The ‘RTB’ at the heart of Digital Marketing

Tata Digital’s Akshay Salaria shares his perspective on how digital advertising really works and why digital spending is steadily increasing


In 2021, India saw its advertising industry rebound from a pandemic hit in 2020. The advertising industry in 2019 was at Rs 68,475 cr which fell to Rs 59,619 cr in 2020. But in 2021, the industry rebounded with revenues of Rs 70,715 cr. 30% of this spending in 2021 was on digital. The share of spending on digital has steadily increased from 12% in 2016 to 30% in 2021. This has largely come at the expense of traditional media like print, radio and TV .

MEASURABLE DATA: A distinctive feature where digital media is different from other media sources is that it is measurable. It works in favor of digital most of the time and sometimes against digital. Every impression a user sees, every click, app sign-up or website purchase can be tracked.

A second feature is the targeting capabilities available digitally. No other media channel even comes close to the capabilities of digital. The list of targeting parameters is endless. I would group these targeting parameters into three parts: demographic, interest-based, and behavioral. “Demographic” targeting includes age, gender, location, etc. Targeting “by centers of interest” concerns the personality, the choices and the interests of the customer. This is closer to the theoretical parameters of psychographic targeting. These interests can range from subjects like shoes, gardening, paintings to rocket science. Targeting parameters also include people who are currently actively searching for a product or service. Ad networks can now differentiate a user who is researching the internet to buy a car from a car-enthusiastic user who keeps up to date with the latest developments in the automotive industry. Finally, there are “behaviour” based targeting parameters which include device used, mobile network, friends/connections a person has, education levels, marital status, etc.


TARGET GROUPS: Each advertiser can choose their own targeting parameters in each of the three categories to reach their target group. Then these can be mixed and matched to create a combination to further narrow the segment and therefore reach even more relevant users. For example, a brand can reach this target segment: men, aged 25-35, living in Madhya Pradesh, interested in sports, leaning towards popular brands like Nike, reebok, owning a high-end phone and doing part of the higher income groups. And this is just one example.

There are endless such combinations that each advertiser can create and use to target their customers for various products and services. In this case, there could be a user who would fall into the target segment of many advertisers. So every time you scroll through your Instagram feed or browse a website, you see ads from multiple advertisers. Indeed, all of these advertisers felt that you were the right person to contact for their products and services.

If you read news on a website where Google is responsible for serving ads by the news publisher, how does Google decide which ad should be shown to you? This is sort of a classic “n is to n” problem where n advertisers want to show ads to n users. To simplify, google has to solve two problems. First: when a person is deemed eligible to display an ad, Google must determine which advertisers are targeting that person. Second: he must then decide who, among the advertisers, will have the possibility of showing his ad.

AUCTION: In my opinion, the real reason for the success of digital is the way these companies (Facebook, Google etc.) have managed to solve the second problem. They found a way to decide which advertiser among n competing advertisers can show their ad. This process is called an “ad auction”. What is an ad auction and how does it work? Let’s start with an analogy.

Reliance is India’s largest company by market capitalization. Many people buy or sell addiction stocks at different prices based on their trading preferences on the stock exchange. Many sellers would like to sell dependency stock at an offer price of X or greater than X. Similarly, many buyers would like to buy dependency stock at an auction price of Y or less than Y. Exchanges hold auctions to match the buyer with the seller by matching X and Y whenever possible, thus completing an exchange. This pairing, if done effectively, is a win-win for all three parties – the buyer, the seller, and the exchange. The buyer can buy at or below the bid price, the seller can sell at or above the bid price, and the exchange is able to satisfy its customers who would continue to use the exchange for future transactions.

Conceptually, an ad auction works on a similar principle called “Real Time Bidding” (RTB). There is one bid for each ad impression served to each user. First, all advertisements from advertisers are identified in the target segment to which a particular user belongs. These ads then enter the ad auction. Unlike the case of the stock market where the amount is the only decisive factor in deciding who can buy the stock, this auction is more complicated. It takes multiple signals which include bid amount, ad quality, ad performance history, advertiser website performance, user location, device and many more. ‘others. It is not clear which of them has what weight in the decision-making algorithm, but they are certainly all very important from the advertiser’s point of view.

This system is also advantageous for all three parties: the user, the advertiser and the advertising network. The user sees the most relevant and contextually served ad, the advertiser reaches their target segment and has a fair chance to serve their ad, and because both user and advertiser are satisfied, both continue to use the advertising network.

WHAT WORKS? To sum up, digital marketing works because each advertiser can identify which user segment to target for their ads, and the ad network ensures that the ads are actually served to those users. And all of this is accomplished by a system, at the heart of which is ‘Real Time Bidding’.

(Data source: Dentsu Digital Report 2022).

Disclaimer: Content Produced by Mediawire