Leveraging Keyword Data for COVID-19 Trends Analysis

Amid the recent COVID-19 pandemic, search has never been more at the forefront of highlighting changes in users’ behavior. We have seen dramatic growth in demand for particular topics along with dramatic drops in others. Keeping marketers and search engines alike on their toes, strategic shifts are being made daily to adjust to the ever-changing landscape that we are collectively experiencing. Many SEOs (and SEMs, we are all in this together), the need for new data has exploded. With many search engine providers not supplying COVID-related search data through conventional tools, we need to look to new sources of data that can truly inform strategies and tactics.

Google Trends for Aggregation and Day Parting Analysis

You’ll see many articles regarding COVID search strategies incorporating Google Trends in the wake of necessary keyword data not being present at this time. But, where many analyses fall flat leveraging Google Trends is in the limitations of data utilized. Similar to standard keyword research, you wouldn’t base your strategy on only a small handful of keywords. The power of Google Trends lies in taking a larger data set and hyper focusing on the granular data points in between the lines. And, with rapidly changing schedules due to work from home initiatives, schools being closed and sleep patterns shifting, having an understanding of volatility in search patterns will help you further evolve your strategies.

To take your Trends analysis to the next level, you can use the unofficial Google Trends API, PyTrends. PyTrends allows for the same raw data exports found on Google Trends without having to manually navigate the user interface and submit keywords one at a time. Using a combination of PyTrends and a handful of other Python libraries, we can aggregate a larger keyword base’s demand together, create a pivot from the raw data and visualize the results in a fraction of the time.

Add your keywords into the list (or bring in a column from a CSV using pandas) and set your time range within the PyTrends function. What you’ll end up with is an easy to navigate data visualization, along with a raw data file which you can use for your further analysis.

For those who are looking for continuous real time data over a particular keyword base, you can set this script to run regularly using datetime qualifiers within the PyTrends function, create an automated command to run this script automatically and path the raw data to a SQL server or separate workflow for data ingestion.

Google Keyword Planner for Forecasting Search Demand

Looking to understand and forecast shifts in demand within your verticals? Consider both historical search behavior data on top of existing shifts. While Google Keyword Planner is used far and wide for both SEO and SEM tasks, taking a step further into data outputs can allow us to create views of the month over month shifts to determine future demand. Simply log into Google Ads, input your keyword base and export historical metrics as a CSV for whatever time range you want to determine forecasts. From there, path your data export into the python script below and from there you’ll be given a series of line graphs showing historical month over month percentage changes in search demand for your keyword set.

Similar to Google Trends, you can take this script a step further by incorporating the Keyword Planner API to pull data for your keywords across separate niches or topics of exploration. From there, sync up this script into a Dash dashboard and have it regularly updated on a monthly cadence with newly updated data. Leave this monitor up and continue running even after the pandemic to create more compelling demand forecasts based on pre/post-pandemic activity.

Conclusion

Whether your business is seeing upticks or downticks in activity during the COVID-19 pandemic, there is a need to meet your audience directly on search. The data you begin to collect and analyze today serves as the basis for pointing your organization in the right direction moving forward. So, if you are trying to either optimize a paid search campaign, or looking to develop/optimize content for organic search during this time, hopefully some of these tools and techniques can help you along the way.

Want to talk further analyzing search or social data? You can always reach me via my LinkedIn or Twitter!