In our fifty-term keyword pull, email marketing and email marketing app both fell ten percent — tied for the steepest decline in the set, against a field averaging around minus five.
The obvious reading is that email is dying. The obvious reading is wrong, and the gap between what this number measures and what people will conclude from it is a useful lesson in reading keyword data at all.
What the number actually measures
Search volume for a term measures people looking for information about a thing. It does not measure use of the thing.
These two diverge predictably, and the direction of divergence is knowable. Search volume for a tool category is dominated by people who do not yet have one — evaluators, beginners, the curious. People who already run email do not search email marketing. They log into the tool they picked years ago.
So a mature category with high retention and low new-entrant rate produces exactly this signature: declining search, stable usage. The decline is a statement about the inflow, not the installed base.
Compare against the shape of the rest of the dataset. Terms with heavy beginner intent are falling hardest: what is marketing down six, digital marketing app down ten, marketing agency down ten. Terms attached to live practitioner surfaces are holding: app store marketing flat, linkedin down one, content marketing down two. That is a pattern about who is searching, not about what works.
The second factor
There is a compounding reason email search specifically fell harder than most: it is the most thoroughly explained topic in marketing.
"What is email marketing," "best time to send," "how to improve open rates" — these have been answered ten thousand times in near-identical articles. That corpus is the ideal input for a generated answer: high consensus, low ambiguity, no need for a live account or original data. The question still gets asked. It stops producing a click.
Categories get hit by this in proportion to how commoditized their explainer content is. Email is at the extreme end. The ten percent is a measure of how solved the topic is, not how dead the channel is.
Why the channel is arguably appreciating
Set the search data aside. The structural case for owned audience is stronger now than it was, for reasons visible in this same dataset.
Discovery is compressing. If the informational tier of search is being absorbed, the reliability of "publish content, get found" declines. Channels that do not route through an intermediary become more valuable in relative terms — not because they improved, but because the alternative got less predictable.
Platform distribution is not yours. Every platform in this dataset can change your reach overnight, and periodically does. An email list is the only asset on the list where reach is a function of something you control.
Attribution is unusually clean. In a landscape where measurement is degrading nearly everywhere, email still gives you unambiguous delivery, open, and click data attached to a known individual. That is rare and getting rarer.
None of which means "start a newsletter" is good advice by default. Email works when you have something worth arriving in an inbox on a schedule. It fails when it is a checkbox — and there is a lot of checkbox email, which is a reasonable part of why the beginner search volume is thinning.
The general lesson
This term is a clean example of a mistake that is easy to make with keyword data and expensive to act on: treating search volume for a category as a proxy for the health of the category.
Three questions worth running before you draw a conclusion from a declining term:
Who searches this? If it is dominated by people who do not yet use the thing, the volume tracks acquisition, not usage. Mature categories decline in search while perfectly healthy.
Is the topic solved? Heavily-explained topics lose clicks to generated answers first, regardless of underlying demand. Steep decline on a commoditized explainer term says more about the corpus than the channel.
Is the whole field moving? In our set, forty-nine of fifty terms declined. Against that baseline, minus ten is roughly twice the field average — meaningful — but it is not the minus-ten-from-zero that it looks like in isolation. Always read a term against its field.
Run those three and most scary keyword numbers turn into something more specific and more useful than "the channel is dying."
What we would actually do with this
Not "abandon email." Not "email is back, everyone" either.
The defensible version: the falling search volume is a signal about how you get found, not about what to do once you have someone's attention. If discovery through search is thinning, the value of converting the traffic you do get into a channel you own goes up. That is an argument for email being downstream of content, which is where it works best anyway.
And the meta-lesson generalizes past this one term. In a dataset where nearly everything declines, the informative move is not reacting to any single number. It is reading the shape of the whole field — which is what we did, and which turned up a seed bias large enough to have hidden TikTok and Reddit from the entire export.
clownfish101 is a content ops loop: generate on tone, publish semi-automatically, recover your real results by screenshot, feed them back into the next batch. Measurement you own, on data that is actually yours. See the loop →