Roast my Video Editing Startup

[sofia555]

My startup wants to address the following problems related to editing personal videos:

a) Users have hours of footage but don't know the content of each clip or have a good way to organise it.

b) Users don't know how or have the time to create a high quality, professional looking short video clip from their footage.

I'm working on a service that uses various algorithms and visual recognition to organise and extract the best footage from within larger video clips. I've also developed an online video editor that stitches clips, adds music, transitions, effects, etc...

I'm following my passion for action sports (surfing, skydiving, snowboarding etc..) and will initially target action sports film makers (i.e. GoPro owners) but see this as something that could be expanded to larger audiences - family, holiday, etc...

I have put together a landing page which outlines how it works: http://sickedits.com

Thank you for your time.

 

[cristianalexandr105]

@sofia555 Some problems you may encounter:

Internet upload speed tends to be slow in home connections. In the US, the current average is around 1MB/s (source). If somebody returned from a trip with 16GB of footage, it will take them 4.5 hours to upload, they may loose their patience.

AI recognition of what is "Interesting" is not a trivial task. You can teach a computer to recognise patterns (like hand written letters, or the shape of a smiling face), track objects or even get physical properties like speed or acceleration from a video, but "Interesting" isn't something that translates into math that easily.

Video processing might be a mine field of patents by the likes of Apple, Adobe, Sony and Google.

Video editing is very resource intensive. It might now work well over an internet connection or when partially executed in the users web browser.

Your most direct competitor would be iMovie. It doesn't require uploading videos and it is running natively on the computer, so it will always be faster. On the other hand, it doesn't have AI recognition (the hard part), and it is limited to only Mac owners, so a very limited target audience. You can be more accessible than iMovie, but it might be difficult to offer a better movie editing experience.

 

[sofia555]

@cristianalexandr105 Thanks, this is great feedback and you echo some of my own concerns:

Uploading speed - I'll need to prove this is indeed a barrier but one idea was to have a kind of hybrid app that maybe compresses the footage before uploading it or maybe even do all the AI classification locally. Also in the future hopefully upload speeds will improve.

What's Interesting - agree, this is pretty subjective which is why it will only provide assistance and not be fully automated. Over time based on what people determine to be interesting and responses to video edits the system can be retrained to be more accurate. Also by initially focusing on a niche market it's easier to tag what is interesting. I.e. Surfer is standing on surfboard. Again this is only an area that is going to improve over the next couple of years.

Patents - not really looked in to this but will keep it in mind

Resources - the first thing I've built is a scalable server based editing/rendering engine, users resources are not tied up in any way which is a plus. Also I'm thinking of only allowing short videos, 1 minute so as to limit resource usage and anything longer could get boring.

There are a few competitors i can think of but none of them meet my needs at least. Magisto is a great service but it is very limited and everything is automated - it decides what is interesting. They go overboard on effects and overlays too in my opinion. Quik (bought by GoPro for $50m) is also nice but its an app and only works with mobile footage at present. GoPros own app for desktop is also really cluncky and crashes a lot. The workflow and templates leave a lot to be desired.

So there seems to be a market and some available solutions but nothing that feels quite right.

 

[runestar]

@sofia555 I think there might be something in this space, but I don't think your idea is quite it. I think there is a need to easily take everyone's videos from their phone and put them on a DVD as an alternative to youtube. Yes people still use DVDs. It's the equivalent of printing paper photos out.

 

[sofia555]

@runestar Thanks for the comment. I'm sure there are still people out there that use DVD but it is an ever declining market. It is now pretty common to just stream directly off your phone/tablet/PC to your TV via Apple TV, Chromecast, Xbox or Smart TV - your DVD/BluRay player may even support this - the technology DLNA has been around since 2003. At the very least the app I am building will be able to create a video that can be streamed to the devices above.

 

[runestar]

@sofia555 This is true, DVDs are going to continue to decline. But there is a problem with storing files digitally. Maybe DVDs aren't quite the answer, but the problem is mainly organization and access. It's very difficult and time consuming to organize your files in such a way that they will later be easily recognizable and accessible. Putting them on a DVD is like making a mix tape. The digital version of that is a playlist. But for example, my wife takes 10 photos and 10 videos of our kid every single day. The easy solution for me right now is to just drop them all onto a dvd. Otherwise to watch them on the TV requires fancy thumbwork.

 

[hu7231]

@sofia555 I think this general space of video editing + AI is a very interesting and valuable one.

And the action sports market does seem like a potentially good early adopter market to start with. Though I wonder if something outside the consumer market and in the business market might offer an easier early fit, easier market to open up their wallets if it simplifies how they create content for their business. But that may not be were your interest is.

You can offset some of the processing-hungry tasks -- such as using encoding services that you can connect to via their API and pay a monthly fee.

Re: selling the idea on your landing page in these early days: you might consider adding a quick video hero section background? http://dissolve.com/video/Slow-moti...alty-free-stock-video-footage/001-D187-30-010 (or one of your own of course)

I'm at a service talking about video -- with no video.

And better, if possible without being misleading: you might consider an example edited video -- even if it's produced from a highly modified version of what the real AI that will eventually power things.

Best of luck.

 

[sofia555]

@hu7231 Thanks for your feedback. Interestingly a bunch of sites doing a similar thing all launched business focused services in the last few months. I guess it's a telling sign that there is demand there and where the money is to be made. It's not quite what I was aiming for but the technology could easily be adapted that way.

Thanks for the link to Dissolve, some good footage on there. Definitely agree I need some examples and a background video would look great too. Will get to that on the next version closer to launch.