Evaluating Startup Ideas

October 20, 2009 | 3 Comments

Someone just asked me, I responded with the following in less than 2 minutes [possibly 1 minute]. What would you tell someone that asked how to evaluate his/her startup ideas?

Evaluating ideas — talk to people; tell them your idea; what problem does it solve. Great if it solves a problem you personally have — because then you know a need exists [if you didn't find a solution already in all your research]. Start small and ugly. Don’t go into debt — 1 in 100 ideas that are executed actually become something. [I made that number up, but it's close enough -- don't gamble everything you have; the idea is to get other people involved so you have less risk... if others put money up, or will buy your thing, or use it, then you know you have something]. Stick away from pitching friends — they’ll likely nod their head, think you’re crazy, but won’t tell you because they don’t want to hurt your feelings or burn out your enthusiasm/passion that you obviously have for whatever you’re pitching them.

Posted by Steve Poland | Filed Under Ideas 

100+ Web Start-up Business Ideas

April 29, 2009 | 12 Comments

With this recession and unemployment so high, there’s likely a lot of people looking to start their own business. Hopefully one of these ideas of mine can provide some inspiration to just one of those people.

I’m formerly a TechCrunch writer, a serial web entrepreneur, and an idea guy. I can’t believe I’ve posted 100+ web start-up business ideas in the past couple years. Below is a compilation of them. They are all for the taking and if they aren’t perfect, hopefully they give you some ideas of your own. If you’ve seen any of these implemented, please post in the comments as I’d love to check them out. And if you know other ‘idea’ websites, please post in the comments for others to visit.

I’m currently working on my own start-up, InSeconds, which is launching a beta in early June 2009. If you have a website, you’re going to want InSeconds (I hope!). Be sure to sign-up your email address to be apart of the beta.

My other posts about start-ups, inspiration, etc, can be found here.

Posted by Steve Poland | Filed Under Ideas 

IDEA #100 – Twitter License Plate social network

April 29, 2009 | 7 Comments

Can’t believe I have posted 100 ideas on this website now. As with any of them, they are all for the taking — I hope they bring you inspiration, or I hope it’s an idea that you’ll want to build! Let me know if you build any of them, I’d be excited to hear it! :)

It’s 8:45am and I was just leaving my friend’s house when I saw the parking sign that said “No Parking 9am-4pm on Mon, Tues, Wed”. It’s Wednesday. His car was parked on the wrong side. I shot him a text message and he switched to the other side of the street.

I’ve been in my car and seen tail-lights out while driving, or tires that are really low, or mufflers practically dragging, or wanting to offer sympathy to someone who’s car was obviously broken into, or someone that left their lights on in the parking lot, or a cop going down a line of cars that are parked illegally giving tickets (I wish I could warn those people!). I often try to get the attention of the people in the car, but typically it’s an utter FAIL. Sometimes I’ve seen single hotties in a car, wanting to know if they truly are single. Or I’ve seen people rocking out to a song and wanting to know what it is. Or I’ve seen a classic Mustang that I wanted to give props to.

There’s a solution that could happen to all these pains, if it were to be created and get visibility in the mainstream. There could be a Twitter service that someone builds, maybe with the ‘auto‘ twitter username (which Jason Calacanis happens to have reserved; I have ‘autos’).

I’d imagine that I could tweet “@auto NY:ABC-123 you need to move your car! you’re parked illegal, it’s Monday”. Anyone could follow ‘auto’ and they could specify which license plate numbers they want to be alerted of via direct message, with whatever was said. Each state, country, would need a proper syntax, but someone could figure that out. Here’s a list of ones in the USA.

This can then be extended to its own website domain — imagine Dogster, but for Autos. Each license plate will have its own webpage, people could send in photos as well (via twipic w/ “@auto NY:ABC-123″ in the subject, which will then get pulled into this webpage). People can post make, model, color — and even link actual people to the vehicle. [Monetize the site via some Google AdSense, or displaying similar cars that are for sale on eBay. Lots of money in auto advertising vertical.]

Obviously this could have a brutal side as well, of people bitching to each other about their driving skills. 

Anyhow, even if some of my friends aren’t on Twitter or whatever, I could follow their license plates and be alerted via direct message if someone says something about them [and if it's important, I could forward to my friend].

Any other thoughts on this idea? Put them in the comments. Otherwise, who’s building this out? Go for it, I’ll be you first beta tester.

Posted by Steve Poland | Filed Under Ideas 

Advertising for Real-Time Search (aka Twitter)

March 3, 2009 | 5 Comments

Real-time search is going to change things up on the web. To me, real-time search is essentially “conversational search” — what are people talking about right now? What events, what trips they are planning at this moment, what answers they are trying to find to looming questions, etc.

Twitter is all about conversation and what people are doing now. Twitter has real-time search, but hasn’t made it widely available yet — imagine when Twitter puts this search box at the top of every Twitter page; it’s going to be a game changer.

Some companies have been trying various methods of helping users of Twitter make some money, but generally putting ads in their feeds. I think this will work if they insert a link that is 100% relevant to what ever the person is tweeting about.

But what really needs to happen is another Google AdWords, but for real-time search. Text ads to people doing real-time search queries, are going to read/look different than search queries at a search engine. Some company needs to start figuring out what these ads are going to read/look like — and also how people will be searching the real-time web, because it’s going to look much different than their search habits on Google. People will use different keywords/phrases and ways of phrasing their queries.

Some searches will stay the same (i.e. “tickets for Purdue Ohio State game”), but others will change (i.e. instead of “hotel recommendations in Buffalo NY”, they might search on how others would be saying it, “hotel great Buffalo NY” — to get people twittering ‘The Mansion Hotel has been great in Buffalo NY!’).

Posted by Steve Poland | Filed Under Ideas, analysis, inspiration, twitter 

IDEA #99 – Group Learning of Anything in Weeks!

February 3, 2009 | 1 Comment

Honestly, I don’t know much about the online learning sector — maybe this is already done. But the post ‘Learn Ruby in 3 Weeks‘ by Al Abut inspired this post [despite his post being 3 years old].

He says in his post that he’s planning to learn Ruby on Rails in 3 weeks. He planned to use the book ‘Sams Teach Yourself Ruby in 21 Days‘, which is available on Amazon.com. He said he was going to start 2 weeks from the day he made this original post, which provided enough time for others to buy the book and get ready for the learning. He was then going to make a blog post each day regarding the learnings, which then others could obviously comment on and discuss.

What a great idea. I could see this idea evolving in various ways — there’d be a central site for all these group learning subjects to take place. Just like how you’d signup for a class on the Internet, or for an in-person classroom class to learn something, you’d do that on this site. Maybe someone(s) posts a subject with a curriculum — which uses a course book they designate, and/or uses various online reading (video, photos, etc) materials that are available for free.

Maybe these curriculums are in a wiki, so that others can edit them and add to them — eventually creating the best curriculum to learning any subject.

Why isn’t there a free online university? People could volunteer as “teachers” — maybe “students” pay a $5 or $20 course fee, which goes to the volunteer (much like how referrees in intramural and bar leagues get paid; or like how Meetup.com charges for their group meetup pages).

Anyhow, if you wanted to learn Ruby on Rails — wouldn’t you rather do it with a group of others? I’m one that is motivated easier by knowing others are doing something with me and relying on me.

Thoughts? Don’t forget, TechStars is accepting applications for the summer, this could be your idea to use!

Posted by Steve Poland | Filed Under Ideas 

IDEA #97 – Tagging All Photos with people, places, objects

January 29, 2009 | 1 Comment

Fred Wilson discusses a (great!) band that he went out and saw the other night — and he brings up a great idea in it, which seems pretty obvious and I don’t understand why it’s not going on now.

“You’ll probably see a few AVC community members in this photo. Too bad we can’t all tag ourselves in it.”
RAA Goodbye

People embed photos all across the web — on their blogs, MySpace pages, Kodak pages, Walgreens pages, Picasa pages, Flickr, etc. Facebook is great with their photo tagging process. Google’s Picasa now lets you do it, but I think the people have to be Google users?

Anyhow, my idea on this would really only be for blogs and it does get hairy. Blog owners would include a snippet of JS on their page — when an image is found, a visitor could click on a part of it and tag it with a user or anything else (celeb; object; etc). But it gets hairy, because it might see any image — even a heading image; so somehow the blog owner would have to designate that the image could be tagged. Then where do the tags come from? Maybe you could tag it with a user from twitter or mybloglog, or any other service that has public users (last.fm; disqus)?

Or maybe you use Facebook Connect on your blog pages somehow, and are able to tag the photos with Facebook users? (I don’t even know if that’s feasible).

Maybe this idea isn’t practical at all :)

Posted by Steve Poland | Filed Under Ideas 

IDEA #95 – URL Filter for filtering Adult websites using Google SafeSearch

January 28, 2009 | 7 Comments

It seems to me that there should be a web app with an API that I can easily send it a URL and it’ll toss back to me whether that URL is an adult website, or has been scanned and has adult keywords on it, or the URL has been officially designated by the company as a website with adult content.

I realize no filter will be great, but I’d like to have an initial filter. Google has ‘SafeSearch’, which will eliminate URL results if you have it turned on.

Does anyone know any actual web services that do what I want?

Otherwise, here’s an initial solution (which someone could create quickly I’m sure and I think it’d be useful to some people; ah-hem):

Do a Google search with the parameter of either ’safe=off’ or ’safe=active’ [these are hidden vars, but also appear to work in the query string].

What you could do is a ’site:(url)’ search with SafeSearch turned off, then with it turned on. If Google shows anything with SafeSearch off for that URL, then you know Google has the domain in its index. If then it shows nothing with SafeSearch turned on, then you know Google has blocked the site for adult content.

Here’s an example with SafeSearch on and subsequently, off.

I guess the easy way to parse Google’s results for whether this is works is the keyword phrase “did not match any documents.”, along with “Make sure all words are spelled correctly.” [just to make sure that phrase wasn't in the actual results].

I’m likely going to get this programmed as I need it for my startup. I have a great domain for this idea (IMO), FiltURL.

Update 1: Sam built this quick and it works great! We’re going to put it up at FiltURL — anyone want to make this page look semi-decent graphically? I’m picturing the homepage of this site to have a URL box with a submit/go button. The homepage will have info on the app, as well as how to use it [like tinyurl]. We’ll also throw some adsense code on there. There doesn’t even have to be a 2nd page, we could just show the result in ajax on that page if the user inputs a url on our page manually.

Update 2: Nate comments below that AWIS can provide this info at $0.15 per 1000, but I wonder how limiting this is, because Alexa’s site doesn’t contain every single URL out there (and does it tell if a site is adult or not — if so, wouldn’t it be listed under their ‘adult’ category?). It would need to be tested. The Google route is good, but as mentioned in the comments, Google may block the IP; but we’re not profiting from Google.

Posted by Steve Poland | Filed Under Ideas 

IDEA #92 – Virtual BarCamp (meeting of tech minds)

January 23, 2009 | 1 Comment

A virtual BarCamp went down in 2006 where it was in real-time and people logged in, blah blah blah. This is not what I have in mind.

I want someone to create a simple website that likely taps into the YouTube API, and allows people to record 2-5minute videos on any topic they want, just like they would at a BarCamp. It allows others to learn what interests you, and you’ll likely make new techie friends. Plus you’ll learn from other’s presentations!

Visitors of the site can then easily browse through the videos in a random nature, or they could browse by category. [Maybe you just make it all random, as it would be at a real Barcamp -- but maybe you show 4 thumbnails of the videos coming up, along with their titles]

Entrepreneurs could use this site to find possible startup partners. Employers could use this site to find applicants/employees. People could use this to find possible love relationships.

Posted by Steve Poland | Filed Under Ideas 

ClickJacking – ideas for this sneaky hack (Twitter, etc)

January 22, 2009 | Leave a Comment

Here we go with another little browser exploit/hack. It’s called ClickJacking and as Scott Jangro puts it, “like carjacking, but with clicks.” Scott has a great post with a screencast that shows how ClickJacking works. Thanks to @esnagel for making me aware his post.

Basically, you can create an invisible (opacity=0) IFRAME over a ‘click here!’ image/button/link [or over anything you want]. In the example (which comes from James Podolsey’s blog), they show a ‘click here!’ image, which has an IFRAME that’s invisible and shifted using CSS, so that the ‘update’ button of twitter.com/home is over the ‘click here!’ image. The update then has whatever you want in it. The user then clicks the ‘click here!’ and actually submits the update without knowing they just did that.

This of course assumes the user is logged into their Twitter account, which most users are — if they have Twitter.

There are some sneaky things you could do with this — blackhat of course:

  • You could do this with Facebook as well, because most users are logged into Facebook. Stuff could go to the user’s newsfeed possibly.
  • Tweet out something actually relevant, with an affiliate tagged link to earn commissions. [Although that'll be trackable back to you]
  • You could have the user click a search result without them even knowing it, and collect the CPC. Or force a click on Google AdWords. Or force a click on a display ad.
  • You could potentially force the user to thumbs up / stumble a webpage you determine.
  • You could submit a comment to a blog with prepopulated data that you filled in (maybe pulled from a DB, so that you already had different data for every user you scam into clicking).
  • You could submit positive or negative reviews — prepopulate a review box and get the user to click it.
  • You could get the user to click an Amazon, EBay, etc, referral affiliate link, so that a cookie gets set on the user’s machine.
Of course, I’d tie this script in with the DOM hack that can tell you what websites the user has visited in the past (i.e. ‘twitter.com/home’ to figure out if they are likely logged into twitter; or ‘amazon.com’ to see if they visit amazon; etc), then I’d have my IFRAME somehow relate to this new knowledge of the user.

But you could do whitehat stuff — that ‘click here!’ button could be almost like a ‘ShareThis’ link on a blog article page. You could encourage users to ‘tweet’ that they are reading this current webpage — they’d simply have to click a button ‘tweet this!’, and it’d post as a tweet, without the user leaving the page they are reading. Rather than the user clicking a link that includes what the tweet would be, the user then going to their twitter page with the update filled in, and then having to click ‘update’.

What ideas does this hack give you? Share in the comments.

Posted by Steve Poland | Filed Under Ideas 

IDEA #91 – Captchavertising (”captcha advertising”)

January 22, 2009 | 2 Comments

Once again, I’m cleaning up my email and draft posts — this was one from a year ago (Jan 29, 2008). I’m not going to take any claim to the idea. Boris (of The Next Web Conference; and also he started TwitterCounter) and I were spitting back a few emails, because he had a script he wanted written for his site (I’m unsure if he’s still doing that business, I don’t see the live demo anymore; but this idea was hot).

He was at the time doing ‘Captchavertising’ (I don’t know if he coined this term, but I have an email from 1/29/08 with him using it), but it doesn’t appear he’s offering Captchavertising on his advertise page (although he may still be doing it). Basically, when an advertiser purchased ads on his site, he’d then use their company name in his captcha’s on the site.

I think this is a great idea that could be used across the web. If any of the comment systems are using captcha’s (IntenseDebate, Disqus, or generic commenting/contact us forms, etc), this could be a way to monetize them.

I know we’re all off saving literacy with recaptcha, but this could be a way for Publishers/bloggers to earn a few extra dollars each month.

The captcha could be the company name or slogan (limited to like 15 characters) and a number next to it. It could also show their logo; or the captcha could be their logo.

So someone… anyone… if you can figure out a business model for this and an easy way for Publishers/websites to use it with their contact forms, forms, comments, etc…. go for it!

Posted by Steve Poland | Filed Under Ideas 

Next Page »