I had a long conversation the other night with an entrepreneur about her product. While sitting at a bar near the NY Tech Meetup, we brainstormed a slew of new features.
I confessed, though, that none of the ideas I gave her were original. I see as many new ideas and try as many sites as anyone and I'm really good at pattern matching. I can reappropriate useful tools from other sites for relevant situations pretty quickly, making me quite the snappy little regurgitator.
So, if you ever meet with me to talk product, this is what I'm going to suggest you build in:
(I think they can almost universally apply to any site.)
It's a simple concept:
- Built up a database.
- Structure a limited amout of link dense template pages
- Make it all public.
On BR, you find a baseball player, all the teams he's played on, the rosters for those teams, more players, more stats and teams... player, team, player, year, team, player... click click click... You can go sideways around that site forever. Same with IMDb... movie, actor, movie, director... click click click.
So what does your database have?
How few templates could you build to expose the whole thing?
Sportsvite for example, helps you manage rec sports teams. They have fields, games, teams, players, sports. A public exposure of those templates could get them a lot of traffic, especially with fields. Ever try to Google for directions to a softball field? It's a nightmare. A Sportsvite page with that field, directions, schedule of games, teams that have played there, photos from games, and some hooks to sign up for Sportsvite would quickly rank pretty high. You could then click through to the sports played, the teams, even the localities.
2) The selfish data sucking helper tool - Hubspot created a brilliant lead gen tool for themselves and others with its Website Grader. You submit your website, it grades you on various SEO stats and requirements, and then gives you the option to save a report by giving your e-mail.
Not only is it a pretty useful tool, but its going to be great leadgen for their sales team.
The key here is that if you want data, you need to provide something useful first...and for your first user if you're a social site, it's not going to be intros to relevent other users. That's why del.icio.us and Flickr worked so well. They both worked just fine as tools for yourself, so you didn't mind that you were also contributing tag data. So what kind of stats and info could you give to your first user that gets them to handover valuable data to you? User #1 has to have something useful to do, and calculation/discovery tools that ask for a user's data are a great way to fill your empty database while still providing value.
3) Kiss the ring management tools for groups - Angelsoft is software that helps angel groups manage their deal flow. The company could have just as easily published an open website whereby companies post their own financing needs and angels just go searching. That would have eliminated the need for angel groups, many of which survive through paid membership.
Why not do that?
Well, first of all, creating direct to consumer businesses on the web are hard. It's hard to generate traffic. Plus, you'd piss off a lot of angel groups who would badmouth you by trying to get in on their gig. But, if you can build something useful for existing powerholders in your space, people who already have traffic, it may be better off than going out on your own and everyone wins. The key is if you can piggyback off groups like this and still somehow leverage your own business to be disruptive.
I'm experiencing that in Path 101, where rather than compete with career planning offices in colleges, we'll most likely build some tools for them to use to make their lives easier. This way, they're more likely than not to promote us as well. Still, I'm not depending on them either.
4) Crawled data - If you can find a way to gather data that's already out there to pre-populate your website, it's a highly cost-efficient way to build your business. Otherwise, you often wind up building big empty databases. Google does it with local reviews. I don't think very many people would care to put reviews into a Google database, but there's nothing stopping Google from just crawling reviews it finds on Citysearch and other similar sites
This way, again, you have something to show user #1, instead of 100% relying on your users to build your business for you
5) Revolving e-mail door for data - Creating user generated content sites depends on the ease of being able to get content into the system. One of my favorite ways of doing this is to rely on e-mail. Contrary to narrowminded belief, e-mail is not dead.
In fact, more and more people are getting smartphones so they can be connected to their e-mail 24/7. Sites that take advantage of easy ways to add content via e-mail will benefit from this trend. Disqus, for example, now allows me to respond to comments posted on my blog, and post them up on the web, 100% over e-mail. They've gotten past the need to force everyone to visit their posting page, because they realize that the value is in the data. The more people comment the more useful the service gets for everyone, so allowing comments to get responded to more easily by just letting me post by e-mail is brilliant.
6) People like me - People like me can be a simple feature or a hard one... it all depends on how many PhD you want working on it. I'd be very interested in seeing people like me across many of the sites I visit, and I think last.fm does it best. (Amazon shows me what people like me bought, but not who those people are.)
I have friends on last.fm, but I don't get much value out of them because they don't listen to the same music that I do. However, my "neighbors" are awesome. It's a totally passive feature and I didn't have to do anything to get a list of people with similar music tastes for me to explore, friend if I want to... that's what social should be about.
I want that on Twitter. Do some text analysis. Who twits about the same stuff that I do? How about on Newsgator? Who reads what I do? Anytime you have a lot of data, you can instantly make a site more social, without all the fuss of "friending" by showing me people like me.
7) Temporary accounts - Path 101 had a great meeting with David from Tumblr about integration. It was pretty obvious that he's had this in his mind for a while--how can he make his site friendly for partnering?
People who have intense registration processes often find it difficult to create partnerships. Who owns this data? Will you pass it to me? Is a partner account a real account? Whose logo will they see? It's completely stupid because no one really "owns" the user anyway. You want to own data, not users.
As long as you get data back, you should do all the white label partnerships you can (unless perhaps you're building a network tool that just would splinter an audience and degrade a product). In order to do that, you need to develop a way for your system to work without a full registration.
What's the least amount of data you need to run an account for someone? Answer: a single unique identifier and a way for a partner to authenticate their sign-in. That's it... no name, age, birthday, zip, etc... If you can get those from a partner, great... if not, now you have more users and more behavioral data (plus all the content) and you should be able to monetize that.
Blogged with Flock