Show HN: I built a dataset of 25k+ Gumroad products to spot market opportunities(gumtrends.com)
gumtrends.com
Show HN: I built a dataset of 25k+ Gumroad products to spot market opportunities
https://gumtrends.com
64 comments
Heh, I remember doing this for WordPress plugins back in 2010ish. Basically look for plugins with high install rate but not been updated in a long time. Fork it, add on paid services, then market the hell out of it on Reddit/etc. I was making ~4-5k a month from that before I had to give it up due to a COI policy at work.
That's amazing! I thought about whether to keep the data to myself and build something or share it and sell access to it. I hope I made the right decision! (Also I just learned what a COI policy is)
I'll probably purchase the data :soon:... honestly. I no longer work for that company (after spending well over 5 years there) and I've been thinking about doing something like this again.
Whoa, was your work related to wordpress plugins?
You mean the COI? You could say that. Basically I wasn’t allowed to take payment for any WordPress related activities after accepting employment.
And it was worth it to give up the $4-5k / month instead of going to an employer without a COI policy and keeping the dual income streams?
Yeah. Some very pretty golden handcuffs.
Aside: This is a great title.
It shows what you made (a dataset of Gumroad products), why/to whom it matters (to spot market opportunities), and it implies how (dataset/scraping). Nice!
It shows what you made (a dataset of Gumroad products), why/to whom it matters (to spot market opportunities), and it implies how (dataset/scraping). Nice!
Thank you! I looked at the previous Show HN submissions and tried to get a pulse on what worked better.
How very appropriate for a post about market research!
I don't sell on Gumroad, but I do sell on Etsy, and use a somewhat similar software (albeit much more robust) called eRank to help spot trends and get ahead of the curve. You should look into amending this product with some more robust features:
* Utilize the API to let a Gumroad seller import their actual Sales numbers, to help further refine your reviews to sales numbers.
* Connect with google to spot search trends that lead to Gumroad products
* Run some sort of AI analysis to help the seller understand their copy is failing them for SEO, etc.
Check out eRank, and look at their offerings, and you can get a lot of ideas on how to grow this product from $50 once to $10+/mo.
https://help.erank.com/features/
* Utilize the API to let a Gumroad seller import their actual Sales numbers, to help further refine your reviews to sales numbers.
* Connect with google to spot search trends that lead to Gumroad products
* Run some sort of AI analysis to help the seller understand their copy is failing them for SEO, etc.
Check out eRank, and look at their offerings, and you can get a lot of ideas on how to grow this product from $50 once to $10+/mo.
https://help.erank.com/features/
Thank you so much this is really helpful! I was a bit worried about spending weeks on something that nobody wanted, so I launched a MVP to begin with. If there’s enough interest I’ll work on more features!
I would still keep a one off or license based pricing though, I feel a collective subscription fatigue
There are so many options on that link? Which is the best one to find product ideas?
It is a beast of a tool. I still only use a tiny fraction of it's potential. I use it to spot Trends, which are the right hand side of the dashboard (its the current most searched for keywords on etsy - 2 weeks ago it was "trump mugshot"). Then in the main portion of the dashboard, it summarizes my shop (inventory, missing/unoptimized tags, missing images, sales to date, global and national ranking, a few other less useful stats).
When you sign up, you connect your shop to your profile. After that, you can look for competitors and add them for comparisons.
When you sign up, you connect your shop to your profile. After that, you can look for competitors and add them for comparisons.
I have been looking through IndieHacker reported revenues for inspiration like this. What I found is a surprising number of ideas work and make $5k+ MRR and so not to get too hung up on the idea as long as it is feasible that people will pay for it.
For example someone is charging $30/m to convert PDF bank statements to CSV, and someone a similar price to effectively static site generate off a Google Sheet. It is all in how you sell it I guess! These are not trivial products but they are not things you need a team to build.
My thought is to build quick, get in front of people and then iterate on that feedback. Instead of building stealth then finding no one wants it.
Well done on this launch!
I was tempted to build something like this off IH but felt that I am not adding much value since IH already spill the beans. Although maybe running an LLM over each idea to spit out a summary might be cool (maybe I will do that).
I have some other ideas along the idea of build X but sell the side effects, the problems solved along the way.
For example someone is charging $30/m to convert PDF bank statements to CSV, and someone a similar price to effectively static site generate off a Google Sheet. It is all in how you sell it I guess! These are not trivial products but they are not things you need a team to build.
My thought is to build quick, get in front of people and then iterate on that feedback. Instead of building stealth then finding no one wants it.
Well done on this launch!
I was tempted to build something like this off IH but felt that I am not adding much value since IH already spill the beans. Although maybe running an LLM over each idea to spit out a summary might be cool (maybe I will do that).
I have some other ideas along the idea of build X but sell the side effects, the problems solved along the way.
Thank you! I think you're so right and I used that spirit to launch Gumtrends. I was worried about the reaction of people (about bugs, the scraping, missing features) but I'm so happy I launched anyway.
What’s the legality of reselling a dataset like this?
I would assume it really comes down to Gumroad's ToS and what they "allow" on their site, but it is a bit different if he was buying data and reselling it. He's not, he's taking publicly available data and adding value on top of it.
If I were gumroad, I'd love these tools to be out there, it drives more people to their platform through more content. I doubt they'll get upset about this.
If I were gumroad, I'd love these tools to be out there, it drives more people to their platform through more content. I doubt they'll get upset about this.
Is this a clone of trendsvc, as advertised on the home page?
"Originally, I took pre-orders for my Trend Reports on Gumroad. But I received... exactly $0. So I changed tactics: I made half of my report free, and the other half paid. Today, 99% of Trends.VC revenue is recurring in the form of annual and quarterly subscriptions.” Dru Riley sells business insights and expertise"
"Originally, I took pre-orders for my Trend Reports on Gumroad. But I received... exactly $0. So I changed tactics: I made half of my report free, and the other half paid. Today, 99% of Trends.VC revenue is recurring in the form of annual and quarterly subscriptions.” Dru Riley sells business insights and expertise"
Honestly I hadn't heard of trends.vc until now (and I never scrolled that far into the Gumroad homepage)
I have another website that is focused on trends (social media related) and I know it's a big market so that's why I built Gumtrends
How did you collect this data? Did you just use webscraping? What tools did you use for this?
Yes just webscraping. Just used Laravel, cause that's what I'm most familiar with. The site is Laravel + AlpineJS (with Vite)
I tried to scrape them and found they had some pretty robust anti-scraping implemented. Did you have to do anything to overcome it?
Exactly, did you build your own solution for this, maybe based on open source projects, or do you rely on external service for scraping?
I'm curious too since I'm developing a web scraping API product (https://scrapingfish.com).
What measures?
Can you also see the actual reviews of the products, or just the total number per star? If the reviews are also available there's potential room for sentiment analysis to automatically determine the common complaints about those products.
Unfortunately Gumroad does not display any text reviews as far as I'm aware
Why invest this effort? Couldn't gumroad just shut it down or outperform you tomorrow?
I have discovered other ideas like this but the lack of leverage makes it really hard to justify the squeeze
I have discovered other ideas like this but the lack of leverage makes it really hard to justify the squeeze
They could, but you get in while the iron is hot, make your money, and get out.
I have the same question for Skio. I see it on HN Hiring sometimes. They build subscriptions for shopify.
Surely the intended end-game is to be bought by Shopify with companies like that (or the equivalent of Shopify in their case)?
I don't think this is quite the same personally, since it seems to me more like Gumroad is just what OP picked initially for MVP, and could 'easily' (and with probably exponential value-add?) support more platforms later.
I don't think this is quite the same personally, since it seems to me more like Gumroad is just what OP picked initially for MVP, and could 'easily' (and with probably exponential value-add?) support more platforms later.
That's a fair question, and that's why I didn't build out everything I had in mind (trends, AI based suggestions, etc), I just put together the best MVP I could in a week and put it out there.
How are you estimating sales? Or how accurate is it?
I've calculated the ratings-to-sales ratio on products that do have official sales numbers, and used that to estimate it for the ones that don't.
I used the median ratio as a guide.
It's hard to tell how accurate the estimates are until some Gumroad seller confirm their numbers to me, but I think with the data I have this is a good guideline.
One surprising thing I found is that the median ratio is the same across all categories, I thought it would vary much more between different niches.
It's hard to tell how accurate the estimates are until some Gumroad seller confirm their numbers to me, but I think with the data I have this is a good guideline.
One surprising thing I found is that the median ratio is the same across all categories, I thought it would vary much more between different niches.
That seems incredibly unreliable. Justin Bieber videos can have ratings in the dirt and billions of views. I can imagine that the are plenty of esoteric items on gumroad which solve something super specific really well for a small audience, or vice versa.
Sorry I was a bit unclear on my previous message. By ratings I mean number of reviews. So I'm not using the average star rating at all, just the pure reviews count.
Got it. That makes much more sense.
I have a Gumroad product and would be curious what you estimated for my sales vs the actual sales. Email in bio if you want to talk :-)
Sure! I’ll email you :)
Did you try to cross-validate your approach?
Not yet, if I see enough interest in my product I’ll work hard to improve the estimates.
Insanely cool, love these analytics types of tools. Have you done this for any other markets or had any in mind?
thank you! Not yet but I have some ideas for sure
How'd you build the website? Using AWS for anything? Where's the database live?
Haven't used AWS. Just a VPS, with Laravel and AlpineJS + TailwindCSS with Vite on the frontend. Database is MariaDB, the DB is just on the server.
I'm new to this and would like to build a service like this one day. What exactly did you use Laravel, alpinejs, tailwind, and vite for? Was it necessary to have 4 different resources?
Laravel is propbably for connecting the db with the frontend.
Tailwind is for styling.
Alpinejs is for sprinkling interactivity on the site.
Vite is probably the bundler for all the frontend stuff to minify, vendor etc
Tailwind is for styling.
Alpinejs is for sprinkling interactivity on the site.
Vite is probably the bundler for all the frontend stuff to minify, vendor etc
On point.
Awesome, site looks great. This is the type of stuff I love to see on Hacker News.
Thank you so much!
Do you have an affiliate program? I'd be interested:)
Just followed you on twitter, I'll DM you
Sorry for the cynicism, but part of me wonders that if the market opportunities the tool finds are so valuable, why isn’t the person behind this pursuing them instead of selling this data?
No problem about the cynicism, I appreciate any questions.
Honest answer: I know Laravel, I know how to use a template to make a nice looking landing page, and I know how to plug in Stripe to receive payments.
I don't really write books or sell courses, Notion templates or any of the things that are big on Gumroad, so for now it's better for me to provide this information and charge for it rather than trying to learn how to write an ebook.
There are many who make their fortune selling shovels rather than going digging for gold :)
I like that you show a 4-star review first :)
Haha honesty is the best policy!
is there some way i can reach out to you? i just bought and have a few questions
sure just email at [email protected]
"Texture Brushes by Gal Shir for Procreate and Photoshop"
That is quite some brush!
That is quite some brush!