Walmart sold out of its Mothman dog costume by the second week of October. The costume — a hooded plush hat with red glowing eyes — was the breakout product of the season, and the runner-up was a Wendigo plushie cape that sold through three production runs at Petco. Halloween 2025 is, by my count, the strangest pet-costume year in the seven that I've been tracking the data on. Mothman, Wendigo, Skinwalker, Derpy the Tiger from KPop Demon Hunters, an unexpected resurgence of the actual Babadook costume from the 2014 film. The pet-costume aisle is essentially a folk-horror gallery this year. And if the pattern I've been tracking holds, the pet-naming aisle is about to follow it.
The six-month aesthetic lag
I've been tracking the relationship between pet-costume aesthetics and pet-naming aesthetics for several years, and the lag is about six months. The aesthetic that defines Halloween costumes in October typically defines first-time pet names by April or May of the following year. The mechanism is, I think, that pet costumes function as a low-stakes aesthetic test — owners try out a particular vibe on their existing pet for a weekend, decide whether they like it, and then bring the residual aesthetic preference into their next pet-acquisition decision in the spring or summer. The next pet they get gets the name that goes with the costume they bought last fall.
The pattern was clearest in 2022 to 2023, when the cottagecore costume wave (mushroom hats, fairy wings, vintage florals) preceded a measurable bump in pastoral pet names — Daisy, Poppy, Clover, Hazel, Pip — in NYC and Seattle municipal pet-licensing data. It was visible again in 2023 to 2024, when the dark-academia costume wave (tweed coats, leather satchel attachments) preceded a small but real bump in literary-coded pet names like Bram, Dorian, Cordelia, and Atticus. The lag is not perfectly tight, and the magnitudes are small, but the directional correlation has held in three consecutive years of data I've personally cleaned.
What folk horror predicts
If folk horror is the dominant Halloween aesthetic of 2025, the spring 2026 pet-naming wave should look quite different from the cute-fluffy default that has dominated American pet naming for fifteen years. The names that fit the folk-horror register — short, slightly ominous, evocative of the woods or the unexplained — are starting to appear in pet-licensing data already, in low absolute numbers but with rising trajectories. Hex appeared in NYC dog licensing data forty-one times in the first three quarters of 2025, up from twelve in all of 2023. Crow is showing up. Gloom appeared eleven times. Wyrm, six times. Mothra (close but not identical to Mothman, which is unlicensable as a name without permission), nineteen times.
None of these are anywhere near Bella territory. Bella was the top dog name in NYC's 2024 data with thousands of registered instances. Hex with forty-one entries is not threatening Bella. But the registers are different from each other in a way that matters: Bella is a stable, multi-decade name that does not move much year-over-year. Hex is a name that moves rapidly in percentage terms because its base is so small. The percentage-rate growth in folk-horror names is, I'd argue, the leading-indicator data, and the absolute-rate growth in cute-fluffy names is the lagging-indicator data.
What this is not
It is not the case that Bella is in trouble. The cute-fluffy register has lasted for fifteen years and shows no signs of imminent collapse. What I'm describing is a small fraction of pet-naming decisions migrating to a different register, not a wholesale shift. Bella will still be the top dog name in 2027. The folk-horror register will probably remain a niche position in the pet-naming chart, occupied by the kind of pet owner who makes deliberate aesthetic choices and is happy to be slightly outside mainstream taste.
It is also not the case that pet costumes are a perfect predictor of pet names. The lag is real but loose, and the magnitude of the costume-to-name effect is much smaller than the costume-to-name effect on, say, the actual costume aisle in subsequent years. What costumes do well at predicting is the texture of the next aesthetic — they tell us whether parents are leaning cute or weird, soft or hard, traditional or experimental. They do not tell us specifically which name will move.
Why this year's costumes are unusual
The reason 2025 feels like an inflection point in the costume data, more than other years, is that the Halloween aesthetic has historically tracked summer cinema relatively closely. The big Halloween costumes of 2018 were dominated by superhero films. The big costumes of 2019 had Stranger Things echoes. 2022 had multiverses. 2023 had Barbie and Wednesday. 2024 had a similar mix of superhero and pop-cultural reference points. 2025 has none of that, or rather, the cinema-driven costumes are present but they are not dominant. The dominant register is uncategorized folk horror — Mothman is not from a 2025 movie, Wendigo is not from a 2025 series, Skinwalker is not a single piece of media. The aesthetic is operating without a primary driver, which suggests it has internal energy that doesn't depend on any single cultural product to sustain itself.
This is rare and, in my experience, predictive of more durable downstream effects. Aesthetics that survive without a driver tend to produce more sustained naming consequences than aesthetics that depend on a particular show or film. The cottagecore wave was driven by Pinterest and Instagram rather than any single cultural product, and the resulting pastoral-pet-naming bump has lasted for three years. The folk-horror wave is similarly unanchored, and I'd expect its pet-naming consequences to persist into 2026 and possibly 2027.
What I'd predict for 2026 pet-licensing data
Three things, in order of confidence. First, Hex, Crow, and Gloom should each cross 100 annual registrations in NYC's 2026 dog data, up from their current double-digit positions. Second, the broader register — short, hard-edged, slightly ominous one or two-syllable names — should accelerate as a category, with new entries like Ash, Wraith, Reaper, Coven, and Dread appearing in low numbers. Third, the food-name register that has been the dominant non-Bella aesthetic for three years (Mochi, Dumpling, Pickle, Olive) should continue but slow noticeably as a percentage of total new registrations, ceding some growth to the folk-horror register.
The honest qualifier, as always, is that pet-naming data is messy and the magnitudes I'm predicting are small enough to be lost in the noise. The structural argument — costume aesthetics predict naming aesthetics with about a six-month lag — is the more durable thing I'd defend. The specific names will be whatever they end up being, and reading the data carefully in 2027 will tell us which of my candidates moved and which did not.
Why I find this interesting beyond pet data
The reason the pet-naming pipeline is worth paying attention to, and the reason I keep coming back to it, is that it tends to be a leading indicator for shifts in human-naming aesthetic preferences as well. The cottagecore wave that hit pet names in 2022 to 2023 is now visible in human girls' names — Hazel and Poppy are climbing in SSA data. The folk-horror wave, if it produces sustained pet-naming consequences, may eventually echo into human-naming choices five to seven years downstream. We won't see Hex on the SSA top 1000 anytime soon. But we may see, in 2030 or 2031, an aesthetic shift in human girls' naming toward shorter, sharper, slightly darker registers that could be traced back, by careful reading, to the Mothman year.
That's a long way out. The first test is whether Hex really does cross 100 in NYC 2026. The longer test, the one I find genuinely interesting, is whether the cute-fluffy era of American naming — pets and humans both — is finally starting to crack from the edges inward. The pet data is the place to look first.
Data source: NYC Dog Licensing Dataset + Seattle Pet Licenses. Analysis by NamesPop.
Found this helpful?
Share it with someone who’s picking a name.