There has been a rise in content creator all around the world, espicially in Canada1. Content creator, sometime known as influencers, is where one generate money through the use of content to communicate with their target audience2.
For content creator, they usually need to find ideas to generate content for the right target audience to earn a living. While doing so, many of them faced creator's block. It causes them to unable to generate creative content/ideas 3 and this will distrpt their source of income.
One of the ways they find idea is to look for current trending topics or things and Google Trend can be one of the ways to generate ideas for content creators.
For this project, I will be focusing on Canada and Google Trend will be used.
For this project, there are 2 things to find:
- trending searched terms
- top searched terms
With these data, the content creators facing creator's block, they can use the result above to catch the attention of their target audience and at the same time, generate more income.
The Google Trends datasets are taken from Big Query. Big Query is a data warehouse cloud platform developed by Google itself 4 . It is free for public to use and SQL is used to clean and query the required data. The data is then saved as csv file for analysis. As I am using the free version of Big Query, there is a limit to much I can use SQL for cleaning, data wrangling analysis and visualisation. Therefore, I used both Excel's power query & power pivot and tableau for additional cleaning, data wrangling analysis and visualisation.
| Feature | Type | Dataset | Description |
|---|---|---|---|
| country_name | object | Big Query | Country of the dataset. |
| region_name | object | Big Query | Region in Canada. |
| refresh_date | int64 | Big Query | The date of when the data is refreshed. Meaning it is where the new set of terms, country, region and combination of scores are added. |
| week | int64 | Big Query | The first day of the week. It is used as a time series for the analysis. |
| rank | int64 | Big Query | The rank of the top 25 terms for the current dataset, which is based on the refresh date. |
| percent_gain | int64 | Big Query | Percentage gain (rate) at which term rose compared to previous date period. |
| term | object | Big Query | The human readable identifier for a term, i.e. 'Acme Inc'. |
| score | int64 | Big Query | Presents the maximum search interest of the term for the time and location selected. The index of the value 100 will imply that in the specified time range, a specific term was trending mostly where the peek is, i.e. where the score was highest. |
| category | object | Excel | Category of the term. |
| latitude | int64 | Excel | The latitude of the region. |
| longitude | int64 | Excel | The longitude of the region. |
Check out the dashboard that was done using Tableau!
I have also link a photo of a sample of the google trend from googe for comparsion!
Big Query google trend's dataset only stores the top current 25 terms based on the refreshed date. Meaning, any terms that is less than 25 is not analysed and any terms that is previously in the top 25 terms but not in the current terms are not included in the analysis. In other words, I am only able to analyse the current top 25 terms and the timeline of the current 25 terms. As there was a change in google's data collection system in 2022, the data before 2022 might be not. Therefore, more studies and data are required. Previous years were not taken in to consideration as comsumer behavior have changed after covid 5. Therefore, results based on previous years are not used and the dataset only consist of datas after 1st Jan 2022.
Based on the dashboard, the following are the findings:
- "Premier League" or "EPL", which is a english football league system 6, is the most searched term in April 2023
- "Blue Jays vs Angels", which is a MBL (Major League Baseball) team match 7, is the fastest rising searched term in April 2023
This means that content creator can generate idea based on EPL or MBL to attact more target audience and therefore, increase their income.
To further improve the results, a recommender system based on machine learning model (Supervised, UnSupervised and deep learning) can be impletmented and it will be another forthcoming project in a foreseeable future.