1 of 11

SOCIAL BUZZ

CONTENT CATEGORY ANALYSIS

By Claudia Aoro

For Accenture

Claudia Aoro

2 of 11

Agenda

Project recap

Problem

The Analytics team

Process

Insights

Summary

Claudia Aoro

3 of 11

Project Recap

Accenture is on a 3 month pilot with Social Buzz to focus on 3 tasks aligned with the challenges that Social Buzz is facing.

We are developing ways to help you manage the scale of your business, having reached a greater audience in recent years.

To start, we will take part in;

Auditing your big data practice.
Guiding you through a successful IPO
Analyzing your data for top 5 popular content categories.

Claudia Aoro

4 of 11

Problem

1 day

100,000 posts

36.5M posts

Per Year

Of Unstructured Data

Claudia Aoro

5 of 11

The Analytics Team

Andrew Fleming

Chief Technical Architect

Marcus Rompton

Senior Data Expert

Claudia Aoro

Data Analyst

Claudia Aoro

6 of 11

Process

Data understanding

Data extraction

Data Model Processing

Visualization

Data-driven

recommendations

Claudia Aoro

So, how did we tackle this problem?

Well, we approached it in 5 steps:

1. Data understanding - the key to success on any data project is to understand the data in detail. So we took the time to understand the data model and domain of your business.

2. Data extraction - after understanding your business, we then architected what an ideal dataset should look like for this problem and extracted it from the relevant data sources.

3. After extracting the raw data, we needed to process and model this data into a dataset that can precisely answer the business questions and produce analytics.

4. With our new dataset, we used our analytical expertise to uncover insights from this dataset and to produce visualizations to describe the insights.

5. And finally we used these insights to unlock business decisions and to make recommendations on next steps.

7 of 11

Insights

Number of Unique Categories= 16

The split of categories of posts of the users shows an almost equal percentage of between 6 and 7%

Top 3 Categories with the most Watch time are Travel, Technology and Animals.

Claudia Aoro

From your data we found that you had a total of 16 unique categories of posts across your sample dataset. This includes things such as Food, Culture and Sport.

The split of categories by posts of the users shows an almost equal percentage of between 6 and 7%

Take note that this is totally different from the popularity of the categories as this is only by content count.

From this chart, we can see that the top count of posts by % is taken up by technology, animals and travel, in the next slide we’ll be looking at the popularity of these more clearly despite the numbers.

Also, it is interesting to realize that the content categories that people are spending most of their time to watch. Travel takes the lead watch time followed by technology.

But now, onto the main question... which is... what were the top 5 most popular categories of posts?

8 of 11

Insights cont’d

Top 5 Categories are;

Animals
Science
Healthy eating
Technology
Food

There is only a thin 1% margin in the popularity scores of these categories.

Healthy eating and Food : Science and Technology indicates specific interests within larger categories which directs us to the bull’s eye of content popularity.

Claudia Aoro

From our analysis you can see that the top 5 most popular categories of posts were Animals, science, healthy eating, technology and food in descending order.

Healthy eating and food both appear in the top 5 categories, which means that the users are interested in consuming but also being keen for what is healthy for them. Equally, science and technology also appear here. A field of interests with a specific type within it- bet we all know where to maximize on revenue channels like ad space for what most people are into.

A split of this shows the thin margin by percentage of the popularity of these categories. Animals at the top is only 1% higher than the following categories.

Also interesting to note that although, travel had the highest watch time, it is not among the top 5 most popular. You might wonder why this is so.

9 of 11

Insights cont’d

Content type with most popularity is Photos, followed by Video, GIF and finally audio.

Most development could therefore be made in photo quality as this is what users like most.

Claudia Aoro

10 of 11

Summary

Key data points derived;

Top 5 most popular categories
Popularity of content type
Top watch time by category

Claudia Aoro

11 of 11

Thank You!

Any Questions?

Claudia Aoro