taken to deliver the solution may increase, while the resource needed to deliver the minimal viable product may also grow. Since these hurdles can increase cost, solution budgeting must be considered throughout.

Implementing, Testing And Activation

Once a customized solution is built, brands should ensure they have a roadmap for testing. The one thing anyone wants is to turn on a whizz-bang tool and it breaks on the first day because the user or back-end functionality was never tested.

I believe we will see more and bigger brands seizing a competitive edge from custom technology. When many brands feel the pinch due to an increased competitor set and digital disruption, they should try and get every edge they can.

Follow Wayne Blodwell (@wayneblodwell) and AdExchanger (@adexchanger) on Twitter.

1 Comment

  1. Thank you Wayne for sharing your analysis about a major market trend. Now that everyone is doing programmatic, how you leverage your data and technology becomes a real differentiator when striving to reach your performance or branding objectives. The example of Wayfair is highly relevant : leveraging the Appnexus Programmable Bidder is obviously for this brand a game changer and a competitive edge.

    Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>