I see a lot of cloud only services out there, meaning web sites that don’t do much more than connect a variety of cloud services, often called a mashup. For example, taking Craiglist postings and showing them on a Google map was one of the first cloud only services that I saw but there are thousands today. Cloud only is a bit of an exaggeration as there is logic and often unique data stored within the app but the bulk of the value lies in external servers and often with terms of service that allow the owner to pull the rug out at any time. It would be hard at this point for Google to turn off its Map API but they could and thousands of sites would be out of luck or would quickly have to scramble to bet on another service from Yahoo! or MSFT but the same risk would still apply. Now betting on Google right now is a good bet but you are leaving the cards in their hands. At the same time, the opportunity is amazing. Knowing next to nothing about mapping, geo-coding and the like you can quickly build apps that blow people away. Now bring in photos from Flickr, tweets, Facebook friends and thousands of other services that provide APIs and the possibilities are endless. But when you step back and think about the long term costs and viability of some of these services (Google Maps is a bad example as Google is very strong and Maps have a great ad model) it can be scary to bet your business on them. For example, Flickr – that is a lot of storage and bandwidth for sharing those photos and the ads only show up on their site but if you don’t visit Flickr then you don’t see the ad and they don’t make money even though they incurred the cost. Long term, I wonder how viable Flickr’s openness is and yet there are many sites that have bet on Flickr being around and remaining open. Hmm. It is still a better bet than building your own photosharing capabilities but it does make you wonder.