A goodly number of organizations are on a path to incremental transformations of their business to digital. There seems to be a consensus that the user experience is a great place to start. Some organizations are starting with a "Mobile First" approach while others are looking at customer journeys. See http://jimsinur.blogspot.com/2015/05/dangers-in-first-steps-of-digital.html
Still others are looking at end to end processes that usually include value or supply chains. Since there is pressure to get going, decisions are being made on technology bases that might create regret later. While technology should not be the driver, but looking at the opportunities and constraints that each type of platform brings should be an exercise that organizations should be doing. Also organizations will have to pick an approach for integrated versus hybrid. Most likely all platforms will support social, mobile and cloud approaches. If they don't, consider this a warning sign.
Business loves to have a complete application, out of the box to support their business. While this is the dream and might be a good place to start, watch out for picking applications with rigid platforms that are hard to upgrade and difficult to leverage business differentiation. Quite often you are locked into a business model, so these are great for standard business models and functions. Beware that these approaches solidify the silo approaches as well. Look for a liquid infrastructure and easy modification capabilities.
Process platforms are quite often quite capable of supporting multiple styles of processes (emerging or static) and have enough pattern recognition and analytic capabilities to be helpful. These are what I call the "Swiss Army Knife" approach. If they have smart process applications that come with the platform, they are more valuable. If they come with rich analytical capabilities like next best action and cognitive computing features, they can be even more valuable. Make sure there is enough additive business capability to make them desirable.
Analytic platforms come with a richness of pattern recognition and decision support and work with large amounts of "in memory big data" to help guide business decisions. The problem with this approach is that they are not linked to actions and rarely work at the faster speeds required by the "IoT" This is what I call "the brain on a stick approach"
Some platforms are great at the "first mile" and support smart capture and even support multiple content types including "micro content". Make sure there is an integrated analytic and process platform. It is a bonus if they come with smart process applications on top of the content.
Some platforms are great as a data and complex event integration hub and can bring together multiple legacy applications and data streams. This is essential infrastructure for the digital world, but make sure they are more than a "bus stop" and can support being a digital nervous system that can support process and analytic capabilities as well.
Hybrid vs Integrated:
Organizations have to decide if they want individual platform stacks (like many individual oil rigs) that they transport between or a planned integrated approach that easily works at convenient speed and low transportation costs. Some integrated vendors are warning about "Franken Stacks" and others are warning about "Integrated Average". Either way, watch out for "lock in" and inability to play well with others. This is true when you pick a core that might need future augmentation. Some vendors can support multiple stack approaches like an agile process platform combined with applications and just enough analytic capabilities.
Organizations have to start somewhere and the decision on platforms will likely be quick. Organizations will have to decide is they want best of brand which means you have faith in particular vendor stack / platform or a best of breed approach that combines the best of optimized capabilities. Big decisions are looming right now that could create downstream opportunities or problems.