A stock photo side hustle: ¥28,284 in half a year — the breakdown across 5 sites, and the structure of "income with no turning point"

A monthly average of ¥4,714 was disclosed

Stock photo revenue disclosures tend to circulate mostly among outlier success stories. Which is exactly why the first-half 2025 results published by Hakuba Photo (a sole proprietor who left salaried work at age 40 and moved to Hakuba Village in Nagano Prefecture) carry unusual reference value, precisely because the numbers are small. The January–June total was ¥28,284, an average of ¥4,714/month. That’s the return after registering with five different sales platforms.

The breakdown across 5 sites

Laying out the share breakdown makes the shape of this case clear. The three sites where registration is entirely completable in Japanese (PIXTA, Adobe Stock, and PhotoAC) together account for ¥23,218, or 82.1% of the total. Shutterstock, which requires English-language input, contributes 16.9%. Image Mart, meanwhile, moved just ¥290 across the entire half-year.

There is no dramatic turning point in this case

Worth stating upfront: nowhere in this six-month record is there a moment that could be called a breakthrough. No single month spiked to tens of thousands of yen. There’s no sign in the article of any one image driving the sales. What the writer points to is simply a seasonal wobble, sales concentrated in January and April, and somewhat sluggish in March and June. As for the cause, he attributes it to his own inventory and settings: “I think this is influenced by the availability of seasonal-event-related material and the precision of tag settings.”

So what can be drawn from this case isn’t “what caused the breakthrough.” The real lesson is how thin he made the work per image, in order to keep sustained sales flowing across five sites simultaneously without ever hitting a breakthrough.

What’s actually working: an operational design of “spread one photo across four sites”

The same writer discloses his posting workflow in a separate article, and it directly explains the structure behind his income.

He uses only Google Translate, Excel, and copy-paste, no dedicated software. With this workflow, he posts the same photo across four sites (Adobe Stock, PIXTA, Image Mart, and Shutterstock) uploading in batches of 10 on PIXTA and sometimes registering 40 at once.

This is the core of the revenue structure. Stock photo income is built on a low per-unit price, so it only turns into real money once you multiply the number of listings by the number of outlets. But the more images you add, the more time tagging takes, linearly. So he creates the most time-consuming deliverable, the tags and title, just once, and reuses it across four sites. Revenue being split across five sites results from multiplying the exits from a single unit of work, not from sales effort.

Easy to miss: the four-site simultaneous posting doesn’t include PhotoAC/IllustAC, which returned ¥5,557 (19.6% of the total) in the first half. That operates as a separate route. The four sites in the main workflow total ¥22,727, meaning roughly 80% of revenue comes out of one single flow. Conversely, the remaining 20% is revenue earned through separate, additional effort.

Keeping Image Mart running, despite it earning just ¥290 over the half-year, also makes sense from this design. Viewed alone, it would be a candidate for shutting down. But since it’s just feeding in already-created tags at essentially zero marginal cost, its expected value can’t go negative. Conversely, the fact that Shutterstock, which requires the extra step of translation, returned 16.9% can be read as a figure showing what that extra effort was actually worth.

Unseen costs, and the limits of this data

There’s a gap worth stating honestly: the article discloses neither the number of images posted nor the time spent shooting or editing. As a result, the ¥28,284 figure can’t be converted into an hourly rate, and there’s no external way to judge whether this side hustle is actually worth the effort. Average revenue per image can’t be calculated either.

The monthly average of ¥4,714 also needs careful handling. Given that he himself notes the concentration in January and April and the slowdown in March and June, this is a figure that smooths out the swings, not a promise that “¥4,714 comes in every month”. Stock-type income like this fluctuates month to month, and even the writer can’t say for certain whether the source of that fluctuation is demand-side (seasonal events) or supply-side (tag settings).

Going forward, his stated priorities are refining tags and SEO, getting ahead of high-demand themes, and building out non-AI, real-life landscape and everyday photos. None of the three is about “adding a new sales channel”, all three point toward “polishing the existing inventory and settings.”

What can and can’t be copied

What’s transferable is the workflow itself. Starting from a site with AI-assisted tag suggestions, generating tags there, saving them to Excel, and then porting them over to other sites. This design doesn’t depend on subject matter or camera equipment. His criterion of “keep a low-revenue site like Image Mart running as long as it costs nothing extra” is also directly reusable.

What’s hard to replicate is the shooting cost. The lifestyle choice of moving to Hakuba Village itself creates daily, built-in access to snow-mountain and outdoor subject matter. Trying to shoot the same subjects from an urban base would add transportation costs and travel time that this revenue scale simply can’t absorb. It’s fair to read the ¥28,284 half-year total as achievable precisely because shooting happens as an extension of daily life, without being booked separately as a cost.

There’s also a time-horizon limit. Whether his own outlook, that “continuing could double the results”, actually holds true can’t be verified from this half-year of data alone. The next half-year’s numbers will be the deciding evidence, revealing whether the accumulation is really compounding or just holding flat.

Related reading

• 24 Kindle books — how far does stock-type royalty income grow?
• ¥10M/month and “80% of my articles dropped in ranking” — the ceiling and fragility of stock-type income