Learning About Our Networks

Introduction and Objectives

At Katha, we believe that our biggest strength lies in the network we have built and the data we have accumulated so far. At over 15 million, our network is currently as big and more active compared to that of Meta. And the 60 million impressions that we have clocked in the past 6 months have helped us understand a lot about our network, the underlying demographics and interests.

This blog post is about how we ran an experiment to validate our understanding of our audience. We were looking to learn further about their purchasing patterns, and in turn, improve our network segmentation.

Okay, so what did we do?

Over our 6-month advertising journey, for every campaign, we captured information on,

• How many users from each network are viewing our campaigns.
• Engagement from each network.

But to understand our networks' purchasing patterns, we wanted to be able to capture sales data as well. Collecting this from our existing clients meant a delay of over two weeks and a lot of manual effort. Hence we turned to the Amazon Affiliate Program.

We generated an Amazon Affiliate link for Katha across 6 different categories, and posted all Kerala Ads on our platform, requesting our audience to share it in their network.

What did we observe?

We hit an overall 6.5 Lakh impressions across 6 categories, receiving a total of over 22,000 clicks and sales of over Rs. 3.3 Lakh. And of course, the platform learnt a lot about our audience.

What did we learn?

• Birds of a feather flock together Engagement data clearly showed that every network had a favourite segment. For brands, this would mean that network based targeting is as (or more) important than demographics based targeting.
• Consistent traffic across segments The platform was able to drive engagement and traffic uniformly across all categories.
• Sales hinge on brand presence, recall, and established online purchasing behaviour. Most of the sales happened in categories where e-Commerce was already the norm. In other categories, where the purchase is based on brand recall or offline buying experience, engagement did not mostly lead to sales.

In conclusion, Katha's experiment revealed the importance of network-based targeting for brands, consistent engagement across categories, and the significant impact of brand presence and online purchasing behaviour on sales.