Flowcasting the…

Posted on April 6, 2016 by Jeff Harrop

What’s in a name? That which we call a rose by any other name would smell as sweet. – William Shakespeare, Romeo and Juliet

It’s hard to believe that nearly 10 years have gone by since Flowcasting the Retail Supply Chain was first published. Mike, Andre and I had the manuscript nearly complete before we turned our attention to figuring out the title. I figured it would end up being something boring like ‘Retail Resource Planning’.

Then, Andre got us both on a conference call to tell us that he came up with a title for the book: ‘Flowcasting’. For reasons I can no longer remember (or maybe perhaps because it wasn’t my idea), I immediately disliked it. We went back and forth on it for awhile and as time went on, the name ‘Flowcasting’ grew on me.

Then I came up with the idea that we should be more descriptive about the process. It’s a new concept for managing the retail supply chain, right? So, we should call the book ‘Flowcasting the Retail Supply Chain’!

Now I wish I had just listened to Andre and left well enough alone.

Over the years, we’ve written and talked at length about the types of major changes a process like Flowcasting brings about in the supply chain areas for retailers and their trading partners:

Managing all supply chain resources (inventory, labour, capacities and spend) using a single set of numbers based on a forecast of consumer demand at the shelf
Collapsing lead times between retailers and their trading partners by unlocking the power of the Supplier Schedule
Improving supply chain operational performance by providing 52 weeks of future visibility from the store shelf back to the factory and giving manufacturers the opportunity to shift their operations from ‘make to stock’ to ‘make to order’
And so on and so forth

To be sure, the supply chain operational and planning changes are significant – but that’s just the beginning. To make Flowcasting successful, the mindset changes in other areas of the retail organization can be just as revolutionary.

The Merchandising Organization

Historically, the buyers in a retail organization were (and in most cases still are) just that: ‘people who buy stuff’. Their primary accountability is growing sales in their categories, and the conventional wisdom is that the best way to increase sales is to have a lot of inventory. If you think you’re going to sell 10,000 units on an ad, then you buy 20,000.

Flowcasting turns this notion on its head. Because it is always accounting for all inventories in the supply chain and replanning every day based on the sales forecast, ‘Buyers’ must learn to become ‘Category Managers’ who focus on their key accountability: generating demand and providing input to the sales forecast. The change management implications of this change cannot be overstated, as this can be viewed as taking away their control of a key input to their overall success.

Similarly, buyers are accountable for maximizing gross margin on their lines. When negotiating case packs and ordering minimums, they may be inclined to choose the option that gets them the lowest overall cost per unit. But this can be very costly to the business overall if these constraints make it impossible to flow product to the store shelf, particularly for lower sales volume stores. High gross margin doesn’t necessarily equate to high profitability for the business as a whole if folks in the merchandising organization haven’t been given the tools and accountability to think holistically about the supply chain.

Store Operations

Two decades ago, the retail supply chain was distribution centres and trucks. A few years ago, the thinking began to evolve to include the retail store as part of the supply chain. Now, we are finally starting to think of the supply chain as linking the factory to the customer’s hands.

Most retailers measure store ‘in stock’, but what’s meaningful to the customer is on shelf availability (meaning that if a store has 10 units in inventory, but it’s stuck in the back room somewhere, it’s an ‘in stock success’, but a failure to the customer).

One of the biggest questions retailers face is ‘How much autonomy do we give to the store?’ While it’s true that each individual store is closer to the market they serve than the folks at home office, does this mean that giving every store manager a stock ordering screen to use at their discretion will automatically increase customer service? In my experience, the only things it increases with regularity is inventory and confusion.

Even though Flowcasting would generally be a centrally managed process for most retailers, it’s driven from individual store level sales patterns and constraints that may not be known to a store manager. The entire process works much better when people are accountable and measured only for the things that are within their control. For a retail store, that means two things:

Receive the product quickly and get it straight to the shelf
Keep the inventory records accurate

If stores are focusing their energy in these areas, it doesn’t guarantee perfect on shelf availability for the customer, but it makes it easier to trace back where the process is failing and make corrections when there are fewer fingers in the pie.

Human Resources

Flowcasting isn’t just a different calculation for coming up with an order recommendation. It’s a fundamentally different way of operating a retail business. As such, it requires a different skill set and mindset to manage it.

While it might be tempting to fill an open replenishment position with someone with replenishment experience at another organization, you could be trying to fit a square peg in a round hole if her prior experience is with traditional reorder point or push methods. In fact, you may have to invest more time with ‘untraining’ than you do with training.

To be successful with Flowcasting, a person needs to be organized and possess decent problem solving skills. Other than that, no specific ‘experience’ is required – in point of fact, it could actually prove to be a detriment.

First

Posted on March 2, 2016 by Mike Doherty

Normally these newsletters focus on wisdom we’ve gleaned with respect to supply chain planning, specifically as it relates to Flowcasting. This month’s is going to be a little different. It’s time for a celebration and to recognize a first in retail!

We’d like to congratulate our client, Princess Auto Limited (PAL), on their successful rollout and implementation of the Flowcasting process and philosophy.

For those of you who don’t know PAL, they are a growing Canadian hard goods retailer with stores from coast to coast, supported by a multi-tier distribution network that flows product from all over the world into their consumer’s hands.

I remember getting a call from one of the Vice Presidents, Tammy and she said, “Hey I read the Flowcasting book you guys wrote and I really like the idea and concept”. So, I replied, “Great Tammy, I really like it too!”. And so it began.

Just recently they completed the initial implementation and are completely managing the flow of product using the Flowcasting approach and solution.

As originators of the concept, co-authors of the book and retail focused implementers, we could not be more proud of their efforts and achievements. Here’s what they have accomplished:

They are managing the entire supply chain from a forecast of consumer demand, by item, by store (and web store). The forecasting process and solution is elegant, simple and intuitive and is not fraught with unnecessary complication.

The consumer demand forecast accounts for a number of different selling situations and demand patterns, including:

Regular selling products that sell all year round
Seasonal and highly seasonal products
Products on promotions
Slow sellers and very slow sellers
One time buy products that are purchased and sold during a very short time period or when the inventory is available

As an organization when it comes to demand planning they have shifted their thinking and everyone, including the Demand Planners and the Merchandisers, is speaking the same language – “what we think we will sell”.

They use the consumer demand forecast to calculate a series of integrated, time-phased plans (for a 52-week planning horizon) from the store to the supplier factory adhering, like heroes, to the mantra “never forecast what you can calculate”.

The projections of product purchases are shared with their merchandise vendors in the form of a supplier schedule so the vendors have visibility to see future requirements and plan accordingly. The vendors are beginning to use these projections to plan raw materials and production and are adhering to the concept of “silence is approval” – that is, if they see something in their schedule that looks odd, they contact their respective Analyst – otherwise, they are expected to supply.

Product transfers (from Stores to Distribution Centres) and purchase orders (from Vendors to Distribution Centres) are cut, automatically, at the agreed upon lead time between any two locations. Since all partners in the supply chain have visibility they are working to a single lead time between two nodes in the supply chain – even promotional requirements are automatically converted to an order at the same lead time as regular demand. In fact, their thinking has evolved to the point where they understand that, in retail, there really is no difference between a “regular” order and a “promotional order”.

The unit projections at all levels are automatically translated to different languages of the business:

In dollars for finance to aid in budgeting and gaining control of the business
In cube and weight for distribution, transportation and retail operations to provide volume projections and automatically convert the projections to capacity requirements

In terms of planning they are using the Flowcasting process and solution to greatly simplify and improve a number of common retail planning scenarios. These include:

Product introductions
Product discontinuations that phase out products, store by store, based on available inventory and the store specific consumer demand forecast
New store openings to predict future dated sales and replenishment by product
Promotions, including national and regional events
Seasonal planning to ensure that residual seasonal carryover inventory is minimized
Distribution Centre openings whereby future store requirements are mapped to the new DC in order to properly depict the ramp up demand on the new DC, while simultaneously showing the draw down demand on the older DC

To summarize:

They are managing their business to a single set of numbers and have created a dynamic model of the business – driven by the consumer!

I think you’d agree that’s pretty impressive and, in my opinion, they deserve a round of applause and perhaps even a standing ovation for their accomplishments.

Folks, this is a first in retail supply chain planning and a first for Flowcasting. A retailer is managing their entire business using Flowcasting and is already beginning to see improvements in on-shelf availability, inventory performance and operational performance – not to mention starting to gain control of the business by connecting the business plan with the day-to-day operational plans.

To the folks at PAL and to the small, dedicated Flowcasting Team that we’ve worked with, I have only this to say….

Congratulations, and well done!!

Overly Sophistimicated

Posted on February 3, 2016 by Jeff Harrop

There are many methods for predicting the future. For example, you can read horoscopes, tea leaves, tarot cards or crystal balls. Collectively, these are known as ‘nutty methods’. Or you can put well researched facts into sophisticated computer models, more commonly known as ‘a complete waste of time.’ – Scott Adams

If you have your driver’s license, you can get into virtually any automobile in any country in the world and drive it. Not only that, but you can drive any car made between 1908 and today.

You want to make a left turn? Rotate the steering wheel counter-clockwise.
Right turn? Clockwise.
Speed up? Press your foot down on the accelerator pedal.
Slow down? Remove your foot from the accelerator pedal.
Come to a stop? Press your foot down on the brake pedal.

Think all of the advances in automotive technology – from the Ford Model T in 1908 to the Tesla Model S in 2016… Over 100 years and countless technological leaps, yet the ‘user interface’ has remained the same (and universally applied) for all this time.

This is what makes the skill of driving easy to learn and transferable from one car to the next. And all of the complexities of road design, elevation and traffic can be solved by making the decisions on the part of the driver in any scenario very simple: speed up, slow down, stop or turn. Heck, even the lunar rover used the same user interface to deal with extraterrestrial terrain!

Not only that, but because the interface is simple and control on the part of the driver is absolute, there is built in accountability for the result. If the car is travelling faster than the speed limit, it’s because the driver made it so, manufacturing defects (most often caused by ‘over sophistimication’) notwithstanding.

While supply chain forecasting software hasn’t been around since the early 1900s, it’s been around long enough that it doesn’t seem unreasonable to expect some level of uniformity in the user interface by now.

Yet, while a semi-experienced driver can walk up to an Avis counter and be off cruising in a car model that they’ve never driven before within minutes, it would take weeks (if not months) for an experienced forecaster to become proficient in a software tool that they’ve never used before.

The difference, in my opinion, is that the automobile was designed from the start to be used by any person. Advanced degrees in chemistry, physics and engineering are needed to build a car, not operate it.

While no one expects that ‘any person’ can be a professional forecaster, it should not be necessary (nor is it economically feasible) for every person accountable for predicting demand to have a PhD in statistics to understand how to operate a forecasting system. The less understood the methods are for calculating forecasts, the easier it is for people on the front line of the process to avail themselves of accountability for the results. Police hand out speeding tickets to drivers, not passengers.

Obviously, not all cars are alike. They compete on features, gadgets, styling, horsepower and price. But whatever new gizmos car manufacturers dream up, they can’t escape the simple, intuitive user interface that has been in place for over 100 years.

While I’m sure it’s an enriching intellectual exercise to fill pages with clouds of Greek symbols in the quest to develop the most sophisticated forecasting algorithm, wouldn’t it be nice if managing a demand forecast was as easy as driving a car?

Bad Habits, Part 2

Posted on December 2, 2015 by Jeff Harrop

Managing on time performance (inbound or outbound) is a struggle for every retailer. Whenever there’s a delivery failure, somebody asks the inevitable question: “How do we prevent this from happening again?”

Sometimes the cause is known to be out of your control – a freak snowstorm leaves trucks stranded or a temporary congestion issue at a facility. For these types of unpredictable reasons, it’s simply not possible to achieve 100% on-time delivery all the time.

But what about cases where a supplying location is chronically at 80%? What if it’s dozens of locations with this problem (which is not uncommon in retail)?

What we’ve seen is that, faced with this problem, many retailers have developed the bad habit of arbitrarily increasing their planning lead times. In many cases, they will even develop processes that will analyze the demonstrated order-to-delivery times on historical orders and transfers, then automatically update the planning systems with increased lead times in an effort to boost on time delivery performance.

What often happens is that on time performance does improve (for a time), which tends to validate the process. Then results start to slip again, triggering another wave of analysis or an increase in the frequency of the automated logic to ‘stay on top of things’.

The problem with this approach is that it assumes two things:

That lead times are something that ‘just happens’ over which nobody has any control.
Where on time delivery performance is failing, it must be because people aren’t being given enough time to do perform their tasks.

The actual order-to-delivery cycle time in a supply chain is the result of processes. These processes must be routine, repeatable and, most importantly, designed to achieve the goal you’re measuring them against.

The routine and repeatable part likely isn’t the issue in most cases. The amount of time it takes to pick an order or drive a fixed distance doesn’t have a lot of variation (especially when lead times are rounded to the nearest day). Once you know what those standard times are, there really shouldn’t be any need to change them very frequently (unless the processes that generate the results change significantly).

Yes, there are rare, infrequent events that will cause lateness, but that should account for an on-time performance measure in the 80s.

More likely than not, the culprit is that the processes are not designed to achieve on time performance. For example, if there is a prioritization scheme in place that schedules picking and shipping based on any other criteria than the due date, the process is not designed for on time performance.

For example, it’s commonplace for retailer DCs to prioritize promotional shipments to stores with a shipping date of next week ahead of regular shipments that are due to be shipped today.

Suppliers may or may not be prioritizing shipments based on the size of the customer or the negotiated price.

In any of those cases the issue is that the ship date is not being respected, so no amount of additional lead time will solve the problem long term. All an increased lead time will do is lengthen the amount of time between the order date and the scheduled ship date, thereby decreasing the ability to adapt and react to changes. Not to mention the increase in safety stock requirements to cover a longer frozen window.

Like anything in the supply chain, the key to solving chronic issues with on time delivery is to find the root cause of the problem and take the necessary steps to address them. To the extent that the processes are within your organization’s control, that can be relatively straightforward (assuming the intestinal fortitude exists to prioritize all shipments based on a due date, promotional or not).

If the issue is with a supplier, there needs to be a common understanding of how their underlying processes work and what conditions are necessary to deliver consistently on time. This goes beyond using a scorecard to try to ‘shame them into submission’. By adopting Flowcasting and sharing a time-phased schedule of their requirements, retailers can provide very valuable preparatory information to their suppliers that will give them visibility to shipping requirements weeks and months before the purchase order is ever cut.

Don’t fall into the bad habit of setting your lead times based on a data mining exercise. Look at them with a critical eye, identify where chronic failures are occurring and attack the root cause.

Bad Habits, Part 1

Posted on October 7, 2015 by Jeff Harrop

Bad habits are like a comfortable bed, easy to get into, but hard to get out of – Anonymous

Friedrich Nietsche said “Most bad habits are tools to help us through life.”

This is especially true in the world of retail supply chain planning. When retailers embark upon implementing time-phased planning, it is tempting – from a change management standpoint – to tell people “this isn’t much different than what we do today” or “we’re essentially doing the same thing in a different way”.

While this is partially true at the 10,000 foot level, the things that are different are fundamentally different.

This month, we’ll explore a bad planning habit that is particularly insidious, because it is often treated as best practice in a non-Flowcasting world: Using the forecast as a “joystick” to control product flow.

There comes a time in every retailer’s life when product must flow further in advance of customer demand than desired, such as for capacity blowing seasonal peaks, planogram resets or promotional display setups.

In a reorder point world, there is little means to plan these types of things in advance. Replenishment rules such as safety stock apply “right now”, making it exceedingly difficult to do a lot of advance preparation in the system. This leaves 2 options:

Plan all of your safety stock updates in advance and use an Outlook reminder to update the values at just the right time to trigger orders when you want them to arrive.
Spend hours creating manual orders/transfer requests with the future ship dates you want – and pray that things don’t change a lot in the meantime that will require you to change those dates/quantities as it gets closer to execution time.

Neither very good options to be sure.

But wait!

There is a third option. By putting an artificial spike in the forecast at around the time you want product to arrive at various locations, you can ‘trick’ the system into flowing product when you want, but also take advantage of changing inventory levels right up until the lead time fence, thereby letting the ordering happen automatically.

To some extent, this makes sense in a reorder point world. Often, it can be the only efficient and automatic way to control flow. And it doesn’t do too much damage, because the forecast really only exists to trigger an order for an item at a location.

So what would happen if this approach was applied in a Flowcasting world? Not much: just unstable plans, incorrect capacity and financial projections and general chaos within operations and the vendor community.

That’s because, with Flowcasting, the sales forecast doesn’t merely drive ordering for an item at a location. It drives replenishment across all locations in the supply chain simultaneously. It drives capacity planning for the DCs and transportation department. It drives labour planning at the stores. It drives revenue, purchasing and inventory projections within the merchandising teams and finance.

Simply put, the forecast must always be representative of how much you expect to sell, where you expect to sell it and when you expect to sell it. And with the visibility afforded by a reasonable sales forecast, many options become available for controlling the flow of goods as necessary.

In other words, once you make the leap to Flowcasting, the bad habit of using your forecast as a ‘joystick’ to control replenishment is no longer a ‘tool that helps you through life’ – rather, it is one of the quickest possible ways to make your life a living hell.

The Biggest Mistake You’ll Ever Make

Posted on August 5, 2015 by Jeff Harrop

You must learn from the mistakes of others. You can’t possibly make them all yourself. – Sam Levenson

‘How does it work today?’

It’s the first question any self respecting project leader or consultant asks on day 1 of a new initiative. It’s critically important to know the ‘state of the nation’ before you embark on any sort of effort to make the changes needed to improve results.

After asking this question, you set about the task of learning the current practices and procedures and documenting everything to a fairly low level of detail (along with the most common variations on those practices that you’ll almost always find in companies of any size).

In most retail organizations, you’ll hear versions of the following:

For non-promoted periods, our purchasing is mostly automated with a short lead time, but we buy promotions manually with a longer lead time from the vendor.
Our stores ‘pull’ product in non-promoted weeks, but for big events like promotions and shelf resets, we ‘push’ the product out to them .
The system will recommend orders, but our analysts review all of the recommendations beforehand and have the ability to cancel or modify them before they go out to the suppliers.

After a few weeks, you’ve become a pseudo expert on the current state landscape and have everything documented. Now it’s time to turn this retailer from a reactionary firefighter into a consumer driven enterprise.

And that’s when you make the biggest mistake of your life.

After summarizing all the current state practices with bullet points, you save the document with the title ‘Future State Requirements’. You don’t know it at the time, but you’ve just signed yourself up for years of hell with no light at the end of the tunnel. If it’s any consolation, you’re definitely not alone.

Common practices, whether within your organization or throughout your industry as a whole, represent ‘the way everybody does it’. This does not equate with ‘that’s the only way it can be done.’

At the time, it seems like the path of least resistance: We’ll try to do this in such a way that we don’t disrupt people’s current way of doing things, so they will more easily adopt the future state. But if you’re moving from a ‘reactive firefighting’ current state to a ‘demand driven’ future state, then disruption cannot be avoided. You might as well embrace that fact early, rather than spending years of time and millions of over-budget dollars trying to fit a square peg in a round hole.

So how could such a catastrophe be avoided? All you really need to do is dig a little deeper on the ‘common practices’. The practices themselves shouldn’t be considered requirements, but their reason for existing should. For example, look at the three ‘common practices’ I mentioned above:

‘For non-promoted periods, our purchasing is mostly automated with a short lead time, but we buy promotions manually with a longer lead time from the vendor.’

Why does this practice exist? Because, generally speaking, the sales volume (and corresponding order volume from the vendor) for a promoted week can be several times higher than a non-promoted week.

So what’s the actual requirement? The vendor needs visibility to these requirements further in advance to make sure they’re prepared for it. Cutting promotional orders in advance is the currently accepted way of skinning that cat – but it’s by no means the only way.

‘Our stores ‘pull’ product in non-promoted weeks, but for big events like promotions and shelf resets, we ‘push’ the product out to them.’

Why does this practice exist? More often than not, it’s because the current processes and systems currently in place treat ‘regular planning’ and ‘promotion planning’ as distinct and unrelated processes.

What’s the actual requirement? To plan all consumer demand in an integrated fashion, such that the forecast for any item at any location represents what we truly think will sell, inclusive of all known future events such as promotions.

‘The system will recommend orders, but our analysts review all of the recommendations beforehand and have the ability to cancel or modify them before they go out to the suppliers.’

Why does this practice exist? Usually, it’s because the final order out to the supplier is the only ‘control lever’ available to the analyst.

What’s the actual requirement? To be able to exert control over the plan (items, locations, dates and quantities) that leads to order creation, allowing the actual ordering step to be an administrative ‘non event’.

So, once you’ve identified the true requirements, what’s next? A lot of hard work. Just from the 3 examples above, you can see that there are several change management challenges ahead. But at least it’s hard work that you can see and plan for ahead of time.

Woe to those who choose what they think is the easy path at the outset only to suffer the death of a thousand surprises after it’s far too late.

In Praise of Non-Collaboration

Posted on June 3, 2015 by Jeff Harrop

According to dictionary.com, the definition of ‘collaborate’ is: ‘to work, one with another; cooperate’.

Recently, I received an email from Supply Chain Digest asking me to participate in a survey called ‘The State of Retail and Vendor Supply Chain Relations 2015’ – perhaps you received it too. (As a consultant, I never participate in these surveys as I feel it could taint the results, but I’m always interested in what they’re trying to find out).

The Research Summary reads: ‘What is the state of retailer and vendor/supplier relations today? Is it getting better and more collaborative – or heading the other way?’ The implication is crystal clear: More Collaboration = Better.

It seems foolish to argue against that. But that’s just what I’m going to do.

Over the years, supply chain collaboration initiatives were conceived with the notion that ‘two heads are better than one’, especially when there’s uncertainty afoot that could cause havoc in the supply chain, such as for promotions or product launches.

And because trading partners’ planning processes weren’t integrated with a common set of numbers, collaboration was seen as the way to bridge the gap (i.e. ‘we have two sets of plans that are misaligned, so let’s talk our way through it until we agree’). But to what extent is collaboration necessary (or even advisable) when trading partners are fully integrated?

Collaboration is work. Integration is effortless.

All that said, I would tend to agree with the ‘two heads’ argument, so long as all available information is shared between the ‘heads’ This is rarely the case, however, and it’s most often the result of inability to share, rather than unwillingness to share.

Consider a common scenario whereby a retailer and a supplier collaborate on a sales forecast. In order for the demand picture to be complete, the following information must be known and disclosed during the collaboration:

Pricing and promotion strategy for the supplier’s products at the retailer’s stores
Pricing and promotion strategy for the supplier’s competitors’ products at the retailer’s stores
Pricing and promotion strategy for the supplier’s products at the retailer’s competitors’ stores

Point 1 is generally already known by both parties. Points 2 & 3 are at best, unethical (and at worst illegal) for either party to share with the other.

It’s sort of like having two people collaborate on what bet to place on a 5 card poker hand. Each person can only see 3 of the 5 cards with one common known card between them and they aren’t allowed to discuss the cards they can see with their collaboration partner. But the strength of the hand is determined by all 5 cards together.

As a general rule, the retailer faces the customer and they know the breadth of their offerings across all of their competing and complementary products. They may not know exactly what their competitors are doing that could impact their sales, but they pretty much know everything else. Given that the supplier is ethically prevented from providing the one missing piece of information, what value do they add to the collaboration process?

Another example is collaboration on network efficiency to alleviate capacity issues at retail DCs during peak periods. The problem here is that a capacity constraint occurs as a result of a number of straws breaking the camel’s back. Only the retailer has the visibility to all of the straws. How can a collaboration with any single supplier (straw) result in a plan to smooth out the flow for the entire building?

Note that I’m not suggesting that retailers and suppliers shouldn’t be talking to each other when circumstances require it. In the capacity constraint example, the retailer has all the information they need to detect the constraint and figure out the best way to circumvent it. At that point, discussions may need to happen with some suppliers to pull shipments ahead or maybe bypass nodes in the supply chain, but at that point, they’re really just working out execution details rather than ‘collaborating on the plan’.

When you think about it, the supply chain is really not a chain at all – it’s a web. With the exception of private label goods or exclusive supply arrangements, the notion that a retailer and a supplier can ‘act as a single entity in service to the consumer’ is not as easy as it sounds – even with advanced planning processes like Flowcasting. Each retailer offers products from many competing suppliers and each supplier provides their products to many competing retailers.

So if the idea of collaboration is somewhat flawed, then what is better?

Basically, retailers need to get their houses in order and build all of their planning activities around sales at the store shelf (using Flowcasting, of course), incorporating all information they know into the plan – inventories, shipping/receiving schedules, case pack sizes and the like.

Once constructed by the retailer, these plans can be shared directly with suppliers, allowing them to ‘read the retailer’s mind’ without having to second guess or do a lot of ‘collaborating’ back and forth. Discussions between business partners only need to occur when either party foresees difficulty in executing the plan.

Why collaborate when you can integrate?

The Inkblot

Posted on April 1, 2015 by Jeff Harrop

Philosophers have only interpreted the world, in various ways. The point, however, is to change it. – Karl Marx (1818-1883)

What do you see when you look at the image above? A demon skull? An old woman with wild hair? A happening dance party?

That’s the great thing about the Rorschach inkblot test – there’s no wrong answer. Any and all interpretations of an inkblot are “right”.

In many ways, a lot of what have been considered to be key supply chain advances are often treated in the same way. Over the years, I’ve met many people at conferences and seminars who proudly state that “we’re doing CPFR” or “we’re doing S&OP”. After short conversations with these folks, it seems that the definition of “CPFR” and “S&OP” has become somewhat diluted. It seems that any time someone shares a spreadsheet with someone outside their organization, it’s “CPFR” and whenever two people from different areas of an organization are in the same room, it’s “an S&OP meeting”.

Part of this is because these terms have successfully pervaded the supply chain lexicon and are well known. People are either jumping aboard a bandwagon or perhaps unintentionally using the terms as shorthand for something they were never meant to convey.

However, to a certain extent, these concepts sort of had it coming. While the early proponents and practitioners of these concepts laid out some guidelines and principles, there was still a lot of grey area as to how they could be applied to the actual day-to-day operation of a business by the people who actually run the day-to-day business (hint: those people do not occupy the executive suites).

Whenever I speak about Flowcasting, there are inevitably people who come up to me afterward to tell me “we’re basically doing Flowcasting in our company right now!”

I hate that word “basically”.

However, my interest is always piqued when I hear this, so I press for more information. Here are some “flavours” that I’ve heard over the years:

– “We have 52 week time-phased plans at our stores feeding back to the DCs but only for a selection of items.”

– “We do time-phased planning for all of our items in all of our stores and DCs, but they aren’t connected to each other.”

– “We have plans connecting all stores to all DCs for all items, but only over the next 13 weeks.”

– “We use multi-echelon planning for our regular volume, but we have a separate push process for promotions.”

I don’t know what the names of each of these variations might be, but the stuff after the “but” means that it is definitely not Flowcasting.

Flowcasting only works as a sum of all its parts and you can’t pick and choose criteria from a menu and label it “basically Flowcasting”, any more than you can say that any object with 4 wheels on it is “basically a Ferrari”.

So how can you know if you’re truly “doing Flowcasting” without having to read a 300 page book on the subject? Fortunately, the Flowcasting process is detailed and specific, without much room for a lot of “eye of the beholder” stuff.

If all of the following are true, I think you can reasonably claim that you are part of a Flowcasting organization:

– If someone chooses any item in your assortment, any location in your network and any day in the upcoming year, you can enter these 3 pieces of information into a system and it will tell you what your sales, receipts and on hand balance will be for that item in that location on that day.

– For any item at any selling location, you can see a 52 week sales forecast that incorporates all known future demand influences, including seasons and promotions.

– Except in rare cases, all orders and stock transfers are administrative “non events” that trigger automatically with a single lead-time and without any human intervention. Yes, that goes for promo orders too.

– You can view any location in your network and you can see inbound, outbound and inventoried cubic feet, weight, pallet positions, picks, receipts and labour hours for any day in the next 52 weeks – and you can drill down to see precisely which items are driving those high level metrics on that day.

– You can automatically create a financial plan for any item, category or across the entire business for the next 52 weeks, including sales, purchases and inventory investments. Then you can update the forecast for one item and the financial plan at any level automatically refreshes.

You may not be able to get there overnight, but once you’re there, you are Flowcasting. Basically.

Okay, rant over.

Cross Purposes

Posted on February 4, 2015 by Jeff Harrop

A compromise is the art of dividing a cake in such a way that everyone believes he has the biggest piece. – Ludwig Erhard (1897 – 1977)

Last week, I had the chance to catch up with a good friend and colleague of many years. His name is Ian and he is the VP of Supply Chain at a mid-sized retailer (with vast prior experience at a large retailer).

After a couple of beers, he asked me point blank: “Forecasting and Replenishment. One job or two?”

Without hesitation, I responded “Two!” Then I proceeded to describe the differences in skill sets, business relationships and aptitudes between a Demand Planner and a Supply Planner.

“Okay”, he said. “Who does the Demand Planning group report to?”

That’s a very interesting question indeed.

The goal of the demand planning process is to create a sales forecast that is as accurate and unbiased as possible. In retail, the process of coming up with a forecast is typically a “joint effort” between the Category Manager and the Supply Chain Planner.

The Category Manager is measured primarily on sales. Therefore he/she has a tendency to make optimistic projections and bias the forecast upward, knowing that a higher forecast will buy more inventory, thereby (theoretically) reducing the likelihood of lost sales.

The Supply Chain Planner is measured primarily on inventory turns. Therefore he/she has a tendency to “keep the forecast lean” to avoid carryover inventory after a promotion or selling season.

Therein lies the rub. If you give control of the forecasting process to the Merchandising group, Supply Chain feels like you’re “putting the fox in charge of the henhouse”. If the forecasting process falls under Supply Chain, Merchandising feels like they have no control over one of the key inputs that drive their businesses.

So should the Demand Planning function reside within Merchandising or Supply Chain?

Neither.

Think about it. Neither group can be faulted for exhibiting behaviour on which they are rewarded. The problem is that they have competing objectives and the biases on either side can have a direct negative impact on the P&L.

That’s why the Demand Planning function needs to report to someone who has accountability for the entire P&L – either the CEO or, more practically, the CFO.

Remember the goal for the sales forecast: As accurate and as unbiased as possible. By having Demand Planners report into Finance, they can be effective mediators between the competing groups and would have “a seat at the table” when matters relating to the sales forecast are discussed (promotions, product launches, safety stock policies, etc.)

Their job would be to chair S&OP meetings with their Merchandising and Supply Chain counterparts and hear both sides of the story with an objective ear. Without either group having the direct ability to impose their biases on the forecast, they must instead make a convincing case to the Demand Planner to support their view.

As a consequence, the current “blurred lines” of accountability are made clear once and for all:

Merchandising: Stimulate demand and be accountable for sales and gross margin.

Demand Planning: Forecast demand and be accountable for forecast quality.

Supply Chain: Optimize supply and be accountable for inventory turns and availability.

Collaboration Kool-Aid

Posted on January 7, 2015 by Mike Doherty

Theories come easy. Robust models that work in the real world do not. – Mark Payne, Fahrenheit 212

Recently I was asked by a client to describe myself in one sentence. After pondering it, the client smiled when I answered:

“I’m equal part genius, equal part buffoon”.

And I really believe that – at least the buffoon part.

I’ve never been able to dance very well with conventional wisdom. For some unknown reason, I can’t seem to follow along. I’ve always had a healthy dose of skepticism for conventional wisdom – especially in supply chain planning.

Recently a client of mine challenged me on the conventional wisdom regarding demand planning collaboration. First off, a bit of background about my client – let’s call him Ken.

When it comes to demand planning collaboration, Ken has a significant advantage over me.

He has no experience.

Yup, he’s never designed or implemented anything regarding retail supply chain planning (Flowcasting or otherwise).

However, he does know retail. Cold. Ken is a smart dude and a quick study. He understands Flowcasting and its potential.

When we were talking and designing the demand planning process that will drive Flowcasting, the topic of collaboration came up. I did what any consultant would do and immediately offered Ken a large glass of Collaboration Kool-Aid.

After all, the conventional wisdom from every planning expert is that when it comes to demand planning, collaboration is king. Luckily Ken’s got a little buffoon in him too and he started to ask some pretty specific questions – questions that I’ve been pondering as well.

He pointed out that in a Flowcasting world, there is only one forecast – a time-phased forecast of sales to the consumer, by item, by store.

By his thinking a retail demand planner would be accountable for this forecast. And, of course, Flowcasting technology would generate the math forecast for possibly millions of store/item combinations. By his logic no collaboration would be necessary. I agreed.

But what about promotions and new items, I argued? Surely collaboration would be needed there and would improve the forecast. After all, it’s conventional wisdom. Then he floored me with the following, regarding demand planning collaboration:

Outline the specific steps each person would perform in the demand planning process
Prove that more people improve the process enough to warrant the extra effort

It was a difficult exercise and we both struggled to document, on a white board, actual steps different people would take that we believed would actually improve the process and resulting forecast.

After all, in a Flowcasting world, we are forecasting consumer demand. When it comes to promotions, a demand planner would use history to help predict demand. But, argued Ken, so would anyone else involved in the process (like a category leader or even a supplier).

And would their actual forecast really differ significantly from that of the demand planner, if they were unbiased and just trying to predict the sales? And, would their steps to arrive at the forecast be much different? If so, what would the steps be? I struggled. Started to babble. Then started to sweat. I couldn’t really come up with any significant difference. And you know what they say…”if 2 people in business always agree, then one of them is not necessary”.

After more debate, we ended up in the same place even for new items.

Our conclusion was completely the opposite of conventional wisdom – A single demand planner, accountable for the forecast, would produce as good a forecast as a team collaborating on it, with considerably less effort and a much simpler process.

If the demand planner wanted advice or input on a specific forecast, they would ask (possibly a category manager). But, we both agreed – that would happen very rarely. In terms of proving that having people collaborate on the forecast produces a better consumer demand forecast, there really is no documented proof. Only conventional wisdom, supported by an army of Kool-Aid drinkers.

Now I know what my long time colleagues and fellow self-anointed supply chain experts are thinking – wow, Ken really is a buffoon.

By the way, I happen to agree with him.

Guess that makes me a buffoon as well.

Told you so.