Etsy Product Engine
Three Jobs A Digital-Product Seller Does By Hand, Deleted
Research, listing mockups and customer follow-up are the three tasks that cap how many products one seller can run. This engine does all three, from finding the product to asking for the review.
- Industry
- Retail + ecommerce
The challenge
Somebody selling digital products runs three businesses at once, and only one of them is the product.
The first job is research: working out which digital product is worth making before spending a week making it. Guess wrong and the cost is not the file, it is the week, and the seller usually finds out six weeks later when the listing has forty views.
The second is mockups. A digital product has nothing to photograph, so the listing image is the product as far as a browsing buyer is concerned. Every item needs its own, and none of them can be skipped, which makes image production a bottleneck rather than a finishing step.
The third is follow-up. Past customers are the cheapest buyers a shop will ever have, and asking them for a review or offering them something related is the ordinary work that almost nobody gets around to, because it happens after the sale when attention has already moved to the next listing.
None of the three is difficult. All three are repetitive, and togetherthey set the ceiling on how many products one person can run. A seller does not stall because the work got hard. They stall because the same three chores now take the whole week.
What we built
The research. The engine finds digital products worth selling, so the choice of what to make next arrives as an output rather than as an afternoon of browsing.
The mockups. Listing images are generated rather than assembled by hand. That removes the step between deciding on a product and being able to publish it at all, which is the step where most half-finished ideas die.
The follow-up. Past customers are emailed for a review, and then offered something else. Both run against people who have already bought, which is the difference between follow-up and cold email.
The order of those two is deliberate and it is the part most sellers getbackwards. The review request goes first, while the buyer still has the file open and an opinion about it, and the offer follows. Asking for money before asking for an opinion turns the whole message into a sales email, and the review, which is the asset that keeps working on every future buyer long after the upsell is forgotten, quietly never gets written.
One path, not three tools. Research to listed to reviewed runs as a single flow with nobody operating it in the middle. That is the part that matters more than any individual step, because a chain of three automations with a human carrying output between them is still a job, just a more annoying one.
Own tooling, said plainly. The Etsy Product Engine was built for our own business. It is not a client engagement, it is not a product with users, and there are no seller results, revenue figures or listing counts attached to it on this page.
The deliverables
- Trend and product discovery that surfaces digital products worth selling
- Listing mockups generated as images rather than photographed or built by hand
- Review requests emailed to past customers automatically
- An upsell that follows the review request to the same customer list
- One path from research to listed to reviewed, with nobody operating it in the middle
- Assembled on eRank, Apify and n8n rather than written from scratch
The results
- 3
- Manual jobs the engine removes: research, mockups, customer follow-upMeasured 2026
- 0
- People in the loop between finding a product and asking for the reviewMeasured 2026
- 3
- Systems the engine is assembled from, rather than written from scratchMeasured 15 August 2026
What the numbers say
The Etsy Product Engine removes three manual jobs from a digital-product business, and it runs the path from finding a product to asking for the review with nobody operating it in the middle. It is assembled from three existing systems rather than written from scratch.
What is deliberately not here: revenue, listings published, reviews collected, conversion on the upsell. None of those are reported because this is own tooling rather than a client engagement, and no measured seller results have been recorded for it. Counting what a tool does is not the same as measuring what it earned, and only one of those two things belongs on a page like this.
Why it worked
Because it automated the parts with no judgement in them and left the parts that have some.
Deciding what a brand should be, what a product should feel like or which category is worth entering is a decision. Producing the fourteenth mockup this week is not, and neither is emailing a buyer from March to ask how the file worked out. Those two are the tasks that quietly set the ceiling on a one-person shop, because they scale linearly with the catalogue and they never get faster with practice.
The second reason is that the three jobs were connected rather than automated separately. Most sellers who automate do it in pieces: a research spreadsheet here, a mockup template there, an email tool nobody has opened since setup. The gaps between the pieces are where the work reappears, because a human still has to carry the output of one into the input of the next. Research to listed to reviewed is one path here, which is why it runs without anyone tending it.
The third is that almost none of it was built from scratch. It is assembled from tools that already existed, which is why it exists at all rather than sitting half-finished behind a longer build.
The takeaway
If you sell digital products, spend one week writing down every task you repeat and mark each one as either a decision or a chore. The chores are usually more of the week than anyone expects, and they are the whole reason a catalogue stops growing at whatever number one person can personally maintain.
Then automate the chores end to end rather than one at a time. A research tool that hands you a spreadsheet you still have to act on has moved the work rather than removed it, and the seam between two half-automated steps is where a system quietly stops being used.
Questions people actually ask
What are the three jobs a digital-product seller actually loses time to?
Research, mockups and follow-up. Research is working out which digital product is worth making at all. Mockups are the listing images, which have to exist before anything can be published. Follow-up is emailing past buyers for a review and then offering them something else. Each one is unskilled, repetitive and unavoidable, and together they cap how many products one person can run.
Why do listing mockups matter more for a digital product than a physical one?
Because there is no object to photograph. A downloadable file has no shelf presence, so the listing image is the entire product as far as a browsing buyer is concerned. That makes the mockup a bottleneck rather than a finishing touch, which is why the Etsy Product Engine generates them instead of leaving them to a human with a template.
Is this a product I can buy?
No. The Etsy Product Engine is tooling built for our own business, not a client engagement and not a shipped product with users. It sits in this portfolio because it is real and it runs, not because it is for sale. There are no revenue figures, listing counts or seller results on this page for the same reason.
Does automating review requests annoy customers?
It depends entirely on timing and on who is asked. The Etsy Product Engine emails past customers, meaning people who already bought and already have the file, and it pairs the review request with an offer for something related. A request sent to a stranger is spam. A request sent to somebody holding a thing they chose to buy is the ordinary follow-up most sellers simply never get around to sending.
