Happy Easter everyone^{1}! To celebrate, I’m doing two posts this weekend. This post^{2} is about the trends in search volume data relating to Google searches for “Easter”^{3}.

I decided it might be fun to look at some of the search data for searches about “Easter,” and write a post about it for the blog. So that’s what I did. To do this project, I used the freely available software Google Trends. The website allows you to enter a search term and see when it is searched for the most. A search for “Easter” looks like this:

Data source: Google Trends (www.google.com/trends).

This graph shows that most Google searches for “Easter” are done around Easter of each year, which is where you see the spikes on the graph. I decided to run a search for Easter of each year and see when each year was searched for the most. To manipulate the data more easily, I downloaded it from Google Trends, then imported it into Microsoft Excel.

Next I replaced all the “<1” markers that Google Trends inserted into my data with 0^{4}, so that I could actually graph it.

Then, I removed all the month labels^{5} (except for the first column) from the spreadsheet, leaving just the data.

Let’s add some conditional formatting to help better visualize the data. We’ll do it in blue, because blue is a nice color (and green for the year numbers on the top).

Here’s all of the data in one image^{6}. We’ll get into it in more detail later^{7}.

The green numbers on top represent the year that I searched for on Google Trends (for example, “2007” became “Easter 2007”). Each column is one year. Each row represents one month of searches, starting with January 2004, and ending with March 2018. The numbers in each cell represent the search volume for that year’s search (example: “Easter 2008” – ninth column) and that month of searches (example: January 2004 – first row).

Now, get ready for a *lot* of graphs.

This graph shows the number of search results for “Easter 2000” each month since January 2004; we’re starting with 2000 even though the search data only started being recorded in 2004. The y-axis displays the search volume of this term, and the x-axis represents the number of the month. The most searches are done in the first couple years after 2004, since these are the closest years to 2000. The number of searches decreases over the next few years, and after around 5 years, the graph levels off. Note that each year, around Easter, there is another spike in searches, even though the search is for “Easter 2000” and not for that year.

The search results for “Easter 2001” are *much* more chaotic. This could possibly be attributed to the fact that “2000” is a much neater, more commonly chosen number than “2001.” There is still a spike^{8} every year around Easter, but the graph is much more random and the highest point is actually in around 2011, instead of at the beginning.

“Easter 2002” returns to the trend of “Easter 2000,” but with a lower search volume after the first few years.

“Easter 2003” took much more time to level off, but the search volume fell even lower after 2012 or so than the previous 3 graphs did.

Here’s where it gets interesting. Since our data starts with 2004, the first year on the graph this is the first year where our search actually aligns with the graph. There is a spike around (don’t overthink it) Easter that year, and a much, *much* smaller spike around a year later (presumably around Easter 2005).

Now we’ll jump ahead a few years to “Easter 2010” which peaks in 2010. We can see the entire spike here because 2010 is about halfway through our data, as opposed to 2004, which was right at the beginning. There is a steady increase in the number of searches for “Easter 2010” leading up to the actual holiday over a *long* period of time – up to around 20 months. Then, there is an extremely sharp falloff, in which the search volume decreases around *95%* in just one or two months, leveling off for the next 9 months or so.

Predictably, the same thing happens when we increase the year to 2018: the spike moves to 2018. As the number of the year increases, the spike moves across our graph, with the x value of its peak corresponding to Easter of the year number that we input^{9}.

Here, I’ve overlaid (overlain?) the graphs of “Easter 2010” (blue) and “Easter 2017” (red). You can see that they follow a very similar pattern, with an initial spike on Easter the year before^{10}, then a drop, and then a gradual increase until Easter of that year.

This graph contains the search volume for every year from “Easter 2004” to “Easter 2018” (lighter is an earlier year, darker is a later year). It’s a good way to visualize how similar the search trend is each year. On the spreadsheet, you can also see the search volume peaking later and later as the year increases (the darker blue regions are times when “Easter [that year]” is searched for more frequently).

This is the search volume for 2019 – very similar to 2018, but with more visible spikes preceding the peak search volume. Of course, the data only extends until March 2018^{11}, so we never see the search term and the year overlap. This is most likely due, again, to people searching for the Easter of the next year ahead of time.

Amazingly, “Easter 2020” has had a measurable search volume even since 2006, with large spikes beginning in 2013 and increasing until this year. Again, this probably has something to do with 2020 being a more commonly searched-for year (and a more even number).

“Easter 2021” displays a similar trend.

This pattern begins to deteriorate with “Easter 2022.”

The graph grows even more chaotic at “Easter 2023,” but still has the highest search volume in 2018.

Surprisingly, the most searches (by far) for “Easter 2024” were done in late 2017 and early 2018.

Interestingly, after “Easter 2025,” the peak search volume is no longer in March 2018, but instead around 2014. I’m guessing this has something to do with people misspelling “2015” to “2025” or “2014” to “2024.” After 2014, the search volume fluctuates, reaching another peak around 2016.

Here’s the last 5 graphs together.

This graph (and the next one) make it seem more likely that the search volume increases for a previous year because of misspelling while searching, but this is just a guess. The peak for “Easter 2026,” however *is* in 2016 . . .

. . . and the peak for “Easter 2027” is in 2017 . . .

. . . and the peak for “Easter 2028” is in 2018.

The search volume for “Easter 2029” are much more spread out, and it’s hard to imagine why people would search for “Easter 2019” over 10 years before that date. It is worth noting that, though the spikes are around 2005, most of the search volume is concentrated after around 2014.

This is the search volume graph for “Easter 2030” over the last 14 years^{12}. They are almost as random as those for “Easter 2030,” with a higher concentration closer to 2018.

And, just for fun, here’s all of the graphs together. It’s pretty easy to see that the lighter graphs (earlier years) are farther to the left, and the darker graphs (later years) are more prominent on the right. The later graphs are also somewhat more random and chaotic than the lighter ones.

I hope you enjoyed this post. I may do a follow-up post next week including some of the things that I wanted to include in this post, but didn’t have time to. See you on Tuesday with another post. Happy Easter^{13}!

- And April Fool’s Day.
- Good luck loading this post – there are 31 individual images (although the total file size is pretty low, about 1.35 MB).
- You can do this for any holiday or event, not just Easter.
- 1,767 cells were replaced, if you’re wondering.
- Maybe I should have learned to use a macro before starting this project.
- 4 individual screenshots stitched together.
- The data for March may be
*almost*incomplete, since this data was downloaded on March 30, just before the month ended, but it’s close enough. - I’m getting a
*little*tired of calling this a “spike,” but I can’t think of anything better. - It was
*so much fun*when I realized that half of my data was shifted over by one column, and that two-thirds of my graphs were also broken as a result. The spreadsheets in the beginning of the post still haven’t been updated, but let’s leave that as a little secret between you and me. - People searching for “Easter 2017” on Easter 2016 for whatever reason.
- Until Google invents a future-telling AI, at least.
- “Easter 2049” sounds like a good name for a sci-fi movie.
- This post said “Easter” 44 times, if you’re wondering (make that 45).

*That’s a new record for how many footnotes I have in one post. Hooray.*