What happened to CPFR?

Lora Cecere: Supply Chain Shaman

Go to any supply chain conference, and you will hear it.  Yes, the term collaboration is bandied about. It is over-used and often over-hyped in discussions largely without meaning.  So, what does it mean?  And, what happened to the supply chain collaboration initiatives of the 1990s?

Let’s start with the definition.  The greatest success in supply chain relationships is when true collaboration happens.  What does it look like? It is a when a sustainable win/win value proposition.   Six elements are required:  resources, skills, joint vision, leadership, a plan and aligned incentives.

The problem is that the so-called “collaborative programs” of the 1990s focused solely on process missing the mark on these six elements.  The tenants of VMI and CPFR were well-intended, but they fell short in building true collaborative relationships.  Let’s take a closer look.

Why did CPFR not gain wider adoption?

Many tout success, and many conference presentations expound on benefits; but back home at the office, the teams are confounded.  In the late 1990’s it was all the rage.  Yes, CPFR (Collaborative Planning Forecasting and Replenishment), over-hyped by many, has fallen short in delivering the promise

The results are clear.  After ten years of active projects, collaborative planning forecasting and replenishment failed to reach its promise for three reasons:

  • Laborious. Just too much work for the benefit.  The added costs did not measure up to the benefit and the programs were not grounded in the six essential elements of collaboration.  Instead, it was a process implemented in the absence of the core elements of what drives collaborative relationships.
  • Retail forecasts not up to the task. For CPFR to work, retail forecast accuracy needs to be high and with sufficient granularity to ensure analysis.  The dirty little secret with CPFR is that only three retailer forecasts—Best Buy, Food Lion and Wal-Mart—were up to the task.  In addition, the gap in retailer data for perpetual inventories and accurate on-hand data could not give the teams a good starting point.
  • Lack of integration into Enterprise Systems. For most Advanced Planning System (APS)/Enterprise Resource Planning (ERP) deployments, there was no logical connection for the data.  As a result, it failed to make a systemic impact on supply chain excellence.

So, as a result, most CPFR initiatives became 20-year old pilot projects.  They were isolated—lacking integration into corporate demand planning architectures—and only as effective as the strength of the relationship and the quality of available data.

When does it make sense?

However, let’s not throw the baby out with the bathwater.  It would be incorrect to say that CPFR never makes sense. It was over-hyped and over-promised, and applied to situations where there was not a good fit.  So, you might be saying, where does it fit?   When a company has these five stars to align, CPFR can be used to reap great benefit:

  • Significant channel presence. The account needs to be significant—at least 10% of the channel-for the investment to warrant the expense.  The greater the channel presence, the greater potential benefit. It must matter and make a difference.
  • High Demand Volatility: CPFR makes more sense for products with short life cycles, seasonal patterns, strong dependence on weather, and in competitive categories.  It makes less of an impact for products that have stable demand.  Companies benefit from advance warning signals.
  • Strong Retail Partnership. The data is clean, available and meaningful to both parties’ business objectives.  Both companies have strong planning skills and a passion for forecast accuracy.  It is tied and closely coupled to the business.
  • The Tie to Replenishment can make a Difference: Many companies forget the “R” in CPFR.  If he advanced notification from forecast sensing can make a difference in improving replenishment and the other conditions can be satisfied, go for it!  However, not all replenishment cycles can be shifted in concert with the CPFR signals
  • Right Stuff: It makes sense when there is demand architecture to support close coupling of the demand signal.  The architecture must allow integration at the account, ship-to level. 

What now?

The key is to be judicious.  CPFR has a place in driving supply chain excellence; just not the over-hyped promise of ten years ago.  Be more judicious.  Be more realistic.  Make smarter decisions.

See Lora Speak at IBF’s
First Ever:

IBF's Supply Chain Forecasting & Planning Conference: Asia

Supply Chain Forecasting & Planning Conference: Asia

5 Responses to What happened to CPFR?

  1. I agree completely with your comments regarding CPFR.
    Like most initiatives, and/or new thinking (concepts) people take them too literal. They are too black/white in their thinking and implementation approach. They attempt to apply the process across all their products, and if it does work, then they abandon the process completely.
    Instead of segmenting their product portfolio and applying different processes, methods, and skills to each segment they apply one process and methodology across all products. Companies need to focus on products that have high growth and profit potential (usually 20% or less of their product portfolio). Utilize data, domain knowledge (collaboration), skills, and more sophisticated technology to facilitate and improve their ability to predict and replenish those growth products more efficiently. Usually, those same growth products have the most data and information, but also have highly volatile demand (due to competitive activities and market dynamics) requiring the most collaboration and analytics.
    Most technology today can automatically forecast demand for the remainder of a company’s product portfolio due to the stability of demand. Those same products are normally harvest brands that are mature and have very little sales/marketing focus. However, they do have strong trends and seasonality. With the advancement in technology harvest brands can be forecasted rather easily, and can be automated on a large scale. The majority of most companies product portfolios consist of harvest brands. Those brands are the anchors that sustain the company’s revenue base that fund and support the growth brands.
    I agree that collaboration is a key success factor in any demand forecasting process whether it be CPFR, or S&OP. However, you need two key ingredients to enable collaboration, 1) trust among the participants, and 2) trust in the analytics that support collaboration. There are two levels of trust. First there needs to be trust between the CPG Manufacturer and the Retail, but there also needs to be trust among the internal participants in the demand forecasting and consensus processes. The best way to gain that trust is by aligning initiatives and performance rewards. Second, everyone in the process must trust the analytics and supporting technology because without the analytics you will make the wrong decisions, and without the technology you cannot support the process across an entire product portfolio. Most companies focus far too much on the process, and shy away from the analytics blaming the technology for the demise of the process. As a result, with the absence of analytics and technology many companies fall back to Excel and human judgment both of which are not scalable.
    The good news, the data, analytics and technology have all caught-up with the process.

  2. I strongly agreed with this article. It seems to be that CPFR requires good software system otherwise it’s difficult to manage too many SKU involved. Relationship between manufacturer and retailer is also another issue.

  3. When I did a deep look at CPFR several years ago, it included a number of pre-defined, rigid, processes. These reminded me of Rosetta Net – defined by a committee to cover all possible cases. The result is a system requiring so much data, and so complex, that it becomes extremely difficult to implement, confusing to users, and ultimately falls out of favor.

    On the other hand, I know of several major electronics companies who have implemented the CPFR principles: namely capture forecast (their own to suppliers or from their major customers), review changes from the previous cycle, firm up on a common view of the demand.

    Then, if getting demand from customers, they test/simulate the capability of their supply chain to satisfy that demand, resulting in a committment back to their customers (which could result in an adjustment to the demand the customer plans to place on them).

    When communicating with suppliers, the objective is to obtain a committed supply chain back from the supplier, which is then used to feed committments back up the chain.

    The point is, not only CAN it be done, but that it IS being done with EXCELLENT results. However, the process has been simplified from formal CPFR into something that can be implemented and used.

  4. I think if you work in non-retail industry, you may not need software to help with CPFR. Because, the benefit of joint business planning already is already good for CPFR implementation.

  5. currently I start to write my undergraduate thesis. my topic is related to applying CPFR in FMCG company. those reasons in this article are make sense to judge CPFR cannot work perfectly. so which strategy will make sense?

Leave a Reply

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

WordPress Anti Spam by WP-SpamShield