In June, Supply Chain Digest published a groundbreaking report on the evolution of integrated supply chain planning and execution. Here's a quote from the synopsis by Don Bowersox and Nick LaHowchic (authors of Start Pulling Your Chain!). Readers of Flowcasting the Retail Supply Chain may find this theme familiar:
"As Nick LaHowchic (former supply chain executive at The Limited Brands) and Dr. Don Bowersox of Michigan State University recently wrote, if information "was shared fluidly between participating firms in a channel, then a great deal of "anticipation" would be replaced with facts. In a collaborative environment, it would not be necessary to forecast what others are planning to do or what they are planning to buy." This is the critical point: visibility and information sharing will allow trading partners to simply sense and respond, within defined relationship rules."
Click here to read the excerpt. A link to the full report is available in the first paragraph.
"It takes years to build up trust, and only seconds to destroy it." - Anonymous
Everybody hates forecasting.
Who wants to be accountable for predicting the future? There are no kudos for getting it right (or at least "right enough"), but there are opinions aplenty when you miss. With the benefit of hindsight, everybody knows exactly what you SHOULD have predicted to avoid the glut of inventory or lost sales that the mistake caused.
Unfortunately, forecasting can't be avoided. In order for any supply chain to satisfy demand, then that demand must somehow be anticipated in advance.
Wouldn't it be grand if a forecasting system could spit out predictions that are always accurate? We put men on the moon, so surely we can develop a mathematical model that can accurately predict demand, right?
If you're one of the people who happen to hold this view, you're in luck. Virtually every software company with a forecasting module will tell you that they have just the right proprietary algorithm that will allow you to feed in some historical data, then kick back and relax while your inventory drops and your service levels soar. It's a Utopian dream - the forecasts are accurate and nobody needs to be accountable.
On more occasions than we would like to recall, we have been told by software companies that their algorithms are so advanced that no user intervention is required at all. When we ask questions about how easy it is for a user (who may not have a Ph.D. in quantam mechanics) to understand and interact with, the most common response is incredulity.
"These algorithms are extremely sophisticated. The last thing you want is someone tinkering with the results! If you feed in the right inputs, it will produce accurate forecasts without being poisoned by human error."
At this stage, we'd just like to point out that this is NOT a diatribe against software companies. Sophisticated forecasting systems can and do produce good results most of the time and if someone is going to override the forecast, it should be for a damn good reason. However, we all need to realize that people will not trust something that they don't understand.
No forecasting algorithms will ever be perfect and accountability for the results can't be assigned to the software. The first time somebody gets burned trusting the system generated forecast (without knowing WHY it failed), they will never trust it again. And if they can't understand the model or influence its behaviour, they will turn to Excel spreadsheets (which they DO understand) and cut orders manually, thereby reducing the value of the forecasting system to virtually zero.
When people buy cars, they don't just look under the hood to see which has the most impressive engine. If that were the case, we would all be driving Porsche 911s and nobody would own minivans.
Similarly, you can't buy a forecasting package because it has the most sophisticated mathematical engine or the highest number of input variables.
Forecasting is not magic. It's a process that must be managed by people and aided by computers. Over time, executing this process should help people to better understand the demand for the products they manage - not delegate this responsibility to some all-knowing computer code sitting on a server in the data centre.
If you give people understanding and control over the forecasts they need to manage, the accountability will follow.
"With one breath, with one flow
You will know
Synchronicity" - The Police
Synchronization. It's by far the most overused term in the supply chain lexicon. For some reason, everybody wants their supply chains to be "synchronized".
Taken literally, the expectation seems to be that all supply chain activity should happen at the same time. This explains why so many retailers seem to be gung-ho about sharing "up-to-the-minute" POS and inventory data with upstream supply chain partners.
The problem with this thinking is that it tries to circumvent the natural order of things in the supply chain. Products that are being sold from store shelves today were shipped a couple of days ago, put into stock at a distribution centre a couple of weeks ago and manufactured a couple of months ago. So much for synchronization.
Until we can find a way to manufacture, distribute, ship and merchandise inventory all just moments before it is demanded at the store shelf, we must accept the fact that true synchronization is not economically achievable for most supply chains.
So if trying to achieve synchronization is the business equivalent of banging one's head against a brick wall, then what can be done to make the retail supply chain more responsive?
The answer may lie in the second most overused supply chain term: integration. From our experience, it seems that many companies today interpret the word integration (and, for that matter, synchronization) as meaning "to send a lot of information all over the place".
Recent attempts at achieving integration in the supply chain are a lot like aerobatic flying teams. Every maneuver requires a series of precise steps to be performed in sequence, with information being shared among the pilots in the form of radio signals. Through years of planning and practice, it may look as though they are integrated, but in the end, it falls to each individual pilot to interpret what's going on from his own unique vantage point. If something unplanned occurs with one aircraft, there's no automatic adjustment that happens in the others to compensate. As exciting as this is, it's also very dangerous.
In many ways, information sharing strategies like CPFR are like this. Information is shared and guidelines are adopted, but in the end it requires a great deal of effort and relies on a number of different parties to make the same interpretation of events to be successful.
Supply chains should be boring. Unexpected things may happen, but the consequences of them should be as easy to predict as tomorrow's sunrise. When your supply chain is boring, then you know that you've achieved true integration.
When designing how a new process (like Flowcasting) is going to work within your organization, don't you hate being the "next guy". You know, the guy (or gal) who's got to make something work or integrate to something where the previous person didn't care at all about the next guy - er, I mean you. You should design and implement as if you were the next guy.
"Whatever you fear most has no power - it is your fear that has the power." - Oprah Winfrey
When you think of how big the retail supply chain is, it can be pretty scary. A moderate to large sized retailer can hundreds to thousands of stores and tens of thousands of items to fill them. That means millions and millions of opportunities to satisfy your customers - or piss them off.
As we've worked to help retailers adopt Flowcasting over the last decade, we keep running into the same fears. And it's not that people are afraid of doing new things to improve their operation. The problem with retail is that even small things need to be done on a large scale to make a difference.
In our experience, here are the top 3 fears that any retailer needs to overcome in order to adopt Flowcasting.
Fear #1 - Inventory Record Accuracy
There's no magic number for what's an "acceptable" level of inventory accuracy, but we recommend that retailers put processes in place at store level that can achieve 95% accurate inventory within tolerance at the item/store level. And for most items that are scanned at receiving and scanned out the cash, those tolerances should be pretty tight.
The typical response to this is something like: "What?!? Last time we did a physical count, our inventory by item/store was in the high 50s - nowhere near 95%! Do you know how much time and money it will take to bring it up to a 95% standard? That's impossible!"
Without a doubt, getting to a high level of inventory record accuracy by item/store can be a daunting task (particularly when you think of it as a single challenge and don't break it down into bite sized chunks).
However, if you already know that your store inventory numbers can't be trusted (assuming you even have them at all!), then that means you have no idea how much fully costed inventory you have or where it's deployed in your network.
Shouldn't that scare you a lot more?
Fear #2 - Store Level Planning
To make Flowcasting work, you need to make the customer the driver of your entire planning process. To satisfy a customer's needs, you first have to anticipate those needs. Since customers want to go to a specific store to get a specific product at a specific time, that means you need to forecast by item/store. And to line up the entire retail supply chain behind the customer, you need to forecast EVERY item in EVERY store - for 52 weeks into the future.
The usual response to this is a hard gulping sound, followed by: "Are you kidding me? Do you know how many forecasts that adds up to? Who is going to manage all that? Plus the accuracy will be abysmal when you start looking at sales velocity by item/store - we can hardly produce accurate forecasts at item/DC level right now!"
While it's true that such a huge volume of information can't possibly be managed by people on a daily basis, today's planning systems can do most of the day in, day out grunt work and only flag situations that are out of tolerance for user review. As for pure forecast accuracy metrics that measure forecast versus actual for a week, the results will surely be less than stellar for slow moving items at store level. However, because of all the rounding influences that take place before a product gets to the shelf and the fact that the supply plan can be resynchronized daily, high levels of pure accuracy are not really necessary.
The fact is that if you don't have a formalized forecasting and replenishment process in place, the success of your entire supply chain rests with whatever forecast was in somebody's head when they placed an order for more inventory at the store.
Shouldn't that scare you a lot more?
Fear #3 - Sharing Information
The Flowcasting process has no organizational delineation. Its purpose is to plan and manage the entire retail supply chain - as a single entity - from the factory to the store. The simple fact is that, in the majority of cases, multiple organizations are needed to make that happen. This means that retailers will need to share with their trading partners (particularly vendors) much of the same information that is required to plan internally - namely forecasts and replenishment plans by item/location for a 52 week planning horizon.
In spite of the progress that has been made with respect to retailer/vendor collaboration over the last decade or so, there are still many retailers who think that cards must be played close to the chest in order to keep vendors in line. "We can't share our replenishment plans with vendors! If that information gets in the hands of our competitors, we're sunk!"
The simple fact is that information sharing needs to occur between retailers and vendors around important events like promotions and line changeovers to ensure that they are prepared to handle them. No matter how secretive you want to be, a minimum level of trust is absolutely required to do business with any vendor.
And if you're not sharing up-to-date time-phased replenishment plans that give the vendor enough of a planning horizon to make changes to their production plans, then they're just sitting there in the dark guessing and hoping that they made a good guess when your order arrives.
Shouldn't that scare you more?
The easiest thing to do is to continue using the same planning processes you currently have - maybe spruce them up with some system enhancements or fancy reports - then you won't have to face your fears.
But the complacency approach assumes that your competitors are likewise standing still and not proactively tackling these same issues in their operations (ask Kmart, Sainsbury and Blockbuster how well that worked for them).
Doesn't that scare you more?
If you haven't read Pink's other books, you should. His landmark book, "A Whole New Mind" (AWNM) is one of the 5 most thought provoking books I've ever read. And I read a lot. Like 50 or so books a year since about 1982.
Drive is a brilliant follow-up to AWNM. In it, Pink dispels a number of myths about what motivates us professional supply chain planners. For managers, especially folks who are doing project work, it's especially enlightening. For example, ask most folks what's the most important thing to motivate a knowledge worker and virtually everyone would say rewards. Most folks would be wrong.
To find out the surprising truth, grab a copy of the book or check Dan out at www.danpink.com
I recently came across a couple of fantastic eBooks published by Adexa. They are very insightful, candid and in complete alignment with our experience designing and implementing planning processes and systems in retail.
Use the links below to download them for free from Adexa:
Common Pitfalls in Supply Chain System Implementations
When you have symptoms, you go to the doctor, who runs tests based on the symptoms. Sometimes the tests tell the doctor what he needs to know and sometimes they're inconclusive. When that happens, it's time to order a magnetic resonance imaging (MRI) scan.
Unlike an X-ray, this revolutionary medical technology allows doctors to actually see inside a patient and watch organ function, brain activity and blood flow on a computer monitor - in real time.
Like the human body, the retail supply chain is composed of many interconnected and highly dependent moving parts and systems.
And when things go wrong, the symptoms can be acute. But the cause is often less obvious. In those instances, it sure would be handy to have an MRI machine that works on supply chains instead of patients.
Luckily, we now have Flowcasting.
If the symptom is a glut of store level inventory for certain highly promoted items, maybe our promotions are more effective in some stores than they are in others. By planning inventory item by item and store by store, we can pinpoint which stores have inventory left over from the last ad before we send more product to them on the next ad.
If the symptom is inventory imbalances and stockouts during line changeovers, maybe we lack proper visibility to the impact of planogram resets or changes to store ordering rules. By incorporating item/store specific planogram information into our supply chain plans, we can model the exact impact display changes will have - weeks in advance and across the entire supply chain.
If the symptom is chronic issues with on shelf availability, then perhaps it's time to think about ALL products at ALL locations in the supply chain as being on a daily planning cycle. What's the point of having weekly or monthly order reviews when any customer can walk into any store and buy any product on any day? By refreshing our plans for every single product/location in the extended supply chain on a daily basis, we don't even give ourselves the chance to run out of stock because we don't have any order review "blind spots".
The retail supply chain is NOT a black box. It is a series inventory pools and product flows that can easily be visualized and modeled within a system.
In the same way that an MRI image is a digital representation of the tissues and systems within the human body, Flowcasting is a digital representation of the physical stocks and flows in the extended retail supply chain.
Like the X-ray, Collaborative Planning, Forecasting and Replenishment (CPFR) has started us on the road to truly fixing the retail supply chain. But now technology has evolved to the point where we can make it to the end game.
Let's finish the journey, shall we?
Click here to read an excellent article by Andre Martin and Darryl Landvater, describing how Flowcasting is changing relationships between retailers and their suppliers.
