Hacker News new | comments | show | ask | jobs | submit login
Ask HN / Review my startup: Transloadit
25 points by felixge on July 20, 2010 | hide | past | web | favorite | 35 comments
Transloadit is offering file uploading and processing as a service.

http://transloadit.com/

At this point we handle image resizing, video encoding and storage in your own Amazon S3 bucket.

We just launched 1 week ago and have started to sign up the first few customers.

If you are potentially interested in our services, please take our 8 question survey:

http://tinyurl.com/369un29

And of course, ask me anything!




Pro:

- The idea is sound. Value-added services like this are exploding all over the web

- The value-prop is clear

- The feature set is clearly explained

Con:

- As others have said, the pricing would work better as pennies per upload -- makes it cheap for rapid prototypers, whom are going to be a big part of this market

- Uploading via a third party raises privacy issues, even if you're not providing storage

- I suspect the market here is too small: it's the gap between people who know how to integrate your product and people who know enough to build it themselves. However, it's a big Internet and I've been wrong before about the sizes of markets (for instance, Tumblr!)

So go for it, and good luck :-)


Pricing: We are going to make things cheaper.

Privacy issues: It's basically not much worse / better than using co-located servers.

Market size: Since you mention Tumblr, they have actually switched to one of our competitors for video encoding recently. It's a really a very hairy problem, so any company that does not see it as a core-aspect of their business should be a potential customer.

Thanks for the feedback : )!


Very interesting!

My first impression of the homepage is that the graphics are a little all over the place (in terms of consistency, not layout). Your logo has a pretty slick look to it, while the large balsamiq image is rough and 2D, and the "Let us do the processing" images are somewhat rough and 3D.

The forgot password page layout looks off to me, the recover button is displaying to the upper right of the email label.

The signup link at the bottom of the login page directs to a different account creation form than the signup links on the plans/pricing page.

"Proceed to enter credit card" seems a bit rough or robotic (can't really think of a good word to describe it honestly), maybe it could be changed to something like "Proceed to finish signup".

Although I didn't look too deeply, the docs look great!

PS: I'm not a designer, I only pretend to be one on the internet, so take all of that design advice with a grain of salt.


Hey thanks for the suggestions, what browser are you using?

I agree that our visuals are not the best - hopefully we can add a real designer to our team at point.


I checked both Chrome beta and Firefox 3.6.3 and they're both showing the same thing.


Well done Felix!

1. Great intro and explanation of everything.

2. Remove the plans. Charge per transload.

3. Add a form to test-drive a transload.

4. Add a real-time stat to the front page: number of transloads since you started reading this page (or something to that effect).

5. I like the "Let us know what pricing you would like." idea.

6. Build your own feedback form. Just a simple textarea with a button. Don't make the user search through a forum to check for duplicate questions before giving their feedback.

7. Re: "We are also the first commercial software / infrastructure as a service product built on node.js." I know of a web-based app launched on Node in April but you've probably been going in beta longer.

8. Could you figure out a way to be able to just start using transloadit without signing up (with some kind of expiring mechanism for accounts which don't go through with a deferred registration)?


2. and 3.: We are working on that, right now

4. We've processed ~4.000 uploads so far / 60.000 internal jobs (each time a robot touches a file). Great idea, we'll add that.

7. We are the first one that charges money for the service. Let me know if I'm wrong and I'll change that : )

Using without an upload: Maybe, has a lot of potential abuse. We'll consider it so


Re: 7. Yes, the app (http://sexbyfood.com) also charges. "one of the first" would be lol.


Alright, can't verify it - but I changed our post : )


I like the site. Very nice job. I don't know how big your market it is. It appears to be targeted at developers, but we are the ones most likely to roll our own solution to this type of thing. I think I would be more apt to want to buy a one time license and run the software myself, then to commit to a monthly service fee.

Perhaps you should looking at a consumer market where you would do the work and then host it yourself for them giving them a link, or publish it to a subset of popular sites (facebook, flickr, youtube, wordpress, etc.) Your typical users on these sites will not/cannot leverage libraries to do what your service does. Make it easy for them to just upload & publish so they don't have to think about it.

Just my thoughts.


I like the site. Very nice job. I don't know how big your market it is. It appears to be targeted at developers, but we are the ones most likely to roll our own solution to this type of thing. I think I would be more apt to want to buy a one time license and run the software myself, then to commit to a monthly service fee.

When you're launching a startup where video encoding is one of the core component you don't really have the time to roll your own solution.

Doing an automated solution to encode videos is not easy and it takes time.

Having easy and cheap solutions like this make you gain valuable time in the early stage and even later.

Perhaps you should looking at a consumer market where you would do the work and then host it yourself for them giving them a link, or publish it to a subset of popular sites (facebook, flickr, youtube, wordpress, etc.)

Thats exactly what Particles do, they have two services ; one for encoding (heywatch) and the second one spread the videos to multiple services and provide analytics (Heyspread).


I was referring to leveraging an existing solution (open source or otherwise) when rolling your own solution. Not re-writing from scratch

I was involved in a start up a while back where video encoding was a tiny part of what we were doing and while it was painful to get working, I don't know if we would have committed to a monthly service.

Take my comments with a grain of salt since I'm not your target customer.


Good idea to service consumers, but Dropbox already does this.

Maybe charge a transactional fee and let the developer pay for the hosting with S3 themselves (most developers already have S3 accounts). This keeps the revenue directly in line with where you add value.


Do other people hate the vertical Get Satisfaction tab as much as me? I think it really interferes with this design (and many others).


You're not alone. I hate that thing, and I hate the fact that when you do actually fill out a suggestion, it forces you to create a login in order to submit it.


At least I know I'm not alone =). I [x] out that stupid box when it takes longer than 1 second to load, because apparently older versions are not asynchronously loaded, they are loaded at the time you click the feedback button.

It bothers me how it takes you to a new page, bothers me how you have to sign up to do them a favor!. Hey hey , I am trying to help YOU out here!!

I guess that's why HN is so nice, we all already have accounts to run our mouths with.

Getsatisfaction and uservoice seem to make money hand over fist, so clearly I am wrong, and clearly people love this stuff, but that just makes me all the more annoyed out how out of touch I can be with the rest of the planet.

Am I really that weird?

p.s. I think Facebook user interface for anything other than writing on the wall is horrific ... andd .... I don't even have a "real" fb account =\

The end.


Correct me if I'm wrong. But we're using Uservoice (not get satisfaction), and you can create ideas entirely without login.

Otherwise I wouldn't be using this stuff : ). But we are considering to change the style of the Feedback button now!


I am a designer so I'll just talk about the design.

It is excellent, the only change that you really must make is to add a "sign up now" button near the top (above the fold). Everything else is great design-wise.


Thanks! That's a great suggestion! Glad you like the design : )


On my list of things to do today is build an image upload page for user profile images. I've been looking at using imagemagick.

Maybe it's just me, but I have a bit of a NDH attitude towards this kind of service, and particularly at the pricing you are offering.

I'm not saying you won't be successful, that's just my two cents.

I suspect it is going to take me a few hours to figure out the imagemajick stuff. These seem to be quite a few tutorials on how to use it with Rails, so we'll see.

You aren't offering storage, and I actually think that is a good thing, as I wouldn't trust something like user images to a 3rd party start-up.

I guess I'm really wondering who your target market is.

I don't think (but I could be wrong) that the code to handle this sort of file-upload stuff is weeks worth of work. So if I can get it done in two hours, lets say that costs a company $100, and that is code we own, run on our servers, and the only further cost is storage.

If the target market is non-programmers, are you offering enough features to make managing photos and video easy enough for them?

I can encode videos on lots of sites for free, and get them served without paying any storage or bandwidth fees.

So if we just focus on photos, and the heavy lifting you are doing is the resizing and thumbnails, is that something that a newby would want to do? Or is it something that is so complicated that a programmer wants to avoid doing it?

I think what you are providing is some very basic tools, which is good because it is easy to understand, but I just don't know if the tools you are providing are compelling enough.

I guess I'll go code up what I need for my service, and if it isn't done by this afternoon, maybe you'll have convinced me that I need your service.

But I think that may be a challenge you're facing. Do most programmers think they can do this on their own?


I agree with you, image resizing is simple enough that an average programmer can get it done. The reason we are offering it, is because it fits very naturally with our architecture, and people who handle video through us will probably also like us to handle their images.

We do a few tricky things so. HTML5 multi uploading is one of them. We actually start resizing images as they come in, that means by the time you are done uploading 100 images, we are done resizing them. Another thing is upload progress without flash, that will usually require a custom-compiled web server of some sort. And there is meta data extraction and uploading to S3. If you just need a fraction of all of that, sure - we have little to offer. But we'll try even harder to make people regret their decision to "build it themselves" in the future : ).

Let me know how far you get this afternoon!

Edit: Our pricing will go down substantially. We basically decided to launch with the highest price we may get away with in order to go down with pricing over time instead of up.


Be sure to check out: http://github.com/thoughtbot/paperclip (this also has an amazon s3 integration api)

I share the same sentiment in that I know I could code out a solution in an afternoon and avoid spending 50 bucks a month for the rest of eternity.

But I also do understand what OP is pitching. Services like these make prototyping that much faster, so I see it as a win for the agile development community. If the prices come down to "scrappy developer" levels, then much more people here on HN will certainly get it. But hey, your first job is to make money so what the hell do i know?


This is interesting. Having been part of a startup which focused on video upload, encoding and streaming I think there is a market for this service. Past experience has shown me that one of the unplanned significant overheads of a service like this is the myriad of esoteric codecs out there and what they do to ffmpeg and related encoding tools. There is always a source or device that produces a video file which causes ffmpeg to hang or segfault on transcoding, or to output a flawed transcoded file with poor video or no sound, etc.

I'm curious how you deal with these 2 issues. The first issue being an infrastructure issue, where incoming media can cause problems for your tools, the second is a quality issue, where the output your tools produce may not be what the customer expects.

What is your plan to handle these 2 problems?


Thanks for the feedback. This is exactly the reason we are doing this, video uploading and encoding is one hell of a nasty problem - something people would be glad to knock of their todo list.

In terms of dealing with the 2 problems: We've been working on this startup for 1.5 years by now, so we learned a lot of the things that aren't working by now : ). Basically we fix problems as they come in, try to get advice from the right people, and be smart about the way we deal with the command line tools. Thanks to node.js we can easily "time out" on hanging FFMPEGs and detect other weird error conditions.

Anyway, I think our competition is ahead of us in terms of the encoding, but we'll catch up to them! We got the lead on uploading for the meantime : )


I'm always curious to know why every startup launches with a "Most Popular" plan.

Isn't that lying to your customers?


I'm wondering if something that said "Best Value" would be most honest. Or similar remove the "Most Popular" wording and just make that option bigger.

I suspect the answer to your 'wondering' is that the startups have copied what 37 Signals is doing.


If one plan is "Best Value", it implies the others are not.

"Most Popular" is right. It simplifies the decision-making process on behalf of the user. It leads and reassures and solves the "what do other people choose" problem.


I personally think that "Best Value" is better verbiage anyways. When the news about the economy is consistently bad, you can catch people's attention with the promise of value or savings.


You are your first customer. So, the "Most Popular" plan should be the one you would choose for yourself.


Good point. We also considered "Best Value", but that's only the case for the biggest plan.

We finally decided to stick with "Most Popular" and not worry too much about it. But yeah, we might change that : )


I really like the website, the API is well documented and the features looks good.

In my company we're currently using heywatch for encoding and after a rapid calcul it looks like you're cheaper.

We're paying 5c the video encoding and 10c for HD quality.

I'll have a deeper look this week.


We'll probably even lower our pricing in a few days, so we'd definitely love to talk! My skype is 'transloadit' !


Question: Is there a more secure way to set the parameters other than have them in a hidden form field? It appears that the end user can set the parameters just by changing POST values. I'd prefer that not be the case.


2 things. Step 1 is to "sign" your post using our authentication:

http://transloadit.com/docs/authentication

Step 2 is to use our templates to hide your params entirely from the eyes of your users. Templates use 256 bit AES encryption:

http://transloadit.com/docs/templates

HTH : )





Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: