A New Decision Engine: Hunch, and Guided Analysis for the Enterprise

Timo Elliott
Posted by Timo Elliott

guided-analysis-banner

A new “Decision Engine” called Hunch was launched today. How might this type of technology relate to collaborative decision-making in enterprises?

Hunch’s tagline is “Hunch helps you make decisions and gets smarter the more you use it.” From the web site:

What is Hunch? Hunch is a new way to help people make all kinds of decisions, such as:

  • Where should I go on vacation?
  • What’s the best US college for me?
  • What kind of smartphone is right for me?
  • Which museum should I visit in the Netherlands?
  • What blog should I read?

Results are based on the collective knowledge of Hunch’s users. Hunch already has more than 2,500 possible topics, and Hunch users add new topics every day.

Hunch asks you 10 questions or less, some about what you need or want and some about who you are. Like a human expert, Hunch chooses questions intelligently and adapts them depending on how you answer.

After asking you questions, Hunch offers you a result. You can leave feedback on Hunch’s results and also explore other results based on the input of Hunch users.

You can train Hunch or correct a fact it got wrong. This is how Hunch learns and gets smarter. Topics get smarter as more people train them. So new topics may start …

guided-analysis-banner

A new “Decision Engine” called Hunch was launched today. How might this type of technology relate to collaborative decision-making in enterprises?

Hunch’s tagline is “Hunch helps you make decisions and gets smarter the more you use it.” From the web site:

What is Hunch? Hunch is a new way to help people make all kinds of decisions, such as:

  • Where should I go on vacation?
  • What’s the best US college for me?
  • What kind of smartphone is right for me?
  • Which museum should I visit in the Netherlands?
  • What blog should I read?

Results are based on the collective knowledge of Hunch’s users. Hunch already has more than 2,500 possible topics, and Hunch users add new topics every day.

Hunch asks you 10 questions or less, some about what you need or want and some about who you are. Like a human expert, Hunch chooses questions intelligently and adapts them depending on how you answer.

After asking you questions, Hunch offers you a result. You can leave feedback on Hunch’s results and also explore other results based on the input of Hunch users.

You can train Hunch or correct a fact it got wrong. This is how Hunch learns and gets smarter. Topics get smarter as more people train them. So new topics may start out not so smart, but they’ll get smarter over time.

Users contribute to Hunch in various ways:

  • Don’t like the questions that Hunch asked you? Submit a different question.
  • Don’t see the result you wanted on Hunch? Propose a better result.
  • Looking to make a decision that isn’t on Hunch? Add a new topic.

See the rest of the tour

It certainly seems like useful technology – and might be even more useful in a corporate context. How many times do different people need to make similar decisions, but end up doing it in completely different ways, because there’s no common context or methodology? How much could performance be improved if everybody made decisions as well as the best person (or better)?

Guided Analysis

The enterprise equivalent of this technology has been typically been called “guided analysis”: helping business users undertake analytic tasks by stepping them through a series of actions, and (without the web 2.0 angle) it’s been available in various guises in existing BI solution portfolios.

For example, the BusinessObjects Analytic Applications from a few years ago included ‘storylines” that helped analysts move from one analytic to another:

image

And the applications included workflow tools for predefined lists of analytic steps (for example, this was used by Somerfield supermarkets in the UK to ensure that each business analyst looked at the right analytics when reviewing which products to stock).

image

This technology was then incorporated into the BusinessObjects Encyclopedia part of the standard SAP BusinessObjects platform functionality. Administrators could set up Business Questions and suggest reports and measures that might be useful in answering the questions:

image

And set up “storylines” for the Guided Analysis tab of the Encyclopedia.

image

These storylines are then available to anybody using the portal.

image

Sadly, anecdotal feedback suggests that awareness and usage of this functionality has been fairly low, in part because of the maintenance requirements, but also because a lack of business skills within the IT organization, or because there was simply no agreed best practice to be recorded.

Looking forward, this type of functionality (along with workflow, discussion threads, rating, etc.) is becoming an essential part of an information platform layer designed to provide a consistent set of workflow and guidance across all information sources and activities (e.g. BI, planning and budgeting, collaboration around information management, etc.)

As the industry matures, we’re moving from BI efficiency (helping people access information quickly) to BI effectiveness (making sure that people are analyzing the right things in the right way) and effective guided analysis is an important part of this process.

Human nature means that it’s very hard to spend the time necessary to put together good best practice workflow. Allowing business people not only to collaborate around data, but also collaborate around processes for best practice decision-making will clearly be a big part of the future of “BI 2.0” and business user applications.

[Post to Twitter] Was this interesting? Share with others on Twitter with automatic URL shortening!