You Can Have A More Accurate Forecast

I think we are going to continue our new years theme and focus on the possibilities. I am pretty sure you didn’t toast in the New Year last Saturday night with jaded apathy. Far likelier you thought of the vast accomplishments on the horizon in 2012 for your personal life, your health and your career.  Like new hires and those newly promoted – your intent is to hit your metrics, earn your bonus, set yourself up for future success and have the best year yet.

This optimism stands in stark juxtaposition to what we hear from prospective clients when we tell them we can lower their forecast error by 25%. We touched on this skepticism in previous blog post discussing now being time to raise the bar for demand planning and forecasting outcomes. It very well might be the high rate of IT implementation failures alone that create disbelief but their reasons (dare I say, excuses?) for not moving forward, span from a company process that doesn’t align with individual goals for their position to the simple fact that there are only 24 hours in a day and 36 hours of work. Having dealt with the realities of their roles they seem willing to simply meet the minimum requirements.  In these conversations they initially cast aside what they consider to be “impossible” because of a limited perspective of what could be.  And that is bad for them and that is bad for their company.

Have you stopped believing in POSSIBILITY and settled for mediocrity?

You would never purport that someone who has resolved to lose 10lbs in the New Year would be successful without a scale because you know it’s a critical component of measuring progress towards their goal. You also know that weight is not the only component for good health and because muscle weighs more than fat, you could actually gain weight while improving overall health.  You need the right tools in to set yourself up for success in a complicated environment characterized by huge amounts of data, fast moving market dynamics and a changing customer environment.   Accordingly, without specific improvements, you cannot expect that your ERP will lower your inventory or increase perfect order performance. Nor will traditional statistics based modeling drive a significant, measurable decrease in forecast error as effectively as modeling based on global optimization. Run your antiquated forecasting tool as many times as you want but old fashioned, static and linear models that everyone else uses and incorrect forecasting models will most likely leave you with the same inaccurate and profit reducing, customer infuriating  forecasts.

There are better tools than the one you are using today that will get you to your 2012 goals.  You can have an accurate forecast. You can blow your metrics out of the water. You can get out of the office earlier because you had more accurate data and made smart and informed decisions.  When a prospective client accepts that a 25% reduction in forecast error is possible based solely on our proprietary Interactive Neural Computing, the doors open to the opportunities they can create for themselves.  Our clients achieve 3-8% improvement in pre-tax profitability because they believe in the possibilities.

More about our forecasting and planning software.

Process Redesign – Do it right or go home.

When approached without a clear understanding of the technological possibilities and a relevant examination of the organizational structure and policies (i.e pay)  supporting it, Business Process Redesign is a complete waste of time.

In fact, I will go so far as to argue it is a destruction of shareholder value to invest time, human resources and actual cash in pursuit of improvements that will be minimized and/or never realized.

So what are the risks of the so-called “People, Process, Technology” approach?

  1. You can only contribute your process requirements based on what you know, but you can’t know the possibilities inherent in technology platforms until you’ve explored them
  2. You might design a process that only stays relevant for a year rather than 5 years and/or necessitates expensive upgrades, training and consulting fees with every step
  3. You might end up paying people to perform the old process, not the new process
  4. Your technology may require costly customizations or “work-arounds” to manipulate data outside of system support in order to support the new process


Say your current process involved 4 or 5 steps – it begins with a query to get historical data, extracting that data to a forecasting engine – something like excel or Demantra (no real difference), then running the model, then organizing the output into a useable format, running a series of reality meetings with Sales, Marketing & Operations and then finalizing the execution forecast.

Now, given that background and experience, in a typical process redesign, you would enter a room with big whiteboards or brown paper sheeting taped up and most likely a consultant (internal or external) standing at the front of the room saying – please – give me your requirements so we can design a process.  What are you going to pull from in order to provide requirements?  Your experience – what you know.  Which means in effect you will be paving over cow paths – which might be charming in Verona but not the basis for delivering competitive differentiation for your company.

Perhaps even more frustrating is say that you are able to think out of the box and come up with a truly radical process that cuts out 4 of the 6 steps and if properly executed would increase accuracy by 10% and reduce cycle time by 75% and help improve order fill rates from 90 to 99%?  And then you go out and look for a technical platform only to find there is no technology to support your process? Can you say frustrating loud enough?

So what is the right way?

Technology – Process – People

A comprehensive, holistic approach based on the principle that Technology should support Business Process, and Business Process should exploit the capabilities Technology can provide – Davenport & Short dubbed this recursive view of Technology & Process Redesign “the New Industrial Engineering” — Rather than lay out a step by step detailed process (proscriptive process design) you outline specific outcomes for the project (outcome based process design). 

This would include:

  1. Identifying what is wrong with the current process,
  2. Creating a general vision as to where you think the current process could improve,
  3. Setting specific measurable goals for improvement – including areas to focus investment and amount or degree of improvement desired.
  4. Considering your pathway toward maturity – how much will it cost to improve down the line?


Within this framework, you can then invite technology providers in for conversations and focus on finding a partner with a technology that can provide measurable performance improvements as well as a platform that is flexible so that it can stay relevant for this process as well as future required changes as your business continues to grow and evolve.

Once you have a partner chosen based on technological prowess, flexibility, industry knowledge and compatibility, you can then engage in detailed process design in tight partnership with the technical platform you have chosen. You can then also do a review of compensation structures and organizational design to ensure these will be flexible in supporting new process and performance expectations.

[Quick word of warning:  This does fly in the face of the normal RFP process where a company says, “Hey, we want to do something but we are not going to share the specific details or outcomes we are going to measure nor any of our criteria for success.” What ensues within the responders  is a process of guesswork, misdirection, outrageous claims and leverage which, in many cases isn’t entirely dissimilar from a season of “Survivor.” And yet, after it all, many executives end up choosing based on faulty assumptions about long-term cost savings and NOT business outcomes.]

There has been specific research done on this by a few groups and the statistics are frightening:

70% of process redesign projects fail to deliver on the business case and the budget. Only 30% actually hit the minimum marks!!

The holistic approach, however, produced strikingly different results.

  1. 50% reduction in total project time,
  2. 35% reduction in total project cost,
  3. 70% improvement in technology uptake
  4. 60% improvement in attaining business case


This really seems like a no-brainer but yet, as noted at the start, people are still approaching corporate performance improvement like it’s 1985.

How can we evolve this conversation beyond old paradigms? What can be done to help drive efforts to improve corporate performance such that our companies not only survive but thrive?

More about our forecasting and planning software.

Industry Trends – Beer Distribution and Improving Profit Performance

Beer Distribution is an interesting business: High margin, protected by regulation that has traditionally limited most forms of competition, which leads to an overall lack of incentive to innovate technologically.

Nevertheless, despite the lack of innovation incentives there are some activities occurring that signal the status quo may be changing a little bit; for example, the recent foray of Berkshire tossed into the mix through the purchase of a couple of distributors.

If the current dynamic were to change, for whatever reason, forecasting would be one area that would allow distributors to rapidly improve – even advance – their bottom line performance outcomes.  Currently, on average, there is not a lot of focus on forecasting. Basic practices involve sales people “working” their on- and off-premise customers, while the inventory people make sure they keep enough stock on hand to ensure customer order fulfillment is met. Inventory managers look for opportunities to take advantage of strategically ordering from suppliers that game prices increases, etc.

A heightened focus on improving forecasting and ordering would allow distributors to lower working capital invested in inventory, while maintaining and/or improving customer service.

Customer service could improve in a number of ways; better order fulfillment being the most basic upgrade. On the more advanced side of the equation; distributors could work together with bars and liquor stores to make sure the products stocked, or on offer, respond and adapt to seasonal changes, trends, pricing, promotions, and holidays – making the distributor a value-added supplier.

In turn, the end merchant will become an even more valued customer by providing a more accurate forecast to their suppliers. This helps distributors and their supply chains become more efficient. Ultimately, this virtuous cycle helps set the distributor apart as a better supply partner – making it one that beer manufacturers will want to work with and which has the capacity to make a product successful in a new market.  This allows the distributor to negotiate more favorable terms with suppliers, thereby increasing margin performance. Everyone benefits.

Improving the forecast model will require improvements in technology and process systems – something that owners will have to support. Since distribution sales people are singularly focused on driving volume and taking care of their customers, they do not take kindly to activities such as supply chain forecasting. But their input is critical in order to achieve a “big picture” point of view that will help the entire company. When forecasting is tied directly to how it will help sales people earn more money (working for a higher margin distributor), a critical component of improving forecasting will be realized.

How improved execution-level forecasting delivers layers of value

Business cases, ROI, improved performance – all are nifty catch phrases used to justify investments in people, equipment and software. In our Nov. 18 post, we highlighted the importance of defining what “execution-level” means in your business’ particular case. But what is the value? Where does the measurable improvement come from?

The obvious answer is: in inventory and/or inventory turns and their associated impact on working capital. I think everyone will agree that an improvement in forecasting will improve the amount of inventory carried.

But this doesn’t begin to fully describe the size of the opportunity. Improved execution-level forecasting is the gift that keeps on giving.

More stuff that customers want
An improvement in execution-level forecasting empowers you to improve the makeup of that inventory, thereby increasing customer satisfaction. In addition, it creates the chance to address issues such as perishability, which has significant performance implications in the food and beverage industry; and old or slow-moving inventory, which can greatly impact the cost line of industries like high-tech and fashion. Issues like these almost bankrupted Cisco back in the early 2000s.

Greater production efficiency
Taken further, improvements in execution-level forecasting also allow companies to address production efficiencies by optimizing the mix and duration of each run for manufacturers, as well as purchase orders for distributors. Reduced changeovers, lead times and overtime are other ways that improved forecasting pays off.

Optimize shelf space for revenue
You can also drill down another level to the management of your bill of materials (BOM) and work in progress (WIP). A perfect production plan is useless if the materials or subassemblies are not where they’re needed, when they’re needed.

If your company deals with big-box retailers and their service level agreements, improved execution-level forecasting gives you the best chance to generate the most revenue possible from available shelf space without running the risk of a large end-of-season return or write-off.

And this does not even begin to explore the opportunities for improving the forecasting and demand planning process and the associated efficiency improvements, or the improvement potential in the area of effective marketing spend.

These results are based on a history of obtaining 1, 2 or maybe 5 percent improvement in forecasting — often at levels higher than execution level, such as brand family or category. Imagine the opportunities if the execution level improved by 25 percent or more.

Demand planning vs. forecasting vs. management: an agreement on terms (Pt. 3)

There are a number of terms that are used when discussing demand forecasting, demand planning and demand management — all of which appear to be used interchangeably. Just what are we talking about here? I thought it would make sense to share the terms I use on this blog and at Demand Foresight,  and  articulate the difference between each.

In my previous two posts, I delineated what demand planning and demand forecasting mean to our industry. Lastly, there’s demand management.  We’re hearing this term a lot these days as it relates to the current economic environment. In economics, demand management is defined as the art or science of controlling demand to avoid a recession. For our purposes, demand management is the art and/or science of matching product supply to demand. If a huge snowstorm is forecasted to hit the Northeast, do the Home Depot stores in the region have ample supplies of snow shovels?  It is within demand management that total demand (open orders plus forecast) is matched against capacity to ensure an efficient and profitable approach. Getting this right requires cross-functional senior leadership as this is where revenue and margin meet.

It is important for an organization to understand the distinctions between demand planning, demand forecasting and demand management so that it can determine where it needs to focus in order to achieve substantial and measureable improvements in bottom-line performance.