Blog: James Taylor Subscribe to this blog's RSS feed!

James Taylor

I will use this blog to discuss business challenges and how technologies like analytics, optimization and business rules can meet those challenges.

About the author >

James is the CEO of Decision Management Solutions and works with clients to automate and improve the decisions underpinning their business. James is the leading expert in decision management and a passionate advocate of decisioning technologies business rules, predictive analytics and data mining. James helps companies develop smarter and more agile processes and systems and has more than 20 years of experience developing software and solutions for clients. He has led decision management efforts for leading companies in insurance, banking, health management and telecommunications. James is a regular keynote speaker and trainer and he wrote Smart (Enough) Systems (Prentice Hall, 2007) with Neil Raden. James is a faculty member of the International Institute for Analytics.

December 2008 Archives

Copyright © 2008 James Taylor. Visit the original article at First Look - Sonetto Retail.

I got a briefing on IVIS Group’s Sonetto product this week. Sonetto is most famous as the platform for Tesco’s multi-channel strategy. Sonetto is a product designed to focus on multi-channel operations, especially (but not exclusively) multi-channel retail. They make the very valid point that multi-channel retail is complex. Not only do retailers have multiple channels (stores, web, catalog, mobile, phone) they also have multiple brands, multiple store formats and partner channels (affiliates, amazon etc). This front-end complexity is matched by a complex back-end with multiple data providers, suppliers and fulfillers. To manage this environment they see a set of application areas that enable the acquisition and retention of customers and the development of customer communities:

  • Multi-channel information management
  • Multi-channel infrastructure
  • Supply chain Logistics and Operations
  • Customer Experience
  • Sales and Marketing Channels

    While all these are important, to IVIS multi-channel information management is at the heart.

    Sonetto Retail is a suite of business applications designed for business users so they can manage product, pricing and customer information without any reliance on IT. The suite manages the dynamic information requirements of a multi-channel retail and allows the business to develop a centralized process for collaboration. Business users “teach” Sonetto by example and define rules and triggers that are automatically applied.

    Sonetto Retail uses a ripple-down approach to business rules (defining rules by example and creating rules to handle exceptions as those exceptions are found) to manage a set of core decisions on the product side:

      • Standardize
        Take information as it arrives and add/standardize information attributes. Products often have inconsistent descriptions and lots of data hidden in text fields, for instance, and the rules extract this information.
      • Categorize
        In a multi-channel world most products have multiple categories (wide screen TV, LCD TV, Home Theater etc) and this is rarely supported by the source data. Rules are created to correctly categorize products.
      • Price and Promote
        Pricing rules and rules to describe offers and promotions are created. The promotion rules are based on a wide range of templates such as bundle pricing, buy-one-get-one-free or cheapest item free.
      • Merchandise
        Defining rules for responding to stock levels, clearance sales etc
      • Publish
        When products are pushed to specific channels, additional rules might be applied to decide which channels and in what circumstances.

        The platform consists of 4 layers:

        • Underlying everything is a decision repository with the usual versioning, audit trail, rollback to old versions, see what was relevant when features
        • The Sonetto Enterprise Decision Management Platform (love the name) with a Semantic Search engine, a Business Rules engine, a Categorization engine and a Workflow engine.
        • Built on this platform are applications for Product Information Management(PIM), Price and Promotion Management(PPM), Channel Information Management, Community & Lifestyle Management (e.g. the recipe community at Tesco).
        • Finally a common environment manages all these applications and supports data management, data viewing and transactions (commerce engine)

          Sonetto expects to run in a complex IT environment and supports integration with ERP systems, Content Management Systems, Search engines, eCommerce platforms, Service Desk systems and more. The product is all web-based (thin client) and has a clean, easy to use look and feel.

          I was a little surprised by the lack of analytics. The product is very rule-centric and all the rules must be entered directly. While IVIS has customers who use analytic tools to mine their data for rules, these rules must then be created and managed by business users in the product. Clearly some more integrated analytics and some ability to execute analytic models would be useful, though Tesco has made itself into a case study among case studies using Sonetto so perhaps this matters less than I think it does!

          The business value of all this comes from increasing sales, reducing costs and improving customer relationships (nothing surprising there) and the product offers particular gains in agility (through the use of rules managed by the business), enhanced customer experience (better search and categorization, cross-channel consistency) and increased productivity. Sonetto’s value add comes in three areas:

          • A Business Focus
            They offer a (fairly) natural language interface and focus on making it easy to add rules incrementally plus they allow simple definition of workflow
          • Flexible and Agile
            They have no fixed product definition or hierarchy allowing companies to develop a complex network of categories and they support a very open environment.
          • Productivity
            Through rules, dynamic forms and a focus on avoiding IT delays.

          All in all an interesting product for retailers that clearly works - as I said, Tesco is a poster-child for multi-channel retailing these days. It is not currently sold in the US and their focus is on expanding in Asia/Pacific and Europe.


          Posted December 29, 2008 3:09 PM
          Permalink | No Comments |
          Copyright © 2008 James Taylor. Visit the original article at Merry Christmas.

          That’s it for me - off to the mountains this week. More posts in the new year. Meanwhile don’t forget to change your feed over to feeds.feedburner.com/jtonedm. Happy Holidays.


          Posted December 24, 2008 3:55 PM
          Permalink | No Comments |
          Copyright © 2008 James Taylor. Visit the original article at All Predictive Models Are Wrong - So What?.

          I was struck by an article on Insurance and Technology titled All Predictive Models Are Wrong. Now, while this is a true statement (more or less) it reminds me of a famous Winston Churchil quote:

          It has been said that democracy is the worst form of government except all the others that have been tried.

          In other words I think that predictive models are wrong but they are better than other approaches. They are less prone to racism, sexism and other isms. They can take into account many more factors than a human decision maker. They can and do improve decision making.

          So don’t be put off - predictive analytics can help but, like any technology, you need to remember their limitations.


          Posted December 22, 2008 3:41 PM
          Permalink | No Comments |
          Copyright © 2008 James Taylor. Visit the original article at Have you changed your feed yet?.

          Just a quick reminder that if you want to keep reading the blog in 2009 you will need to change your feed from feeds.feedburner.com/smartenoughsystems to feeds.feedburner.com/jtonedm. For now it is automatically redirecting but that will only last a couple more weeks.


          Posted December 20, 2008 5:47 PM
          Permalink | No Comments |
          Copyright © 2008 James Taylor. Visit the original article at Predictive Analytics are important no matter what IBM thinks.

          Doug Henschen had a blog post on IBM today that caught my eye - Will IBM Add Analytics to its Toolbelt? in which he quoted Ambuj Goyal (who heads up information management at IBM) as saying predictive analytics are overrated. Sadly this reminded me of the old days of IBM - when FUD (fear, uncertainty and doubt) was IBM’s reponse to anything they did not do well. Predictive analytics are not overrated, at least not by anyone who understands them. It is true that predictive analytics, like all good technologies, are sometimes overused by over-enthusiastic supporters and that they can’t do everything. IBM’s lack of this technology is a mistake as without it their solution set is incomplete and no amount of FUD will change that.

          Now Ambuj is right about a couple of things:

          • He’s right about the need for a trusted data layer - bad or mistrusted data does not enable good decision making.
          • He is also right when he says, though not quite in so many words, that decision making takes more than data mining and predictive analytics - it takes optimization, business rules, adaptive control. This is why enterprise decision management takes all these components together. Where he is wrong is in not including predictive analytics and data mining (technology he does not have) as peers to the technologies he does have (or will have once IBM completes its ILOG acquisition).
          • He is also right about the power of vertical solutions and vertical templates to accelerate time to value and address issues of skills shortages.

          To argue, however, as he does that Cognos is a platform for advanced, automated decisioning is disingenuous. Like any BI platform, Cognos is not the place for automating decision making as it is human-centric - designed purely to present information to a human decision-maker. It is aimed at decision support not decision management.

          In the article Ambuy says:

          Predictive modeling has become a hammer for these vendors, and in many cases it’s overkill. For example, we were recently talking to a client that has created a 100-terabyte model, using software from one of those vendors, to do prediction. They’re using an amazing amount of computing capacity because the only method that they had available was a statistical model coming from one of these vendors. But the problem could have been solved much more simply with a policy engine, which could have been created for less than $1 million rather than requiring tens of millions of dollars of expense.

          Are they very good at constrained logic programming or Markov decision processes or rules engines? No, they are not.

          Well of course this is why I talk about decision management not just analytics. The important thing is to focus on the decision and then figure out how to solve it - hence decision management.

          Ann Milley of SAS commented also and said Hey, IBM, It’s About Competing on Analytics! Ann makes some great points:

          • She emphasizes “continuous learning and improvement” or what I call adaptive control
          • She reiterates that this is about “more than prediction. It’s all about delivering business value”

          It is, in fact, not about competing on analytics, but about Competing on Decisions (to use a great phrase that Neil Raden came up with and on which we presented recently).Analytics matters a great deal to companies competing on decisions, which is where IBM is wrong, but it does take more than analytics, which is where SAS is wrong. Business rules, optimization, data mining, predictive analytics and adaptive control - the necessary ingredients for Enterprise Decision Management and business success.


          Posted December 19, 2008 9:17 PM
          Permalink | No Comments |
          PREV 1 2 3 4 5 6 7 8 9