It is estimated that, by the year 2020, 34 billion devices will connect to the internet, a scale so massive that regulation and control by a single company will be near impossible. Given that no one organisation will dictate how the internet functions or what it looks like, within those 34 billion devices, each of us will interact with our own, highly-personalised virtual environment.

As such, we should not be thinking about the internet as if we are all navigating the same place. Rather, the experience of ‘going online’ is developing into something unique and malleable. This complex new ecosystem will bring many challenges for marketers. Beyond providing consumers with personalised content, they will need to customise this content to complement individual online experiences from moment to moment. Not simple.

Earlier in 2016, my CEO Brian O’Kelley declared that ‘programmatic [advertising] is dead.’ While swift and scalable, programmatic was equally limited in scope, failing to provide consumers with an online marketing experience that was as individual as their behavior. Internet marketing of the future, he predicted, would be ‘programmable.’ But what does that look like, and how does it work?

Broadly speaking, programmable technology is programmatic plus intelligence – that is, programmatic, with the addition of powerful, algorithmic technology that enables the customisation of advertising – or, 1:1, personal engagement.The programmable internet is more intelligent than the static internet, and the sum of its possibilities for advertising is far richer and more robust.

In order to keep up with the dynamic ecosystem of music and video streaming, interactive gaming, social networks, audience platforms and the Internet of Things, successful programmable internet marketing requires several elements:

#1 Own your real-time data

Marketers and publishers should use their proprietary data, rather than relying on third parties, in order to most effectively acquire, engage and monetise their audience. Whether it is used for the creation of new campaigns, new messages or even new products, data plays a key role in the creation and targeting of programmable advertising, as well as enabling better user engagement and more sophisticated yield management.

#2 Leverage algorithms and machine learning

Marketers need to go beyond reliance on human intuition if they are to consistently match the right creatives with the right people at the right moment. Algorithms, machine learning and audience data bring real-time insights to the programmatic process, exceeding the capabilities of an individual marketer or marketing team. Brands that embrace machine learning will be able to engage and monetise their consumers more intelligently. In other words, we need to leverage machine learning to achieve human intuition at scale.

#3 Be part of an ecosystem

The age of the programmable internet needs a dynamic and open ecosystem in which brands can work with tech vendors who help them create seamless, personalised experiences for each consumer. Brands that give their data over to walled gardens, or even third parties, hand over the keys to their castle. Instead marketers need to the opportunity to work with partners who transparently enhance their offerings.

When we started talking about programmatic and real-time bidding ten years ago, many people responded with muted enthusiasm, but the figures show that 90 percent of marketers in the UK are now leveraging programmatic. The future is programmable; those who leverage programmability themselves and through relevant partners will have ultimate say in shaping the consumer internet of tomorrow.

Nigel Gilbert

Nigel Gilbert


Nigel Gilbert leads AppNexus in Europe.