Some feathers were ruffled at the beginning of this year when I suggested that in 2019, anyone buying media should be building their own algorithm. But with many media companies still running their programmatic advertising campaigns through a ‘one-size-fits-all’ or ‘off-the-shelf’ Demand Side Platform (DSP), is it unreasonable to expect brands to be able to do this?
So as we approach the half-way point for 2019, it seemed apt to share an example of a company building their own algorithms in practice - and the benefits that follow.
Typically, brands and media companies look to build their own algorithms in order to improve the performance of their campaigns. This often happens through optimising to non-standard or proprietary data points, allowing them to customise their buying to their own needs.
And, since they’re buying more efficiently, focused on their unique data, they simultaneously reduce their media-buying costs. Ultimately, a successful strategy boils down to collaborating to find the right ingredients to create the optimal programmatic buying strategy.
It’s a recipe that isn’t pre-packaged, as brands are extremely aware, involving significant effort, technology, people, and cost in order to execute impactful data-driven programmatic advertising.
Indie media buying and consultancy, Iotec, is a solid example of a DSP looking to empower the advertisers buying on their platform with precisely the recipe they need for success.
Originally, Iotec built its own, proprietary DSP back in 2014, recognising early-on the value of having its own proprietary decisioning. However, the team there wanted to expand it beyond its existing functionality, with the hope of further developing the businesses' unique selling points: bespoke machine learning technology that set ethical standards throughout the ad buying process. The catch, however, was that this feat would only be possible if it was able to maximise the efficiency of its DSP.
Understanding the importance of transparency and data science to drive incremental customer value, Iotec made the decision to switch to Beeswax's bespoke bidder build.
Cost savings and efficiencies
Iotec's operations and infrastructure costs have significantly reduced since then. Typically, DSPs have three fundamental sets of functionalities:
- Connecting to supply sources
- Campaign set-up and targeting
- Decisioning around bidding.
While the first two of these are tricky (and expensive) to maintain, they don’t offer any real, tangible advantage to a buyer. The true benefit comes when the buyer is able to make much smarter decisions in a highly accurate way on when, where and how much to bid. With Beeswax taking care of the first two functionalities, the Iotec team had time to focus on the bespoke machine learning technology to drive better performance for their customers.
Since the switch, Iotec has saved 90% of its previous costs surrounding operations and support, and 100% of the costs associated with development and integrations. This significant reduction has allowed them to invest and focus in the areas that will allow them to make a difference to their clients.
With Beeswax taking care of the demand side infrastructure, iotec’s Data Science team was able to build custom algorithms for each of their advertisers. This bespoke machine-learning technology, deployed brand-by-brand to iotec’s customers has improved iotec’s retargeting performance by extending their reach between 20-50%.
What’s more, with full transparency into every auction, iotec is able to fine-tune their algorithms in real-time. The “machine” generates intelligent insights, enabling the team to support their marketers in building a far clearer picture of their audiences and the factors that drive them. This approach has empowered iotec’s brand clients to move away from a “one-size-fits-all” approach and move towards personalised creative and messaging to effectively provide more engaging and more relevant content to consumers.
Looking to the future
As more and more brands increase their investment in programmatic advertising, they are also increasingly looking for technology that they can customise to suit their unique needs.
Some brands are keen to try taking this in-house, but for others, the right move is to find a partner with the expertise and skills to build custom, bespoke algorithms with their business needs at the very heart.
While the in-house movement still fills industry headlines, the realisation from many that blindly taking everything internally is costly, complex and lengthy, is becoming ever more apparent.
Despite the large sums of money invested in innovative technology to help drive personalisation, brands and marketers are still struggling to see returns on investment. And it boils down to one core challenge; expertise. Blending both technology and human intelligence is the future for brands who want to stay relevant and keep afloat of consumer demand.
As 2019 continues, we are starting to see brands make smarter changes around their data capabilities, leaving media companies having to step up to the plate to meet the requirements of their customers.
Cadi Jones is commercial director for EMEA at Beeswax