Interviews, insight & analysis on digital media & marketing

The future of personalised video advertising

Matt Shearer, Product Director, Data Language

Successful hyper-personalisation at scale has been the ultimate goal for both B2B and B2C marketers over the past ten years. Now, the rumblings of change are coming. Consumers are asking, and trying to answer, hard questions about balancing the usefulness of personalised content and their data privacy. Businesses are struggling with the same concepts, with the added motivation of increased sales and revenue.

Although marketing personalisation has been a resounding success in monetary terms, increasing ROI on marketing spend by up to 8 times and lifting sales by 10% or more, according to McKinsey, it isn’t always smooth sailing. In 2018, 75% of US consumers said they found personalised adverts at least somewhat creepy, and the conspiracy theory that Facebook listens to your private conversations and uses these to display highly targeted adverts remains popular with the platform’s users. 

Another common complaint, often shared on Twitter in exchanges like this, is that personalisation is achieved through data on consumers’ past purchases and website visits. This can lead to consumer frustration when a one-off purchase becomes the focus of hyper-targeted personalised adverts. For businesses, it is a waste of advertising spend. These two challenges of privacy and relevance are heavily intertwined – and advertisers need to resolve them. Happily, the technology to do so already exists. 

Ethical personalisation

Internet users’ understanding of data privacy issues has increased significantly in recent years, and will continue to do so. This has no doubt in part motivated Google’s commitment to phase out support for third-party cookies within the next two years. This will be a huge step-change for the advertising industry, because, as touched on above, personalisation is largely driven by advertisers’ ability to track users’ purchases and website visits. 

There is a better option. Advertisers need to diversify from tracking user activity to instead offering contextual recommendations based on what is relevant to a given consumer in that moment. This satisfies data privacy concerns, and, through use of properly structured and connected metadata, can increase ROI – particularly when it comes to personalised video advertising. 

To be clear, the switch here is that advertisers can deliver relevant messages without snooping or tracking user behaviour, simply by understanding the context of moments within videos. This might not be “personalisation” as marketers currently understand it, but it is more personally relevant, and less disruptive than the classic tracking approach. The key will be in giving video creators the capability to describe what their content is about.

How advertisers will “know” where to advertise – Structured metadata 

Metadata is descriptive data that identifies what your assets and content are about. It is a key, but often overlooked, part of best-of-breed marketing and advertising campaigns. Take an instructional cooking video as an example. Tagging the video with ‘dicing onion’ in an online video player using structured metadata can provide your advertising platform with valuable new context. Not only does the system know that dicing is featured in the video, but also that dicing is a technique, is used in X amount of other recipes hosted in the database, and requires certain types of tools in the process (i.e. a knife). 

This ability to understand the techniques and ingredients within the video, thanks to the backbone of structured metadata, enables advertisers to target consumers at the moment when their purchasing intent is highest. To illustrate, let’s again take the example of a consumer watching an instructional cooking video. At the point at which the chef is chopping onions, a personalised advert for a chef’s knife or onion dicing tool can appear. While the user is watching the video is the time that they are most engaged with the cooking process prior to replicating the recipe themselves – making their purchasing intent highest at this moment. 

This also opens up personalised advertising to go beyond specific goods and services that the user has already searched for. Instead, they can be shown recommendations for what is most relevant to them at that specific point in time. An onion dicing gadget may not be directly featured in the instructional video, for example, but may be of interest to someone learning how to cook. 

A welcome additional benefit of this structured data approach is that it enables advertisers to gather new and highly context-specific analytics on the items that users are interacting with, giving new insights on where the most ROI can be achieved. In an uncertain economy, this level of precision will be crucial for understanding where spending will add the most value. 

Video moment technology enables advertisers to break videos down into a series of snapshots or ‘moments’. This optimises the viewing experience by allowing viewers to flick through to the section of the video they particularly want to see. The combination of structured metadata with video moment technology is particularly powerful. One recent implementation by Jamie Oliver’s marketing team increased click through rates by 4.9% and generated a staggering 9.6% conversion rate with the ecommerce vendor.

Personalisation need not and should not disappear entirely – it’s too valuable to both consumers and businesses. The way that companies approach and implement personalisation programmes, though, must evolve with the times. The technology is available; so what are advertisers waiting for?