Discuss Home · Bootstrapped Podcast · Scribbleton Personal Wiki · HelpSpot Customer Service Software

Vidgen - Generate Summary Videos of HTML Content


#1

https://vidgen.io

This is a project I’ve been working on that generates basic slideshow videos using HTML content. I’m releasing it this week in hopes of making it eventually profitable.

Confession time. I’ve made the typical developer-turned-wantrepreneur-dude-lets-just-code-this-thing mistakes:

Worked on it heads-down for the last few months? Yup.
Gathered a bunch of market leads, email addresses, phone numbers? Nope.
Tested my “business hypothesis” by contacting potential customers before making this thing, and discussed how they might want something like this to work? Double nope.

Dammit.

Before writing any code, I probably should have reached out to publishers that use video to market their content, and figured out:

  • what their current workflow is like
  • how they’d prefer it to work
  • what it would take for them to switch (features, pricing, support, etc)

and then build the thing based on that.

But, I did research the existing competition before making it. I think their successes gave some validation to my idea. They are mostly recent startups. Some of them have received millions in funding, and have big name clients (ie. many of those subtitled videos you see on USA Today were made a competitor).

I’m hoping it will appeal more to smaller content creators as it’s far more accessible and cheaper than the competition. But the next step (should’ve been 1st) is find out if that’s actually true by talking to them, and then improve the quality of the product based on that.

Feel free to leave some sugar-free feedback.

FREE STUFF I’m trying to expand the “Example Video” gallery on the Vidgen site. So if you generate a decent video of your content and want it for free, contact me at [email protected] with the video’s url, and I’ll enable it to be downloaded. I just ask that you put it on youtube/vimeo and then send me the embed link. (* limited time only)


#2

Hey @timbowhite, gave it a whirl! Unfortunately, i didn’t work :frowning:

My first attempt was with ‘google.com’ and it spit back an error saying something about adding a correct url. It looked correct, but then figured what you meant was making sure to add the scheme so i retried with ‘https://google.com’ and it started running…

…but then it errored out.

So then i decided to give your site a whirl :slight_smile: by adding ‘https://vidgen.io’. It finished, but the video generated had nothing to do with your site. It looked like some generic demo video.

Anyways, that was my experience with it. Tried it and it didn’t work for me.

Hope that helps!


#3

Hey Chris, thanks for the feedback. Vidgen is meant more for online content, ie. blog posts, articles, etc. Pages lacking a narrative/text/images probably won’t work out very well.

That being said, I should probably ensure that submitting the site’s own homepage actually creates a watchable video :slight_smile:


#4

No prob! Hope it helps…

Gotcha! I didn’t get that in all honesty :slight_smile: Probably just me though…

Wondering if this isn’t a faint signal saying that your own landing page should have more “narrative” :slight_smile:

Also… just tried it again with a different site and get:

There Were Problems. Your IP address has reached the allowed 10 submissions in a day. Please try again later.

I think the concept is great and having something to demonstrate to potential leads/customers is wonderful, it looks like with a bit more time invested on the software side you could have something solid to show off! :slight_smile:


#5

First time I tried it, I got told I had reached the daily limit…

Confession time. I’ve made the typical developer-turned-wantrepreneur-dude-lets-just-code-this-thing mistakes

Like most of us, I suspect, if we’re honest… :wink:


#6

Haha good point!

Also… just tried it again with a different site and get:

@SteveMcLeod Oops, should be fixed now, the IP submission limiting was buggy.

I think the concept is great and having something to demonstrate to potential leads/customers is wonderful, it looks like with a bit more time invested on the software side you could have something solid to show off!

Appreciate it!


#7

Wow what is this sorcery? Fantastic idea Tim. I’m sure this is counter-intuitive, but I think you should also provide a manual option. I just made this video and you can see, while the tool does clip important sentences from across the post, it does not make a coherent video in the end : https://vidgen.io/vjm/a7Wyv36XxzB6U

Maybe a manual option would help me pick the important sentences myself so that the video makes more sense.


#8

Wow what is this sorcery? Fantastic idea Tim.

Thanks much!

I’m sure this is counter-intuitive, but I think you should also provide a manual option.

Not at all counter-intuitive. I’ve actually already started working on this. As you mentioned, the automated summaries and image selection aren’t perfect, and potential customers are going to want their generated media perfect before they’re willing to pay for this service.

Unfortunately, my work on the manual video customization has stalled as I’ve had to focus on projects that are making money. ETA for its release is looking to be the end of May.


#9

maybe pre-sell it first to your list if you have anyone on it?

there are companies who send custom slideshows and/or videos in each cold email, your service is somewhat similar to those, you might find them interesting w/r/t pricing.


#10

@timbowhite how is it going? Have u got any paying customers? Who are your competitors?

So im the same as you. Im a dev. And i hate all this preselling idea. Im not saying it doesnt work at all. Im just saying im no good at it.

My validation was there were competitors whom i thought hadnt captured the entire market and that there was room to improve upon. Luckily im half decent at seo and knew i could convert 5% of leads through optimization of the target keyword. You should think about doing the same. I never have to speak to customers apart from email really. And never have to ‘sell’ people. They come to us because they are searching for the tool. They are essentially 80% sold.

Keep going mate


#11

Hey thanks for the encouragement! I’m still working on the project when I can, but it’s not making any money right now.

I probably should brush up on my SEO skills. However, isn’t the key part of getting to the top of the SERP powered by getting linked to by relevant and reputable websites? A.k.a. link juice? I’ve tried submitting the project to multiple “startup just launched” list and directory websites. Didn’t amount to much traffic. I probably should also reach out to some relevant blogs and news sites to see if they have any interest in writing about the project, or allowing me to do a guest post.

The other thing I’ve noticed (based on my other projects) is that google typically doesn’t give a damn about your website until it’s been live for at least a year.


#12

Quality backlinks (as the SEO community refers to links on other sites to your site) are nowhere near as important for SEO as they were. In late 2013, Google made a major change to their algorithms. Since that change, Google now tries much harder to understand the relevancy of the info on your site. In 2017 it is all about the content.


#13

Really interesting concept with a lot of potential.

A couple notes on my likely personal needs:


#14

@SteveMcLeod this was the Penguin update, right? So google is essentially trying to independently validate and verify content quality with far less dependence on backlinks? l suppose I’m out of the loop - overall their results seem to be working pretty well, with a few exceptions like this one, check out the 1st organic result, yikes):

@michaeljamescalkins thanks for the detailed feedback, really appreciated. I agree 100% with your points. This thing probably isn’t ready to be a product until each one of those is addressed.

I’m working on implementing Amazon Polly TTS right now - here are some samples of a few of voices, would you mind telling me if you think they sound too robotic?

Kendra: http://mamchenkov.net/wordpress/wp-content/uploads/2016/12/english-us-kendra-linux.mp3?=1
Bryan (British): http://mamchenkov.net/wordpress/wp-content/uploads/2016/12/english-british-brian-linux.mp3?
=2
Geraint (Welsh): http://mamchenkov.net/wordpress/wp-content/uploads/2016/12/english-welsh-geraint-linux.mp3?_=3

And if they are too robotic, do you think people would be willing to pay significantly more for videos narrated with real human voices instead of TTS?

Audio sources from: http://mamchenkov.net/wordpress/2016/12/04/amazon-polly-text-to-speech-in-47-voices-and-24-languages/


#15

I feel like Kendra was the best sounding by far. A choice between male and female versions of this person would be best.

I like the idea of automation because I can have my video instantly. (or audio? separate feature?)

Honestly you’d need to ask around to see what people would be willing to pay for this. It’s a very new concept.

Good job on this!


#16

Nice job with this man. As mostly a dev myself, I assume that you at least got some valuable learning experience with video editing/automation/scraping/parsing/etc and now probably have a semi-decent set of tools and knowledge that can be used for all kinds of video related apps.

Now, I personally wouldn’t really see the need to convert a website or even html into a video, but can see some other (or additional) potential ideas using this or similar tech.
For instance, it may be useful to allow people to connect to their fb/dropbox/gphoto/iphoto/wordpress/joomla/zapier/etc folder/albums and create a video from the pics and some sort of README.txt or a textarea, website, etc.
Content writers, news outlets, bloggers etc could potentially have a need for something like this that’s simple to use to quickly insert a video into their posts/news without many delays or even 3rd party humans involved. Hell, regular people could even use it to spice up their feeds and social profiles. Don’t forget those share my video buttons and/or upload to yt functionality.
As far as I see, reading the narration from a text file instead of from html should be trivial, and as for the possibility to access social networking folders, albums, etc, that would be API/plugin research for you :slight_smile:

Anyway, really good job on this, regardless of current monetization (or lack of).
My suggestion is just that you find a way to connect with more places for video content instead of just url/html.


#17

Thanks! Building this thing has definitely been a learning experience.

For instance, it may be useful to allow people to connect to their fb/dropbox/gphoto/iphoto/wordpress/joomla/zapier/etc folder/albums and create a video from the pics and some sort of README.txt or a textarea, website, etc.

All excellent ideas. I was going to support word docs, powerpoints, pdf’s, etc. but your suggestion could be a higher priority as it’d have more immediate utility to many more people. I agree - eventually, Vidgen should just be able to convert any media types/storage into a video.

Additionally, I think a free Wordpress plugin (as you mentioned) could really help it gain some traction.

Thanks again for encouragement and suggestions!