Development of Indicators for the Media and/or the Cultural and Creative Industry.
Disclaimer: This is a fictional project, I'm not part of Podium Podcast's marketing team, I'm just a fan of their podcasts.
As a leading company in the world of Podcast in Spanish, we are the marketing team of Podium Podcast's and we went to Pol Gubau's studio as part of his audience analysis program to solve a couple of doubts that have arisen in order to update our content in the coming months.
Who are we?
Podium is an exclusive sound content platform part of the Prisa group where we broadcast different kinds of programs, but always based on the highest quality and standards.
Our website was published in the middle of 2016 and has allowed us to reach a large and solid network and listeners in different countries of Latin America.
Mobile development
Besides our web platform, we have also developed a mobile app for both IOS and Android where you can listen, download episodes as well as read the blog or exclusive content, even so, our users continue to prefer the web by number of visits.
Our podcasts
We also have a strong presence on Spotify and its podcast rankings.
How are we different from the competition? Apart from having great contributions to the podcast world, we are totally free, you don't need a membership to join our platform or to download episodes.
Are you familiar with Estirando el Chile, La Escóbula de la Compassula or Aquí hay Dragones? Then you already know us!
Our problem
We are now at an interesting point in time, we have completed several successful narratives and we are looking at where and how to expand our product, which product, what categories of podcasts are most interesting to the audience and what can bring us more listeners as well as not losing the current ones.
Our audience
Following the guidelines that Pol made us see, we will have to plot some indicators.
First of all, let's select a topic that we proceed to measure: 'We want to find out for which category of podcast there is more interest in future months.'
Now that we know the topic to develop, let's look at the problem we are presented with, how and for what purpose we will analyze it.
Mainly we want to to get ahead of what our listeners want. This is one of the is one of the reasons why Big Data is so important to us as a company, we are not interested in any individualized user case, but it is crucial for us to know the consumption patterns of our platform.
The solution
This is the main reason why we developed a mobile application and a website where you have to login. Spotify gives us the listeners number and rankings, but we don't really know the users on this platform.
Having our own website allows us to store user data, to see which podcasts have been listened to the most, but also which have been rated the highest, which ones have been abandoned before...
From what we see in our basic Spotify rankings, we see that the most listened to podcasts are informative and entertaining talk shows with several young people, so our initial hypothesis is that if we make another podcast with these characteristics will also be liked. As indicators we have used the number of reproductions of past podcasts. Our information between category and popularity allows us to see where we fit.
Let's see an example of what it would look like:
Disclaimer: I don't have access to Podium Podcast's data, so these are aproximated numbers.
To analyze these numbers we do not need complex data, only with the number of listeners per year we can make a graph and see the trend of these categories.
The future
We can see how the estimates for 2024 and 2025 are to dedicate more funds to Entertainment than to Horror.
We can also look at how to analyze more complex data, for example, let's say we want to analyze which episodes our users have liked the most.
How do we analyze this? There is no "rate this episode" button in Spotify, but even if there was it would be unbiased because you would vote right after watching that episode without having recently watched the previous ones. So we will have to use to use more concrete objective data.
How do we know if a user likes an episode?
We can highlight the following:
-
Which episodes have been listened to the most (per unique listener and in total, to see which episodes people listen to multiple times)
-
How long they take to remove the episodes, the closer they get to the end of the episode, the more entertaining or interesting they find it.
-
We can also take into account the ratings on platforms like Ivoox that allow a rating system as well as positive comments.
How do we know that the comments are positive?
We can resort to the usual AI to classify us between haters and lovers, as for example Oracle's OCI Language.
By combining and analyzing these multiple simple indicators we will reach much more complex and ophisticated ones on which to base the decisions we make as a company. as a company.
Conclusions
We have seen how we can use simple indicators to make decisions, but also how we can use more complex ones to make more accurate decisions. And this would be a summary of our strategies, thank you very much for reading and as we always say, the best is yet to come.