Implementing an Innovation Process

I came across this nice blog on Innovation Process Framework, by Jeffrey Philips (via Innovation Weblog)

innovationprocesssoftware.gif

The blog is a nice read and tries to outline a framework for Repeatable Innovation. Towards the end Jeffrey appeals to the readers to provide feedback.

If you care to, please comment or provide your feedback. I think if we practitioners, consultants and interested bystanders can create a consistent vision for the future of innovation and the tools and processes necessary for success, we can help our clients and business partners become more successful.

I have been experimenting with a few tools and some ad-hoc processes for innovation (in small product groups). So let me start out with a few tools and see how we can start putting together, elements of this framework.

You can start with any simple content management system (Drupal, Plone, Dotnet Nuke or even a Wikimedia engine).  It is also possible to use commercial portal products like Sharepoint, BEA or IBM portal servers. Let us see how we can go about building a prototype of the tools required to bootstrap your Innovation Process based on the framwork described by Jeffrey.

1. Trend Spotting

You can use several products that exist in the marketplace to track trends. The tools I list here provide you information to detect trends. Here is a list.

  • Google Alerts- A service to receive alerts based on certain keywords
  • InfoMinder – Our product to track specific web pages for changes (you can optionally specify filters) and receive notification. Unlike Google or alerts, InfoMinder is specific to the pages you want to track.
  • DiggdeliciousTechmemereddit or any of your favorite social bookmarking service (you can look for specific trends or retrieve information using tags)
  • Technorati or Google Blog Search tools
  • Tag Clouds (many of the services mentioned above provide tag clouds that tell you the more popular trends) or you can create your own tag clouds.
  • Google Trends – A product from Google that allows you to see trends based on searches
  • A set of high level Text Mining and Tech mining tools ( a subject that deserves almost a blog of its own)

A combination of these services and other customer serivces, can be used to perform trend capture. You need to figure out a way to make sense of trends from these different pieces of information (Trend Spotting). Fortunately many of these tools provide RSS streams or APIs. You can easily integrate them with several content management systems.

2. Generate Ideas

You can set up a workflow where people with the role of Generators, look at the captured trend information, combine it with other sources and generate ideas. These can be either stored in any relational database like MySQL, Postgres SQL.

3. Capture additional Information
In the system, Ideas are just a specific type of document with certain metadata like creator, date of creation, source of idea, description etc. It will be nice to add the capability for anyone to tag ideas. Based on tags and other criteria, ideas can be routed to Evaluators.

4. Evaluate Ideas
The evaluators can add comments, additional tags, classify the ideas to be further researched and send them back into the system. With each iteration, the circle widens. Ideas are further validated, combined with others or split into multiple ideas and put back into the system. Since Ideas trigger ideas, this process of combining and splitting will work well.

5. Develop and Launch

Stakeholders are found, prototypes built, ideas developed and launched as products/services.Your content management system can be used as a record keeper in this phase. In every step of the process from ideation to launch, it may be worth engaging small communities of users. Connecting to social tools like Twitter, Facebook, LinkedIn may be a good way to build and grow these communities.

6. Workflow/Process Automation

This is functionality built into several content management systems. Ideas can move from one stage to another (nascent, researched, validated etc.)

7. Idea Archetypes

One of the important aspects of the design of Idea Archetype is the progressive addition of information. Some ideas are listed here:

  • State – specifies the current stage of the idea. As it goes through the system, the state of the idea keeps changing
  • Strength – an indicator of the strength of the idea. As ideas float through the system and gather support, the strength can be progressively increased. Support increases this value and opposition decreases this value.
  • Next Steps – For each idea there can be a sequence of steps which can be started by the creator of the idea and collaboratively edited by others. For example, the legal department may add a patent search as a next step

8. Process Maps

Argument mapsConcept maps and other mapping tools can be loosely integrated (most of them export data in XML, JSON or CSV  format).

9. IdeaLogs

Ideas can also be published in blogs (private if they are meant for a small internal groups). Many portal products or content mangement systems come with their own blog software. You can also integrate some of the popular blogging software like WordPress.

10. Wikis as Collaborative Knowledge Bases

Wikis can be used as a knowledge bases to share, collaboratively edit and archive ideas. Wikis are alternative to idea archetypes,  mentioned earlier. Many of the wikis now provide templates for creating structured pages.

Any portal framework that supports content management, custom content types, workflow, collaboration, authentication can be used to jump start the Innovation Process in an organization. It is easy to bootstrap an innovation process using this framework and existing tools in a few weeks.

The best approach is to start with something as simple as a portal, set up some simple workflows, use a single page with extensible metadata as a basis for collaboration.

Update

Pretty much everything I described here can be done using many other portal frameworks, as well. One of recent favorites is Drupal especially since it has started providing support for RDF ( core language  for the semantic web as well). You can also custom build this framework using web frameworks like Rails(Ruby), Django(Python).

Leave a Reply

Your email address will not be published. Required fields are marked *