What's the difference between normalized data and absolute data?
Google Trends uses a methodology called normalization to adjust the raw search data, turning it into an interest ranking from 0-100, with 100 representing a keyword's peak popularity during the selected time frame, 50 representing half of that peak, and 0, insufficient data.
Our Interest endpoint returns this normalized data, directly as Google Trends shows it in their web app. Our Interest_enriched endpoint returns the datapoints as absolute search volume numbers, in addition to other metrics like YoY growth and MoM growth.
If you are looking to do custom calculations on how interest has changed over time, the Interest endpoint is a cheaper and faster solution. If you need absolute volumes, or any of these other growth metrics, the Interest_enriched endpoint is the right choice.
Why am I seeing slight variations in historical data?
Google Trends uses a sample dataset (a subset of data representing the whole) rather than the entire dataset, a decision explained in its documentation. This leads to faster loading times and keeps costs down.
However, this also means that sometimes when you make the same request after some time has elapsed, the historical data may look ever so slightly different. This difference – and the variance in general – is more visible with lower volume search queries (see the Law of Large Numbers).
High-Volume Keyword
Medium-Volume Keyword
Low-Volume Keyword
Practically speaking though, there isn’t much cause for concern. The differences between the data are generally in-line with how statistically significant those signals would be.
For example, with the low volume keyword chart above, the data is coming from merely a few dozen searches. With such an exceedingly small sample size, even though the data is drastically different, it doesn't represent a meaningful difference that would even allow allow it to inform any decision-making.
Need Support?
If you encounter any issues or have questions, don't hesitate to reach out to us at [email protected]. Our support team is ready and eager to assist you.