supply chain insights

Welcome in the Big Data Opportunity

by Lora Cecere on July 9, 2013 · 0 comments

Big Data Outlook

Big Data Initiative

The term big data is moving up the charts as a “hot topic.” This week, we published our second quantitative study on the evolution of big data concepts, it gave us the opportunity to talk to supply chain leaders on the evolution of technologies and the use of analytics in the Race for Supply Chain 2020.  We wanted to understand where companies are at in the adoption of big data concepts and we had a good time writing the report on the research.

The study was completed by over 120 respondents in the period of June-July, 2013 and the complete study can be accessed on the Supply Chain Insights website or through slideshare.

What Did We Learn?

We believe that the adoption of new concepts for big data is a step change for supply chain teams. It is not about force-fitting new forms of data into applications based on relational databases. Is cannot be treated as an evolution.

It requires change management. It is about small and iterative projects using new forms of analytics. The projects have to be based on a business problem and the focus needs to be on continuous learning. This is quite different from the traditional waterfall project approach of mapping “as is” and ”to be” states and managing a large project against a goal. Companies have to be open to the outcome and invest in innovation through analytics.

To drive success companies have to sidestep the hype. While the powerpoints on big data concepts abound, very few technology companies and consulting partners have built solutions to harness this opportunity, and even fewer supply chain leaders are ready to have the discussion. We find the work by Aster Data (now a division of Teradata) and the work by IBM and Enterra Solutions on cognitive learning engines to be promising. We are also encouraged by SAS’s work on unstructured text mining, Bazaarvoice on the translaltion of blog and sentiment data, and the work by APT on test-and-learn strategies is exciting. We are also encouraged by the work on cold chain and serialization by a number of consultants working on sensor data and counterfeiting.

The Findings

In the study, big data was defined as volume that is at least a petabyte, working with a variety of data that goes beyond traditional structured data, and building processes that are based on an increased velocity of data that is associated with real-time flows. In the results, we see several trends:

-The biggest opportunity is not with the volume or velocity of data. Instead, it is with the management of opportunity associated with new forms of data. In the study,  76% of companies see big data as an opportunity and 12% see it as a big data problem.

-Databases are growing, but they can be managed. 15% of companies have a database today that is at least at a petabyte. The largest databases are not Enterprise Resource Planning (ERP). Instead they are in the areas of product management or channel data.

-The work is starting: 28% of companies have a Big Data Initiative today with 37% planning to implement a big data team.

-Companies that have worked on a data-driven culture have a leg-up. Organizations that have active teams on Master Data Management are more likely to have a big data cross-functional team. 54% of companies with Big Data initiatives believe that big data techniques help with MDM.

In the words of a supply chain leader, the path forward requires a disruption. It is for this reason that we have asked experts from the Department of Defense to join us to discuss these findings on our webinar this Thursday at 1:00 PM EDT. We hope to see you there.

We will also be continuing the discussion at our Supply Chain Insights Global Summit in Scottsdale, AZ on September 11th and 12th. There will be several panel discussions on the future of analytics, the evolution of new technologies for healthcare, and the evolution of digital manufacturing with real-time data.  The event will sell out at 150 participants and the cutoff date for the discounted room block at the Phoenician is August 8th. So register soon to reserve your seat. (The event is now sold-out for technology providers.)

Bait and Switch

by Lora Cecere on April 17, 2013 · 2 comments

Bait and switch:  A form of fraud.  A dishonest marketing tactic where a marketer advertises a very attractive value proposition and then switches the offer to something else after gaining interest by the buyer.

Source Merriam-Webster Dictionary

DDSN. DDSC. DDVN. CDSN. The acronyms keep coming…. The cadence does not stop. Everyone seems to have a new one. Today, they swirl in the market forming a fog. The term demand driven has become vogue again, but what does it really mean? And, should it be taken one step further to orchestrate bidirectionally market-to-market in market-driven value networks?  Or will companies stumble on the path by mistakenly implementing supply-centric processes and calling them demand-driven initiatives?

There are a number of newly anointed experts writing articles about becoming demand driven.  They are piling up on my desk.  As a writer of research on demand-driven supply chains for over eight years, I find many amusing.  I like the idea that this old concept is gaining new steam; but unfortunately, too few people writing the articles really understand the concepts. Instead, I see a behavior that I call bait and switch. The article is written and the story is spun, but the solution offered is a supply-centric solution based on yesterday’s technology.  The original principles of a value network that can sense, shape and translate demand with near-zero latency are being lost in the fog.

 Why is this happening? The market for large ERP programs is slowing. The gravy train is coming to an end. User satisfaction with planning systems is low.  The market shift is towards analytics, but this new market is confusing. It is still early.

Supply chain leaders feel stuck. Their current technologies are inadequate. They are struggling to manage the challenges of  simultaneously driving growth, improving profitability, absorbing complexity and reducing cycles. Frustration is mounting.  The concepts surrounding demand driven sound right.  Companies are interested.  As a result, articles are written proclaiming demand-driven results and then the reader is given a solution that is anything but demand driven.  Each time that they are published, the Shaman sighs and chuckles in her little apartment in Baltimore.

 The first definition of Demand-driven Supply Chains was pushed into the market by AMR Research (now part of the Gartner Group) in 2004. What the articles that flood the market do not tell you about is:

  • Slow Adoption. Eight years after the evolution of the concept, there are only a few companies making progress on demand-driven concepts. If asked, I would only cast a vote for the demand driven work that is happening at Cisco Systems, General Mills, Pfizer, PepsiCo, Procter & Gamble, and Kimberly Clark. Each of these pioneers would tell you that it is hard work. No one company—technology provider or supply chain line of business leader—has figured it out. Most have implemented the concepts in parts of their businesses. The most successful have used best-of-breed solutions. (We will show how the adoption of these practices have improved market capitalization in our webinar on April 25th.  Join us for the launch of the Supply Chain Index.)
  • Hard Work. Many companies that have started demand-driven initiatives have abandoned them.  The rewards are high, but the cultural barriers are difficult.  They are sometimes insurmountable.
  • Misunderstood.  A frequent reason for failure is a lack of understanding of the basic concepts of demand latency, sensing, shaping and translation.  As a result, many well-intentioned companies have mislabeled supply-centric initiatives as demand driven.
  • Demand-driven Concepts Are Not an Evolution. They are step change requiring either the redeployment of existing technologies or the purchase of new platforms.  Details matter. Data model structures are the difference between success and failure. Today’s architectures are inside-out not outside-in, and to be demand driven, the process focus needs to change. This often means a reimplementation of APS, and a change in focus for the company.
  • Be Careful of the Word “Integrated.”  The promise of the integrated supply chain sounds attractive, but tight integration of the supply chain has reduced agility and made the supply chain response less flexible. Today, due to tight integration, only 10% of companies are satisfied with their “what-if” modeling capabilities, and only 23% can model supply chain profitability. Both are essential.  The goal should be synchronized demand and supply with role-based dashboards, workbenches and optimization engines that allow users to work across the supply chain.  To accomplish this, demand has to be sensed, shaped and translated.
  • Change Management Issues Are High.  The largest challenges are in the redefinition of process flows from inside-out to outside-in. Demand-driven concepts are expansive they extend from the customer’s customer to the supplier’s supplier, but the areas of sales and procurement are often very resistant to the demand-driven concepts. To do this companies need an end-to-end leader.  Only 1% of companies have defined this role.
  • Most Have Defined “Demand” too Narrowly.  Demand in the demand-driven network is about much, much more than forecasting.
  • It Needs to Be About More than Demand.  Supply is volatile. Shortages abound. It is for this reason that I have defined market-driven value network processes in the book Bricks Matter.  The definition is: An adaptive network focused on a value-based outcome that senses and translates market changes (buy- and sell-side markets) bidirectionally with near real-time data latency to align sell, deliver, make and sourcing operations.

As we move forward, there are no silver bullets. There are no well-defined industry platforms.  I coach companies to take the following steps.

  • List All the Forms of Demand Data and Map Its Usage. This includes unstructured text data (this can include data from social networks, ratings and reviews from blogs and websites, and channel data), weather data, and transactional data.  Some supply chains also have inputs from the evolving world of the Internet of Things where machine sensors transmit frequent streams of data. This is the case for heavy equipment, vending machines in the field, and smart shelves.
  • Map the Process Outside-in from the Channel Back. Start with the channel, and map the requirements of the channel. Evaluate how to reduce latency by using downstream data to sense demand and implementing demand translation technologies to make the downstream data usable.  These technologies include the work by Terra Technology and ToolsGroup.  (While SAP has purchased SmartOps and is marketing a demand sensing/demand translation offering, I have not been able to validate the solution through references.  It is clear that math matters.  Neither Oracle or JDA references were able to meet the challenges in the field.)
  • Build What-if Analytics.  Technologies like Kinaxis and Steelwedge are frequently undervalued for supply chain visualization and what-if analytics. Cloud-based analytics for sourcing and the management of supplier networks are evolving and should be embraced. Consider solutions from GHX, Elemica, E2Open and SCA Technologies to improve end-t0-end visibility.
  • Design the Network. Actively design the network with clear push/pull boundaries and right size buffers.  The strongest solutions in the market continue to be Llamasoft, Insights and JDA.  And, the strongest consulting partner for network design is Chainalytics.  I also like the work that is happening at the Demand-driven Institute on the redesign of manufacturing to be more demand driven.
  • Focus on End-to-End Orchestration. Build processes that enable the alignment between demand- and market-shaping levers to orchestrate end-to-end bidirectionally through outside-in horizontal processes. Actively orchestrate demand through shaping, and the supply response through the market-driven levers below.  Charter the end-to-end process manager to orchestrate a market-driven value network that connects and orchestrates bidirectionally between markets.
  • Use New Forms of Data. Embrace Digital. Think long-term on the use of digital signals.  Map the use of mobile/social and eCommerce on the future of the digital path to purchase, and the impact of machine-to-machine interfaces in manufacturing on digital manufacturing.  It excites me to see the revitalization of manufacturing applications to be more demand driven based on the Internet of Things in process industries and 3D printing in the discrete industries.
  • Experiment with Best-of-breed Technologies. This innovation is not going to come from the large players. It will require large manufacturers to take risks with smaller players like Applied Predictive Technologies, Enterra Solutions, Orchestro, Retail Solutions, and Signal Demand.
  • There Is No Substitute for Leadership.  Success happens when there is an inspired leader that believes that the supply chain needs to own the supply chain from the consumer/user to the supplier’s supplier.
  • Focus on Building Horizontal Processes.  These bridge the gaps between functions. The four main horizontal processes to tackle are revenue management, sales and operations planning, supplier development, and corporate social responsibility.

 In summary, progress on supply chain cycles and margins, and balancing the trade-offs of complexity, has stalled. Over the last decade, the only metric that we have improved is revenue/employee (see below).  Leaders do not know what to do to power themselves off of this horizontal plateau. The gap between what we have and what we need has widened.

Processes are evolving. Technologies are changing. There is no clear definition of what drives value. In an effort to try to drive progress, the system integrators and technology providers have started providing their own research. The problem is that it is not OBJECTIVE, and lacks research rigor. It is largely self-serving and is confusing the market.

As the research firms have consolidated, primary research is lacking. Consortia research has not filled the void. While these organizations have the reach, the organizations of APICS, CSCMP, GMA and SCOR lack the understanding of research processes.

I want to help.  It is for this reason, that I have built this new company, Supply Chain Insights. I believe that supply chain matters. I can see the impact on balance sheets through the successful implementation of demand-driven concepts; and it’s even greater when the concepts are balanced by market-driven levers. I look forward to sharing these with you in our Supply Chain Index webinar.  While it does not provide all of the answers, I look forward to sharing what we are seeing. We want to stir a healthy debate in the market, and would love to have you join us.

One of the things that you will never find us doing is promoting a bait-and-switch program. Our goal is to help supply chain leaders gain first mover advantage.