Over the years, digital marketing has started transitioning to audience buying from site buying. Agencies can target audiences with relevant ads as per the choice of viewers in their online journey.
In an e-meeting with the media, Bharat Khatri, Country Head of Xaxis India, discussed how to move to outcomes in 2020 with new-age programmatic.
Khatri mentioned about the 4Ps of new-age programmatic, improved artificial intelligence, and omnichannel and personalised approach can build brand expectations on business outcomes.
According to Khatri, brands are moving towards performance marketing for business outcomes. Application-based platforms such as Urban Clap or Zomato are running videos on digital not to get impressions but to get business outcomes. Not just by getting installed but the final result is getting subscriptions or product purchase.
4Ps of new-age programmatic: In ( ) is a digital media landscape point of view
People (recent updated and relevant) is data.
Product (technology) that makes a difference
Platforms with (quality) are media.
Performance (real-time) is measurement.
How technology works
In the image below, on the left is an advertiser and on the right is a publisher. In between is technology, which is trying to connect the requirement from one side to the other.
Example: Pepsi is an advertiser that has an agency called Mindshare; they are using an ad server like DCM, which is a Google double click campaign manager to serve their ads.
The server is buying through a demand-side platform where it will put a demand for targeting youth of the 18-20 age group, watching music on YouTube because the advertiser has created a new music video. Then the demand-side platform will put the demand out in the pocket and the advertiser can buy the audiences directly through the publisher, ad networks or Google ad exchange.
How to move to outcomes in 2020
Omnichannel approach: Brands have to think holistically and leverage multi-channel synergies for a holistic approach to its advertising strategy. “Personalised and targeted communication is good, but brands will have to make its presence felt at all the places a consumer is spending his time,” Khatri said.
Gather Data: Get your data together and understand how to activate first-party data.
Impressions on digital campaigns give data obtained from the consumers, like what handsets they are using, the speed of the internet, the network they are using.
These data help in mapping consumer behaviour and actions towards an ad. Khatri said that each impression of a digital campaign is consuming about 2000 data points about the consumer.
Personalise effectively: Personalise your campaign to speak to your audience.
Leverage Artificial Intelligence: Leverage AI technology to surface insights and achieve better optimisation towards business outcomes.
Khatri said that improved AI consumes data, makes predictions, and takes further action for output, which businesses are looking. Machine learning will only consume data and make predictions. “AI doesn’t take personal information, it only asks for ad ID, cookie ID or obtains event IDs,” said Khatri.
The four data types agencies work on are device data, engineered data, advertiser data and event data.
“Any advertiser who runs different ads with different creative colour, subject matters, etc., have segmented their ads into a different bucket like fashion, auto and travel. Whenever anyone likes any of those ads an ‘ad ID’ gets generated at the back end,” Khatri said.
Ad ID is an identifier that helps understanding consumer behaviour which they are doing in the in-app environment. Earlier, the agencies were using cookies to identify consumer activity but the cookie was only able to use the searches in the browser environment.
“The way the market is changing in India most of the time, people are spending their time in in-app. To track consumer behaviour in the app environment, agencies need consumers’ ad ID,” he said.
Through ad ID, Google can send personalised ads.
Xaxis has moved towards advertising ID and is able to understand the mobile behaviour and lot more elements along with the location. “We combine both offline and online and create the outcome of product purchases,” said Khatri.
According to him, event data keeps a track on time of the day, day of the week when the engagement has happened. The agency and advertisers get to understand the position of the ads.
In a typical environment, there is a lot of ad on the desktop and mobile; event data helps the agency to serve the ads in a better and lesser manner.
Xaxis has its in-house proprietary AI engine, Co-pilot, which optimises different campaigns and delivers better efficiency. “Co-pilot right now works on all beddable media, where we understand all sort of historical data, people in the past who have visited the brands’ website or made a purchase. What was there ad behaviour, what kind of ad they like, the content they are watching and mobile they were using. Using that information, we can train our AI engine to produce an output and deliver the best outcome for the advertisers,” Khatri said.
Xaxis is working with lots of brands in India to define its Custom Outcome Indicator.
AI turns Input into output.
In the image Input, A is the raw feature, and output B is the output AI is making with models, a trained algorithm in our daily lives.
Now take Input A as the data points and output B as the output a brand is looking. Model and algorithm is the mathematics in between, which is taking all the inputs and converting them into output.
Khatri explained that working with models gives the best results as models are trained algorithms that make predictions and has converted a lot of inputs into outputs.
Businesses to Find the Signal