October 11, 2019

400 MILLION DOWNLOADS

Ryan Badger @ryanbadger

On August 25th 2019 we surpassed 400,000,000 file downloads.

Monthly page views are averaging 30 million.

22 million unique users per month are using the site.

Google Adsense revenue is $40,000/month.

Subscription revenue is $10,000/month.

Running costs are around $4,000/month.

Team size: 1.

Monthly profit: $46,000/month.

  1. 2

    How are you achieving a running cost of only $6k per month? Are you using S3 or something else entirely?

    I'm especially curious as my own service stores videos and my backend (s3) gets really expensive.

    1. 2

      Yes I had to learn that lesson fast, s3 is GREAT for cold storage (I backup every file there and store it forever for $1/TB) but the bandwidth fees are unmanageable, so I manage my own bare metal servers with automated load balancers and fail overs. I have boxes in ever major region so when a file is uploaded it's routed to the closest server for speed.

      1. 1

        An impressive result!
        Could you please tell how many files you store and how big they are?

        I'm trying to understand how you organized it all (from a technical and software point of view).
        If I understand correctly:

        • you are using S3 cold storage for backup
        • you transfer almost 1 PB of data per month
        • you have your own bare metal servers

        But S3 cold storage has long access time (and big bandwidth fees), so I think
        it is only used in case of problems with your servers.
        In that case, where do you store the files? On your servers ?
        But it probably would require hundreds of TB of disk space ....
        How do you do it at such a low cost?

        1. 1

          At the moment, there are around 5.5M files hosted. Sizes vary massively from a few bits to 100GB. The average across every file is around 1MB, but we currently host a few hundred TB of files.

          You're correct, we use AWS s3 deep archive glacier for backups, but the files are served from our own hardware, which means we have 0 data transfer costs.

          We have life cycle rules setup, which means once a file falls out of activity (no downloads for x amount of days) we migrate it to the deep archive storage, when a new download request is received, we move it back to our own servers and serve it from them, until it is archived again, and so on. (this doesn't apply to paid users' files, they are always kept available on our hardware so that they can be accessed at any time without retrieval delays.)

          I won't give away where we got our servers from, but it was the only supplier I could find that didn't impose bandwidth charges! Without them our product could never be possible.

          1. 1

            I understand. How do you deal with failures? Are the files on your servers saved in one or many copies? What if the server disks fail? Do you restore files from S3 storage?

            1. 1

              Fortunately, in 3 years we've never had a disk failure or data loss. That said, in the event of it occurring, we have backup procedures to guarantee data recovery.

              1. 1

                Can you write something more how you solved it? Of course, disk failures are rare. But reports show that you can expect about 1.5% failure. So with the scale, it's more and more likely. I checked the s3 transfer costs and they are HUGE. Supposing that a 10 TB disk has failed, the cost of transferring 10 TB from s3 storage would be 90K USD ... So if you can say something more how you prepared for such a situation. I ask because I recently had a problem with the backup of my server and it was a big failure :)

                1. 1

                  Of course I was wrong. transfer costs for 10 TB is 900 USD so there is no big problem ... But I'm still curious about your solution. Eg what file system do you use on your servers ...

  2. 2

    Wow, that's impressive. Is your story on IndieHackers yet or what??

    1. 3

      seriously.. where's the interview?? I want to listen to this one.

      1. 2

        I kind of hate talking about myself. Also I am a terrible role model. Everybody I know works harder than me, I think I just got lucky.

  3. 2

    Great! Congratulation! "Running costs $6000". What that covers? And what is your backend stuff?

    1. 1

      The costs are made up of:
      My own bare metal servers
      AWS (RDS, Elastic Beanstalk, S3 for Glacier storage (every file is backed up))
      Memcachier
      Sendgrid
      Xero
      Cloudflare
      Intercom
      Gmail

      Actually the costs are more like $2,000/month, I was paying a revenue share to a large partner, but that has come to an end now (which means revenue will fall too from October)

      1. 1

        hi Ryan, I also build my app behind the AWS service. We use Cloudfront as our CDN solution, but you did it by Cloudflare. Do you mind sharing your reason why you choose Cloudflare instead of Cloudfront?

        1. 1

          Honestly Cloudflare is amazing. I could/should write an entire article on why it's so great, but there's already plenty out there. The main reason is Cloudflare is mostly FREE. Whereas Cloudfront charge per request and bandwidth, Uploadfiles gets over 1 billion monthly requests, and almost 1PB of data transfer a month. With Cloudfront this would cost something like $50k/month. (according to their online cost calculator)

  4. 2

    AMAZING!!!

    What is the main thing people use Uploadfiles for?

    I see you are promoting "Sell your digital products with Uploadfiles" more than any other features. Is this where most of your MRR come from? I mean do premium users use this feature more than any other features?

    1. 1

      Selling digital files is just the most recent major release, it really doesn't have many users, it was more of a test whilst I rebuild my other product, new.shoprocket.co

      70% of the revenue is from Google Adsense, the rest is from paid subscriptions. The majority of users are on the Pro plan for $5/month and it's used mostly for faster download speeds, accessing old/expired files and more storage space.

      To be fair though, I do absolutely zero marketing, I'm sure more people would use the more advanced tools if I actually told people about them!

      1. 2

        Do you advertise the product when people share files? Just curious how people hear about the service if you're doing zero marketing.

        1. 1

          When somebody shares a file, this is what others see: https://ufile.io/o877751q
          We have some call to actions on the page to encourage signups, but our conversion rate for creating accounts is 0.17% and conversion rates to paid accounts is <0.01%

          1. 1

            what is your backup plan if Google Adsense decide to ban you ?

            1. 1

              I get offers every day from other ad networks so it would be easy to switch, I just stick with Adsense because they actually give me a great CPM and payout every month

  5. 1

    That's amazing 🚀

  6. 1

    How do you deal with users who upload ileagal stuff. I had a simmilar site and one of the biggest problems was I was getting takedown requests and Google chrome banned the site a few times because of malware and ISIS stuff going on there.

    1. 1

      lots of automated scripts :)

  7. 1

    Congrats ! inspiring !

  8. 1

    Wow, WILDLY profitable! VERY well done @ryanbadger

    1. 1

      Signed up and uploading some files. I don't see any advertising. Am I missing something? I thought you were earning from adsense?

      1. 1

        Do you have adblocker? :D

  9. 1

    Congratulations!

    I was wondering what the general view is on using Adsense as a source of revenue?

    Uploadfiles.io is making a lot of revenue from it, but I do not see many other startups utilising Adsense? Why?

    1. 2

      Honestly at the peak of our traffic, the CPM was pretty low, I'm always being told that Adsense is "bottom of the funnel stuff" and I should be using larger ad networks, but I've yet to find one that could out perform Google (especially the payment terms).

      Our traffic is much lower this month as we lost a large partner (they closed down their app) and now our CPM is much higher, $3.26 per 1,000 ad impressions. At the peak of our traffic, that was more like $1.

      Also the ad coverage is only 51.66%, meaning only half our ad requests are being filled. (meaning we're missing out on 50% more revenue)
      I'm actually about to try a new ad agency... I will update here if it out performs Adsense.

      1. 1

        Could you please tell us more about partners? Who they are?

        1. 2

          Our largest partner was a small team who hosted their APK with us, it had millions of users and they released monthly updates. We gave them 15% of the ad revenue generated every month, which meant not only could they host their file for free, but they also earned $5,000+ a month from us. Unfortunately they recently closed down their business which means a lot less traffic for Uploadfiles. I need to find more partners like that but I hate sales & marketing...

          1. 1

            Let's talk about sales & marketing maybe? Do you need help with it? Maybe advice? Can do it for u. Just hit me up)

            1. 1

              I wouldn't even know where to begin, it's so difficult to think how to market something that has so much organic traffic, anything I tried in the past had an inconsequential impact.

      2. 1

        I have also heard the "bottom of the funnel" remark, but i think combining it with other revenue models like you do is the way to go.

        Ad coverage percentages is a another reason why having huge a amount of page views are necessary to make it. <100% also means you are not hammering users with ads all the time which i believe is good for UX.

        Could you elaborate on why Adsense has superior "payment terms"?

        1. 1

          Adsense pay you reliably within 21 days, all the other ad agencies I found were minimum 30 days, some 60 days, and a few even longer...

  10. 1

    Man that is awesome. Kudos to you. Inspirational.

  11. 1

    Your result is very, very impressive. I wonder, though, how you succeeded in competing with Dropbox, Google docs and other file hosting providers? Do you have something specific they don't have?
    And additionally: how did you dare to start such a project knowing there were mighty players like Google and Dropbox?
    Thanks!

    1. 1

      Honestly I don't try to compete with them, I just wanted to make something simple and do it well. With Google & Dropbox you have to create an account to upload and share a file, that was my main selling point, that you could just drag & drop and get a short file URL immediately. Once it had traction, I started adding other features to add extra functionality (like the desktop app) which does compete somewhat with Dropbox, but it's not even close in terms of which is better, Dropbox has over 2,000 engineers and a P2P network of file markers and encrypted chunks that give the appearance of impossible speeds. My app is built in Electron and basically does post calls in node :D

      But ultimately in my experience... there are always enough people out there who prefer to try something else. Google & Dropbox completely dominate the space, but realistically an alternative app only needs 10,000 users paying $5/month to achieve the revenue I am generating. It's a drop in the ocean compared to the giants, but it's more than enough for me!

      1. 1

        Thanks for your reply! You just confirmed my thoughts :)) Good luck!

  12. 1

    My mind is absolutely blown. Congrats!

  13. 1

    Congrats @ryanbadger! So what next for Uploadfiles.io? Do you plan to remain a one-person outfit?

    1. 1

      Yes. I've had some interest in the past from potential acquirers, and also VCs who have wanted to invest and help make Uploadfiles "the next Dropbox", but honestly it doesn't interest me.

      I went through Google Seedcamp accelerator in 2014 with my other product (Shoprocket) and we chased that dream for 3 years of being the next big thing, but it's not for me. With all the managing large teams, red tape, constant pitching for more investment, hiring/firing, office space, founder arguments and accounting, we didn't even manage to finish building the core product . I built Uploadfiles on my own in 2 days!
      Yes it was garbage and I stole most of the original website theme, but it worked. I love working alone, making mistakes & learning fast.

      1. 1

        Interesting, thanks for the reply :) And given that it's seemingly the same as wetransfer (or a number of other services), how did you find your first customers? How competitive is the advertising space etc

        1. 1

          At the start I just posted it on Reddit and Hacker News, I tried a few banners (it didn't seem to do much) and tried promoting it here and there, but really I think the growth was mostly just organic...

  14. 1

    Very nice. How do you deal with copyright infringement?

    1. 1

      It's all automated, when a DMCA complaint is received the file in question is immediately removed and the uploader alerted. Repeat offenders are blocked. Saying that, I've been hit hard by a Google block in the last few months, I used to rank number 1 for some great keywords like "upload" and "upload free" etc but now I don't seem to rank for anything. No messages or warnings in webmaster tools though. Typical Google.

  15. 1

    Very impressive! Congrats :)

  16. 1

    Damn, you are rolling in it. Impressive.

    What is your main acquisition medium? Search, maybe? If so what is you rankings on relevant words?

    1. 1

      The vast majority is direct, then referrals. Search & social only make up 8% of traffic. I used to rank really, really well for some great keywords, but was hit hard by a recent Google update (or maybe a ban?) and now I hardly rank at all. I'm sure an expert could figure out what happened and maybe fix it, but I've not idea.