Four Years of Stock Photography, ¥185,000 Total. A Real Record of the "¥4,000 a Month" Median

Stock photography articles on side-hustle media are almost always drawn from the upside tail of success stories. But what you really need when deciding whether to enter is median data: “if an ordinary person keeps at it in an ordinary way, how many yen does it become?” The four-year sales record published by the operator of the travel blog “Solo Tabi Note” fills exactly that gap. Roughly ¥185,000 cumulative, about ¥4,000 a month on average, whether you read that number as “no dream here” or as “the floor for near-total neglect” completely changes how stock photography looks as a side business.

The published numbers

By the author’s own tally, the Adobe Stock portion is paid in dollars, so the weak yen provided a large tailwind, at ¥150 to the dollar it corresponds to over ¥120,000. So more than half of the roughly ¥185,000 cumulative total was earned by Adobe Stock alone, with the two domestic sites (Photo AC and PIXTA) splitting the remainder.

It started with “pandemic downtime”

The start was 2020. The author, who worked at a travel agency, found themselves commuting only about once a week during the COVID-19 pandemic and took up stock photography as a way to use the freed-up time. They shoot landscape photos exclusively. The style is continuous with running a travel blog: photos taken on trips get registered as stock assets as-is.

That means this record is not “four years of seriously chasing income” but “four years of continuously uploading a byproduct of daily life.” The author themselves looks back on it as “the epitome of neglect” and says they have “nothing but regrets”. The one bright spot they cite for 2024 is having added new photos from an overseas trip in autumn 2023. If you skip this premise and read the data as “only ¥185,000 in four years,” you misread what the data means.

What sells on each of the three sites

Line them up and the pattern is unmistakable. What sells are wide-format landmark landscapes (Umeda in Osaka, Himeji Castle, Angkor Wat) and generic backgrounds like sky and clouds. All of these are “imagery that designers and web managers use as backgrounds for documents and banners”, the market moves on an axis entirely separate from a photo’s merit as a work of art. And this tendency, the author says, has barely changed from year to year.

The observation that horizontal (landscape-orientation) photos sell better belongs to the same context. Website hero images and presentation slides are both wide-format. The demand side’s use cases dictate the format.

One more thing the table reveals: the top sellers do not match across the three sites. Umeda, Osaka ranks high on both Photo AC and Adobe Stock, while PIXTA’s top sellers have a different composition, snowscapes and overseas locations (Angkor Wat, Würzburg). Even with the same pool of assets, each site’s buyer demographics and search behavior differ, so what sells diverges. The fact that a product shot of “shoes” shows up among Photo AC’s top sellers is telling: you cannot fully predict in advance what will hit, which is precisely why selling on multiple sites, despite the hassle, is a rational form of diversification.

Reading behind the numbers

The variable that separates outcomes is not shooting skill but “volume of uploads × fit with demand.” That the best sellers didn’t change over four years means the demand pattern was observable from the start. Put the other way around: if you shoot for the demand side (what users need it for) from year one and pile up volume, this median can be beaten. The roughly 10x gap versus the case that built ¥500,000 a year with AI-generated assets can be explained not by talent but by differences in production cost structure and demand selection.

The numbers are also honest about the fact that years without volume don’t grow. New uploads in 2024 were 854 (less than half the 2,000-photo goal) and revenue stayed flat at around ¥4,000 a month. Stock-type businesses are almost entirely determined by “the amount of assets placed × demand per asset.” The author too names insufficient volume as the next challenge: “I need to get at least my main channels, Adobe Stock and PIXTA, up to around 2,500 registered photos.”

And exchange rates worked as a third variable. Because Adobe Stock pays in dollars, a weak-yen phase increases the yen-converted payout without any effort. The skew (over ¥120,000 of the ¥185,000 total coming from Adobe Stock) reflects the currency tailwind as much as asset quality. If the yen strengthens, the same logic cuts the other way. Looking at the exchange-rate effect from only one side is not accurate.

Where “¥4,000 a month on near-total neglect” fits

By the yardstick of evaluating side businesses at an hourly wage, ¥4,000 a month falls into the “not worth it” category. But the nature of this income differs fundamentally from flow-type side businesses in three respects: it keeps coming in with zero additional labor, assets once registered do not disappear, and the shooting can double as part of ordinary life such as travel. As the real data from month three of a resale (sedori) business shows, flow-type income drops to zero the moment you stop working. Stock photography’s ¥4,000 a month should be read as the floor value of “an asset that does not go fully to zero even when you stop the labor.”

Conditions for reproducing this — and the limits

Some of this record does carry over. Anyone who already has a habit of traveling or going out can reach the “median = a few thousand yen a month” at close to zero additional cost. The best sellers (wide-format landmarks plus generic backgrounds) are observable from public information, and steering toward demand requires no special equipment or technique. The low barrier to entry is itself a form of high reproducibility, “whoever does it won’t miss by much.”

The lines it does not cross are just as clear. The path from ¥4,000 a month to a full-time income does not lie on this trajectory. This very dataset shows that continuing the same approach lands you at “¥4,000 a month after four years”. Changing the order of magnitude requires either a shift in production cost structure like AI generation or a demand-first mass-production setup. The weak-yen tailwind, meanwhile, is an exogenous variable unrelated to the author’s effort, and it promises no reproduction. And a single example specializing in landscape photography cannot be extrapolated to genres with different pricing and demand structures, such as people or business scenes.

The greatest value of this case lies in the very act of publishing four years of an “ordinary record”, down to the breakdown of best sellers, rather than a success story. What you should look at before entering is not the dream of the upside cases but this median.

Related reading

• ¥500,000 a year with AI generation × Adobe Stock — the upside tail of the same market
• Real data from month three of a resale business — a contrast with flow-type side income