One of the issues facing any social-network based startup is building a community when none exists already. A classic chicken and egg scenario where you need users to create the content, but without content users aren’t interested in contributing.

After a discussion with my lead developer I realized that we need to feed Other Weather a steady stream of weather forecast data to “salt the tip jar” as they say, or to seed the community. As a result, I did some research online and came across the NOAA website (National Oceanic and Atmospheric Administration) and their weather service XML API.

With some crafty programming we are now taking the NOAA feed and implementing this forecast data into OW for the current and next day forecast. Instead of pulling a forecast for every zip code in the country (currently 99,999) we are pulling a sampling of data across each major zip code range, which means you’ll see weather forecasts spread across the country, with a higher clustering around more populated zip codes (as they are more of them). The NOAA gets their own account (and profile) with Other Weather, and each forecast is available for rating just like a real user’s forecast.

This isn’t a perfect solution however, and one we’re working on improving. My goal is to take in more data from the NOAA feed and implement some sort of scaling so that at each zoom level you can see forecasts covering the map without it getting too cluttered.

My hope is long term people will compete with the NOAA feed and delivery more accurate forecasts which give an more accurate picture of weather across the US.

/Back to work.

This isn’t an exercise in positive thinking, it’s rather a reaction I had when I finished reading this article I found today via Y Combinator Hacker News. The article was posted by Steve Spalding of How To Split an Atom and covers some basic startup shortfalls.

Here’s the run down of each category and how Other Weather fairs:

Obviously, The Product Will Sell Itself

I’m working on this one, and promotion takes time and lots of it, or if you lack time you can spend lots of money. We’re not sitting around betting that our weather forecasting site will just “take off”. 4/5

People Care About Features!

We launched Other Weather with a tight, but I think pretty nice set of features. I’m not adding any features until I “sell” users on this concept. Once that’s done we’ll add features based on user feedback, although if we had the budget I’d add some essential features (not OpenID or any of that). 4/5

What Do You Mean We Don’t Need A Private Jet?

Spending is not a problem when you essentially fund your startup our of your pocket. 5/5

No One Told Us People Would Actually Use This Thing

This is related to features, but I think that the user interface of Other Weather is pretty good, and can actually be used. I posted about one of those features a while back (Multi Day Submit) It could use improvement, but again budget and ’selling the product you have’ comes into play. 3/5

Early Adopters Aren’t Real People

I’ve listened to early feedback on OW, but I’m being wise not to act abruptly and widely to any criticism thrown at my way. This is another area where Other Weather needs to improve, as we’ve had a few requests from users that have been shelved for a while. 3/5

–

So out of 25 possible points, I’ve evaluated my startup at 19. Not bad, but not good either, we definitely need to improve our efforts in listening to early users and adding features that make the process of weather forecasting easier.

One of the main ideas behind Other Weather, and why I think it will be truly superior to traditional forecasting once people take full hold of the concept is Crowdsourcing. Wikipedia unfortunately has a rather crappy definition of Crowdsourcing which doesn’t help people get a better idea of the theory or “neologism” is the Wiki author tags it.

here’s the text:

“Crowdsourcing is a neologism for the act of taking a task traditionally performed by an employee or contractor, and outsourcing it to an undefined, generally large group of people, in the form of an open call. For example, the public may be invited to develop a new technology, carry out a design task, refine an algorithm or help capture, systematize or analyze large amounts of data (see also citizen science).”

The use of ‘outsourcing’ as a term is pretty limiting, as does the idea that it has to be an “open call” (which brings to mind fake corporate campaigns to gather customer data). The second sentence does the idea much more justice and conveys at least a basic description of various uses for Crowdsourcing.

For more on Crowdsourcing, Jeff Howe did a great piece for Wired in 2006.

–

Why do I think that Weather is a prime candidate to help further this idea?

Weather is a local thing, a very local thing. It’s also something that can be predicted well by people who have time to watch and understand the weather where they live. Why not harness this local knowledge to not only predict future weather, but also report current conditions? How useful would a social weather forecasting site (updating in real time) be to people facing a natural disaster like a tornado, hurricane or wild fire?

Which brings us to the “social’ aspect of Other Weather. Yes every startup and dotcom 1.0 entity is tacking on the ’social’ tag these days, but for OW it’s completely different. The need for social contribution is essential to OW, it’s not a tacked on feature at all. Without the social, OW doesn’t can’t exist.

This is why we take care of our forecasters, and will be looking to add to their already excellent set of tools like; Multi-day submit, Personal RSS Forecast Feed, Personal Weather Forecast Widget, links, photo, text and more.

One of the neatest features we have here on Other Weather is the multi-day forecast submit form. When a registered user clicks the “Submit Forecast” navigation link, they are presented with a simple 1 day forecast form with a Weather Summary Key and 4 other fields (Date, Zip Code, Temperature and Description). If the user wants to submit more than 1 day at a time however, this is where things get really slick.

Clicking “Add Another Day” dynamically inserts another (slightly more streamlined) form with AJAX without reloading the page. This allowed the forecaster to quickly add forecasts for 1 or multiple zip codes on 1 form and in 1 action (saving time and resources).

After the user is done, they simply click Preview, which takes them to a confirmation screen where they can review and confirm each forecast as it will appear on the site, before submitting.

This is just a simple example of how basic AJAX can improve an interface and reduce the number of steps a user needs to complete to perform a function in your application.

At Other Weather we believe in the power and wisdom of crowd sourcing for predicting and reporting the most accurate weather.  To accomplish this goal of utilizing a social community we employed an easy 5 star rating system for each and every forecast submitted.

We used an AJAX star rating system to allow users to quickly and easily rate forecasts without time consuming and server intensive page reloads.

These ratings are then aggregated and assigned to the user which allows them to compete for the Top 10 Forecasters list (home page) and overall site rank (on each Profile page)

Too long in fact, which is why I’m putting an end to that and posting today.  The Google map on the home page of Other Weather is looking great, with weather forecasts showing up all around the continental US.

We have many states and towns reporting weather including; Harrisburg PA, Danville VA, Marietta GA, (warm) Captiva FL, (cold) Shreveport LA, (wintery) Saint Louis MO, and (coldest) Grand Rapids MI.

The posting schedule will be more regular from now on as we’re entering our groove again and making plans to promote OW for the winter/spring of 2008.

I finally got the Wordpress theme customized for the OW blog, and decided it needed it’s first real post.  Other Weather dot com is only days from formally announcing the launch, and we’re very excited around here in anticipation if this rather large and important event.

We’ll be promoting OW on social networks, social news and bookmaking services, weather blogs, weather websites and various influential tech blogs.  We’ll also be releasing some sort of Press Release (once I write it), as there are several sites online that allow PR submission and ‘distribution’.

Stick around, this is going to get good.