Changing the game

In 1972, for my 10th birthday, my Mom would buy me a wooden chess set and a chess book to teach me the basics of the game.  Shortly after, I’d become hooked and the timing was perfect as it coincided with Bobby Fischer’s ascendency in September 1972 to chess immortality – becoming the 11th World Champion.

As a chess aficionado, I was recently intrigued by a new and different chess book, Game Changer, by International Grandmaster Matthew Sadler and International Master Natasha Regan.

The book chronicles the evolution and rise of computer chess super-grandmaster AlphaZero – a completely new chess algorithm developed by British artificial intelligence (AI) company DeepMind.

Until the emergence of AlphaZero, the king of chess algorithms was Stockfish.  Stockfish was architected by providing the engine the entire library of recorded grandmaster games, along with the entire library of chess openings, middle game tactics and endgames.  It would rely on this incredible database of chess knowledge and it’s monstrous computational abilities.

And, the approach worked.  Stockfish was the king of chess machines and its official chess rating of around 3200 is higher than any human in history.  In short, a match between current World Champion Magnus Carlsen and Stockfish would see the machine win every time.

Enter AlphaZero.  What’s intriguing and instructive about AlphaZero is that the developers took a completely different approach to enabling its chess knowledge.  The approach would use machine learning.

Rather than try to provide the sum total of chess knowledge to the engine, all that was provided were the rules of the game.

AlphaZero would be architected by learning from examples, rather than drawing on pre-specified human expert knowledge.  The basic approach is that the machine learning algorithm analyzes a position and determines move probabilities for each possible move to assess the strongest move.

And where did it get examples from which to learn?  By playing itself, repeatedly. Over the course of 9 hours, AlphaZero played 44 million games against itself – during which it continuously learned and adjusted the parameters of its machine learning neural network.

In 2017 AlphaZero would play a 100 game match against Stockfish and the match would result in a comprehensive victory for AlphaZero.  Imagine, a chess algorithm, architected based on a probabilistic machine learning approach would teach itself how to play and then smash the then algorithmic world champion!

What was even more impressive to the plethora of interested grandmasters was the manner in which AlphaZero played.  It played like a human, like the great attacking players of all time – a more precise version of Tal, Kasparov, and Spassky, complete with pawn and piece sacrifices to gain the initiative.

The AlphaZero story is very instructive for us supply chain planners and retail Flowcasters in particular.

As loyal disciples know, retail Flowcasting requires the calculation of millions of item/store forecasts – a staggering number.  Not surprisingly, people cannot manage that number of forecasts and even attempting to manage by exception is proving to have its limits.

What’s emerging, and is consistent with the AlphaZero story and learning, is that algorithms (either machine learning or a unified model approach) can shoulder the burden of grinding through and developing item/store specific baseline forecasts of sales, with little to no human touch required.

If you think about it, it’s not as far-fetched as you might think.  It will facilitate a game changing paradigm shift in demand planning.

First, it will relieve the burden of demand planners from learning and understanding different algorithms and approaches for developing a reasonable baseline forecast. Keep in mind that I said a reasonable forecast.  When we work with retailers helping them design and implement Flowcasting most folks are shocked that we don’t worship at the feet of forecast accuracy – at least not in the traditional sense.

In retail, with so many slow selling items, chasing traditional forecast accuracy is a bit of a fool’s game.  What’s more important is to ensure the forecast is sensible and assess it on some sort of a sliding scale.  To wit, if you usually sell between 20-24 units a year for an item at a store with a store-specific selling pattern, then a reasonable forecast and selling pattern would be in that range.

Slow selling items (indeed, perhaps all items) should be forecasted almost like a probability…for example, you’re fairly confident that 2 units will sell this month, you’re just not sure when.  That’s why, counter-intuitively, daily re-planning is more important than forecast accuracy to sustain exceptionally high levels of in-stock…whew, there, I said it!

What an approach like this means is that planners will no longer be dilly-dallying around tuning models and learning intricacies of various forecasting approaches.  Let the machine do it and review/work with the output.

Of course, sometimes, demand planners will need to add judgment to the forecast in certain situations – where the future will be different and this information and resulting impacts would be unknowable to the algorithm.  Situations where planners have unique market insights – be it national or local.

Second, and more importantly, it will allow demand planners to shift their role/work from analytic to strategic – spending considerably more time on working to pick the “winners” and developing strategies and tactics to drive sales, customer loyalty and engagement.

In reality, spending more time shaping the demand, rather than forecasting it.

And that, in my opinion, will be a game changing shift in thinking, working and performance.

Fossilized thinking

Fossilized

In August 1949 a group of fifteen smokejumpers – elite wild land firefighters – descended from the Montana sky to contain an aggressive fire near the Missouri River.  After hiking for a few minutes the foreman, Wagner Dodge, saw that the fire was raging – flames stretching over 30 feet in the air and blazing forward fast enough to cover two football fields every minute.

The plan was to dig a trench around the fire to contain it and divert it towards an area with little to burn.

Soon it became clear that the fire was out of control and the plan was out the window.  The fire was unstoppable so, instead, they’d try to outrun it, to safer ground.

For the next ten minutes, burdened by their heavy gear and tiring legs, the team raced up an incline, reaching an area that was only a few hundred yards from safety.  But the fire was unflinching, gaining ground like a wolf chasing down a wounded animal.

Suddenly, Dodge stopped.  He threw off his gear and, incredibly, took out some matches, lit them, and tossed them onto the grass.  His crew screamed at him but to no avail – when Dodge didn’t listen, they had no choice and turned and ran as fast as they could, leaving their foreman to what they believed to be certain peril.

But Dodge had quickly devised a different survival strategy: an escape fire.  By torching an area in front of him, he choked off the fuel for the fire to feed on.  Then, he poured water on a rag, put it over his mouth and lay down, face first, on the freshly burnt grass while the fire raged and sped past and over him.  In total, he’d spend close to 15 minutes living off the oxygen close the ground he’d just torched.

Sadly, of the rest of his crew that tried to outrun the blaze, only two would survive.

Wagner Dodge was able to survive not because of his physical fitness, but his mental fitness – the ability to rethink and unlearn.  The prevailing paradigm was that, at some stage for an out-of-control blaze, your only option is to try to out run it.  But Dodge was able to quickly rethink things – believing that, perhaps, by choking off it’s fuel line and providing his own small wasteland area, the fire might avoid him.  The ability to rethink had saved his life.

As it turns out, the ability to rethink and unlearn is also crucial for retails survival and revival.

It’s no secret, many retailers are struggling.  The same is true of many retail supply chains.

Do you ever really wonder why?

Lots of people blame retail’s generally slow adoption of new technologies and business models as the main factor, but I think it’s a deeper, more fundamental and chronic problem.

Technology is not eating retail.  Fossilized thinking is.

What’s fossilized thinking?  It’s people – at all levels in an organization – who are unwilling or unable to challenge their long-held beliefs.  Not only challenge them, but be able to rethink, unlearn and change them often.

As a case in point, many people who work in the retail supply chain don’t include the consumer as part of the supply chain.  Yet, if you think about it, the retail supply chain begins and ends with the consumer.  There are even a number of folks who don’t consider the store part of the supply chain.  Once the product has shipped from the DC to the store, then, incredibly, job done according to them.

Don’t believe me?

I won’t embarrass them, but just recently I read a “thought leadership” article from one of the world’s pre-eminent consulting firms regarding the top trends in retail supply chain management.  At #3, and I kid you not, was the growing view that the store was a key part of the supply chain.

Flowcasting tribe members know better and think differently.  The consumer and store have, and always will be, part of the supply chain.  That’s why we understand that, in retail, there is no such thing as a push supply chain – since you can’t push the product to the consumer.

In my opinion (and I’m not alone), fossilized thinking, not technology adoption, is the real disruptor in retail.

If you want to improve, innovate or disrupt then you must…

Constantly rethink, unlearn and challenge your own thinking!

Limit your inputs

Meditations

Marcus Aurelius is widely considered to be one of the wisest people of all time. His classic and colossal writings, ‘Meditations’, is a bible for crystal clear thinking, famously outlining the principles of stoicism.

In Meditations, he asks a profound question that all designers and implementers should ponder often…

”Is this necessary?”

Knowing what not to think about. What to ignore and not do. It’s an important question, especially when it comes to designing and implementing new ways of working.

As an example, consider the process of developing a forward looking forecast of consumer demand, by item, by store. Of course, loyal readers and disciples know that this is the forecast that provides a key input to allow a retail supply chain to be planned using Flowcasting.

You might get your knickers in a knot to learn that in one of our most recent and successful implementations of Flowcasting, the store level forecasting process uses only two key inputs:

  1. The actual sales history by item/store in units
  2. An indication if that sales history was during an abnormal period (e.g., a promotion, an unplanned event, a stock-out period, a different selling price, etc).

Now, I know what you’re thinking. What about all those ‘other’ things that influence consumer demand that many people espouse? You know, things like the weather, competitor activities and any other causal variables?

Counter-intuitively, all these additional ‘factors’ are not really required at the retail store level and for very good reasons.

First, did you know that for most retailers 50-60% of the item/store combinations sell 24 or less units per year. That’s less than one unit every two weeks. Furthermore, about 70-80% of the item/store combinations sell less than 52 units per year, or about 1 unit per week.

Consider the very slow sellers – selling 24 or less units a year. If the last 52 weeks sales was 24 units and so was the previous year’s, it would stand to reason that a reasonable forecast for the upcoming 52 weeks would be around 24 units.

Keep in mind, that as actual sales happen the forecasting process would always be re-forecasting, looking ahead and estimating the upcoming 52 weeks consumer demand based on most recent history.

Now, consider a 52 week forecast of 24 units. That breaks down to a weekly forecast of 0.46.

Factoring in additional variables is not likely to make actual sales of 30 units happen (assuming stock outs were not excessive) – which would be the equivalent of increasing previous year’s sales by 25%. A higher forecast does not mean sales will actually happen.

Even a 10% increase in the forecast would only increase the annual forecast by about 2 units, or about .04 units per week.

Is there really a difference between an average weekly forecast of 0.46 and 0.50 units? Isn’t that essentially the same number? In terms of a forecast, they are both reasonable (in reality, the forecast would also have a pattern to the expected sales and would be expressed as integers, but you get the point).

One of the keys of the retail Flowcasting process is using only a limited number of inputs to build the item/store forecast, while allowing people to easily understand and thus manage it – by a very limited number of exceptions that could not be automatically resolved using system rules.

Add some basic supply information (like inventory balances and ordering rules) to the forecast and voila – the entire supply chain can be calculated/planned from store to supplier – every day, for every planned inventory flow and projection for a rolling 52 weeks into the future.

What’s elegant and inherently beautiful about Flowcasting is that daily re-planning the entire supply chain provides the agility to adjust current and planned inventory flows and ensures everyone is working to a single set of numbers – based on the drumbeat of the consumer. It negates the need to find the ‘perfect forecast’ and, as such, allows us to limit our inputs to the bare essentials.

Marcus Aurelius was right.

Limit your inputs and always ask, “Is this necessary?”

It’s good advice in business, in life and especially store level forecasting.

Experts on the future

In 1971, Judah Folkman, a doctor working in Boston, developed a new approach to treat cancer – essentially by stopping the blood vessels supplying the tumors. Blocking the flow, he concluded, would halt the growth of the tumors.

At the time, the only accepted and endorsed approach to treating cancer was chemotherapy. Dr Folkman’s idea was scorned and ridiculed by the medical establishment consisting of a group of PhD insiders – mostly from the field of biology.

According to Dr Folkman, when he attempted to share his idea and thinking to the scientific community, the entire room would get up and leave – as if, collectively, they all had to take a piss at the exact same time. Over time, the criticism got so bad that special committees were developed to review his ideas and not only judged his idea to be of little value, they also threatened to revoke his medical license if he did not cease – in one letter, writing to him to reject his ideas and calling him a ‘clown’.

Folkman, however, was undaunted and pressed on. Painstakingly, his ideas were slowly starting to be accepted by more “open” thinkers and eventually morphed into drugs available for cancer trials. To his credit, in the summer of 2003, at a major medical conference, the results from a large trial for patients with advanced colon cancer validated Dr. Folkman’s thinking.

The way to treat cancer had been transformed. At the event, the crowd rose in a standing ovation. The presenter, at the time, said, to the effect, “it’s a shame that Dr. Folkman couldn’t be here to experience this” – little did he know that, sitting in a back row, Dr. Folkman had just smiled.

Eventually he couldn’t hide his fame and was asked about his achievement – which had taken the better part of 32 years. Most folks wanted to know how he felt and why he continued on his journey in light of all the criticism and personal attacks.

His answers were and still are very insightful.

First, in terms of the ridicule he proclaimed, “You can always tell the leader of new thinking from all the arrows in their ass”.

And even more profound about why he never gave up: “There are no experts of the future.”

Presently we’re living through an unprecedented time and there are a lot of questions about the future – how will the world look after the virus is subdued and what will the new normal look like? Some of these questions are focused on retail and supply chain management.

How will consumers change their behaviors? How much sales will be transacted online? Will home delivery become even more significant? Will supply chain networks become more diverse and less susceptible to a single country’s supply disruption? What other customer delivery methods will emerge?

All good questions for which my answer is the same as yours, “I really don’t know”. In my opinion, a lot will change, I’m just not sure where, when, how much and how fast.

Remember, “There are no experts of the future”.

What I do know is that human behavior will not change. We are social animals and like and need to acquire stuff. We just might shift, perhaps dramatically over time, how we go about this.

For us supply chain planners – especially retailers – that means having the supply chain driven by and connected to consumer demand will be crucial. As consumer demand shifts and evolves having a complete model of the business and providing longer term visibility to all stakeholders will be a core capability – both in the short to medium term but also longer term to proactively plan for and respond to the next disruption.

Wait a minute…that sounds like Flowcasting, doesn’t it?

The key to being in-stock

Key

Abraham Lincoln is widely considered the greatest President in history. He preserved the Union, abolished slavery and helped to strengthen and modernize government and the economy. He also led a fragile America through one of her darkest and most crucial periods – the American Civil War.

In the early days of the war, there were lots of competing ideas about how to secure victory and who should attempt it. Most of the generals at that time had concluded that the war could only be won through long, savage and bloody battles in the nation’s biggest cities – like Richmond, New Orleans and even Washington.

Lincoln – who taught himself strategy by reading obsessively – had a different plan. He laid out a large map and pointed to Vicksburg, Mississippi, a small city deep in the South. Not only did it control important navigation waterways, but it was also a junction of other rivers, as well as the rail lines that supplied Confederate armies and plantations across the South.

“Vicksburg is the key”, he proclaimed. “We can never win the war until that key is ours”.

As it turns out, Lincoln was right.

It would take years, blood, sweat and ferocious commitment to the cause, but his strategy he’d laid out was what won the war and ended slavery in America forever. Every other victory in the Civil War was possible because Lincoln had correctly understood the key to victory – taking the city that would split the South in half and gaining control of critical shipping lanes.

Lincoln understood the key. Understanding the key is paramount in life and in business.

It’s no secret that many retailers are struggling – especially in terms of the customer journey – most notably when it comes to retail out of stocks. Retail out of stocks have remained, on average, sadly, at 8% for decades.

So what’s the key to finally ending out-of-stocks?

The key is speed and completeness of planning.

First, we all know that the retail supply chain can and should only be driven by a forward looking forecast of consumer demand – how much you think you’ll sell, by product and consumption location.

Second, everyone also agrees (though few understand the key to solving this thorn in our ass) that store/location on-hands need to be accurate.

But the real key is that, once these are in place, the planning process must be at least done daily and must be complete – from consumption to supplier.

Daily re-forecasting and re-planning is necessary to re-orient and re-synch the entire supply chain based on what did or didn’t sell yesterday. Forecasts will always be wrong and speedy re-planning is the key to mitigating forecast error.

However, that is not enough to sustain exceptionally high levels of daily in-stock. In addition, the planning process must be complete – providing the latest projections from consumption to supply, giving all trading partners their respective projections in the language in which they operate (e.g., units, volume, cube, weight, dollars). The reason is simple – all partners need to see, as soon as possible, the result of the most up to date plans. All plans are re-calibrated to help you stay in stock. And the process repeats, day in, day out.

We have retail clients that are achieving, long term, daily in-stocks of 98%+, regardless of the item, time of year or planning scenario.

They understand the key to making it happen.

Now you do too.

Pissed Off People

Jim is basically your average bloke. One Saturday afternoon, about 25 years ago, he’s doing something a lot of average blokes do; cleaning his home – a small farmhouse in the west of England.

After some dusting, it’s time to vacuum. Like everyone at the time, he’s shocked how quickly his top-of-the-line Hoover cleaner loses its suction power.

Jim is pissed. Royally pissed off. Madder than a wet hen.

So mad, in fact, that he took the cleaner out to his shed, took it apart and examined why it would lose suction power so quickly. After a few experiments he correctly deduced that the issue was that fine dust blocked the filter almost immediately and that’s why performance in conventional cleaners dips so fast.

Jim continued to be pissed until one day he visited a timber mill, looking for some wood. In those days, timber mills planed the logs on the spot for you. Jim watched as he saw his wood travel along until it reached a cyclone specifically designed to change the dynamics of airflow, separating the dust from the air via centrifugal force.

BOOM! James Dyson, still pissed at how shit traditional vacuum cleaners were, got the core idea of the Dyson cyclone cleaner. An idea that he would use to eventually deposit over £3 billion into his back pocket.

Unbelievably it took Dyson three years and 5,127 small, incremental prototypes to finally “perfect” his design and revolutionize cleaning forever. Can you imagine how pissed you’d need to be to work, diligently, over that many iterations to finally see your idea through?

Dyson’s story is incredible and enlightening – offering us a couple of key insights into the innovative process.

First, most folks think that innovation happens as a result of ideas just popping into people’s heads. That’s missing the key piece of the puzzle: the problem! Without a problem, a flaw, a frustration, innovation cannot happen. As Dyson himself states, “creativity should be thought of as a dialogue. You have to have a problem before you can have game-changing innovation”.

Second, for innovative solutions to emerge you need pissed off people. People like Dyson who are mad, frustrated and generally peeved with current solutions and approaches for the problem at hand. So they are always thinking, connecting and, at times, creating a breakthrough solution – sometimes years after initially surfacing the problem. So, while it’s easy to say that the “idea” just happened, more often than not you’ve been mulling it over, subconsciously, because you’re pissed about something.

Here’s a true story about Flowcasting and how it eventually saw the light of day as a result of some pissed off people.

About 25 years ago, I was the leader of a team whose mandate was to improve supply chain planning for a large, very successful Canadian retailer. I won’t bore you with the details but eventually we designed, on paper, what we now call Flowcasting.

Problem was that it was very poorly received by the company’s Senior Leadership team, especially the Supply Chain executives. On numerous occasions I was informed that this idea would never work and that we needed to change the design. I was also threatened to be fired more than once if we didn’t change.

The problem was, our team loved the design and could see it potentially working. As I was getting more pressure and “never” from the leadership team, I was getting more and more pissed. Royally pissed off as a matter of fact.

As luck would have it, as a pissed off person, I didn’t back down (there’s a lesson here too – “never” is not a valid reason why something might not work, regardless who says it). One person on the team suggested I contact Andre Martin and he and his colleague, Darryl Landvater, helped us convince the non-believers that it would be the future and that we should pilot a portion of the design. The rest is, of course, history.

The Flowcasting saga didn’t stop there. As we were embarking on our early pilot of the DC-supplier integration, Andre and Darryl tried, unsuccessfully, to convince a few major technology planning vendors that an integrated solution, from store/consumption to supply was needed and that they needed to build it, from scratch.

All the major technology players turned them down, citing lots of “nevers” themselves as to why this solution was either not needed, or would not scale and/or work.

To be honest, it pissed them off, as they’ve admitted to me many times over the years.

So much so, that, despite all the warnings from the experts they “put their money where their mouth is” and built a Flowcasting solution that connects the store to supplier in an elegant, intuitive and seamless fashion – properly planning for crucial retail planning scenarios like slow sellers, promotions, and seasonal items just to name a few.

In 2015, using the concept of Flowcasting and the technology that they developed, a retailer seamlessly connected their supply chain from consumption to supply – improving in-stocks, sales and profits and instilling a process that facilitates any-channel planning however they wish to do it.

Sure, having a reasonably well thought out design was important. As was having a solution suited for the job.

But what really enabled the breakthrough were some pissed-off people!

Covered in Warts

It’s the early 1990’s and Joanne is down on her luck. A recently divorced, single mother who’s jobless, she decides to move back from England to Scotland to at least be closer to her sister and family.

During her working days in Manchester she had started scribbling some ideas and notes about a nonsensical book idea and, by the time she’d moved home, had three chapters written of a book. Once back near Edinburgh, she continued to write and improve her manuscript until she had a first draft completed in 1995 – fully five years from her first penned thoughts.

During the next two years she pitched the very rough manuscript to a dozen major publishers. They all rejected it and believed the story would not resonate with people and, as a result, sales would be dismal.

Undismayed she eventually convinced Bloomsbury to take a very small chance on the book – advancing her a paltry $1500 pounds and agreeing to print 1,000 copies, 500 of which would be sent to various libraries.

In 1997 and 1998 the book, Harry Potter by J. K. Rowling, would win both the Nestle Book award and the British Book Awards Children’s book of the year. That book would launch Rowling’s worldwide success and, to date, her books have sold over 400 million copies.

The eventual success of the Harry Potter series of books is very instructive for breakthroughs and innovation.

The most important breakthroughs—the ones that change the course of science, business, or history — are fragile. They rarely arrive dazzling everyone with their brilliance.

Instead, they often arrive covered in warts — the failures and seemingly obvious reasons they could never work that make them easy to dismiss. They travel through tunnels of skepticism and uncertainty, their champions often dismissed as crazy.

Luckily most of the champions of breakthrough items are what many would describe as loons – people that refuse to give up on their ideas and will work, over time, to smooth and eliminate the warts.

When it comes to supply chain planning innovation, you’d have to put Andre Martin into the loon category as well.

In the mid 1970’s Andre invented Distribution Resource Planning (DRP) and, along with his colleague Darryl Landvater, designed and implemented the first DRP system in 1978 – connecting distribution to manufacturing and changing planning paradigms forever.

Most folks don’t know but around that time Andre saw that the thinking of DRP could be extended to the retail supply chain – connecting the store to the factory using the principles of DRP and time-phased planning.

The idea, which has since morphed and labelled as Flowcasting, was covered in warts. During the course of the last 40 years Andre and Darryl have refined the thinking, smoothed the warts, eliminated dissention, educated an industry and, unbelievably, built a solution that enables Flowcasting.

I’ve been a convert and a colleague in the wart-reduction efforts over the last 25 years – experiencing first-hand the some irrational responses and views from, first, a large Canadian retailer, and more recently the market in general.

But, like JK, the warts are largely being exposed as pimples and people and retailers are seeing the light – the retail supply chain can only deliver if it’s connected from consumer to supplier – driven only by a forecast of consumer demand. Planned and managed using the principles of Flowcasting.

The lesson here is to realize that if you think you’ve got a breakthrough idea, there’s a good chance it’ll be covered in warts and will need time, effort, patience and determination to smooth and eliminate them.

It can, however, be done.

And you can do it.

Godspeed.

Managing the Long Tail

If you don’t mind haunting the margins, I think there is more freedom there. – Colin Firth

long-tail

 

A couple of months ago, I wrote a piece called Employing the Law of Large Numbers in Bottom Up Forecasting. The morals of that story were fourfold:

  1. That when sales at item/store level are intermittent (fewer than 52 units per year), a proper sales pattern at that level can’t be properly determined from the demand data at that level.
  2. That any retailer has a sufficient percentage of slow selling item/store combinations that the problem simply can’t be ignored in the planning process.
  3. That using a multi level, top-down approach to developing properly shaped forecasts in a retail context is fundamentally flawed.
  4. That the Law of Large Numbers can be used in a store centric fashion by aggregating sales across similar items at a store only for the purpose of determining the shape of the curve, thereby eliminating the need to create any forecasts above item/store level.

A high level explanation of the Profile Based Forecasting approach developed by Darryl Landvater (but not dissimilar to what many retailers were doing for years with systems like INFOREM and various home grown solutions) was presented as the antidote to this problem. Oh and by the way, it works fabulously well, even with such a low level of “sophistication” (i.e. unnecessary complexity).

But being able to shape a forecast for intermittent demands without using top-down forecasting is only one aspect of the slow seller problem. The objective of this piece is to look more closely at the implications of intermittent demands on replenishment.

The Bunching Problem

Regardless of how you provide a shape to an item/store forecast for a slow selling item (using either Profile Based Forecasting or the far more cumbersome and deeply flawed top-down method), you are still left with a forecasted stream of small decimal numbers.

In the example below, the shape of the sales curve cannot be determined using only sales history from two years ago (blue line) and the most recent year (orange line), so the pattern for the forecast (green dashed line) was derived from an aggregation of sales of similar items at the same store and multiplied through the selling rate of the item/store itself (in this case 13.5 units per year):

You can see that the forecast indeed has a defined shape – it’s not merely a flat line that would be calculated from intermittent demand data with most forecasting approaches. However, when you multiply the shape by a low rate of sale, you don’t actually have a realistic demand forecast. In reality, what you have is a forecast of the probability that a sale will occur.

Having values to the right of the decimal in a forecast is not a problem in and of itself. But when the value to the left of the decimal is a zero, it can create a huge problem in replenishment.

Why?

Because replenishment calculations always operate in discrete units and don’t know the difference between a forecast of true demand and a forecast of a probability of a sale.

Using the first 8 weeks of the forecast calculated above, you can see how time-phased replenishment logic will behave:

The store sells 13 to 14 units per year, has a safety stock of 2 units and 2 units in stock (a little less than 2 months of supply). By all accounts, this store is in good shape and doesn’t need any more inventory right now.

However, the replenishment calculation is being told that 0.185 units will be deducted from inventory in the first week, which will drive the on hand below the safety stock. An immediate requirement of 1 unit is triggered to ensure that doesn’t happen.

Think of what that means. Suppose you have 100 stores in which the item is slow selling and the on hand level is currently sitting at the safety stock (not an uncommon scenario in retail). Because of small decimal forecasts triggering immediate requirements at all of those stores, the DC needs to ship out 100 pieces to support sales of fewer than 20 pieces at store level – demand has been distorted 500%.

Now, further suppose that this isn’t a break-pack item and the ship multiple to the store is an inner pack of 4 pieces – instead of 100 pieces, the immediate requirement would be 400 pieces and demand would be distorted by 2,000%!

The Antidote to Bunching – Integer Forecasts

What’s needed to prevent bunching from occurring is to convert the forecast of small decimals (the probability of a sale occurring) into a realistic forecast of demand, while still retaining the proper shape of the curve.

This problem has been solved (likewise by Darryl Landvater) using simple accumulator logic with a random seed to convert a forecast of small decimals into a forecast of integers.

It works like this:

  • Start with a random number between 0 and 1
  • Add this random number to the decimal forecast of the first period
  • Continue to add forecasts for subsequent periods to the accumulation until the value to the right of the decimal in the accumulation “tips over” to the next integer – place a forecast of 1 unit at each of these “tip-over” points

Here’s our small decimal forecast converted to integers in this fashion:

Because a random seed is being used for each item/store, the timing of the first integer forecast will vary by each item/store.

And because the accumulator uses the shaped decimal forecast, the shape of the curve is preserved. In faster selling periods, the accumulator will tip over more frequently and the integer forecasts will likewise be more frequent. In slower periods, the opposite is true.

Below is our original forecast after it has been converted from decimals to integers using this logic:

And when the requirements across multiple stores are placed back on the DC, they are not “bunched” and a more realistic shipment schedule results:

Stabilizing the Plans – Variable Consumption Periods

Just to stay grounded in reality, none of what has been described above (or, for that matter, in the previous piece Employing the Law of Large Numbers in Bottom Up Forecasting) improves forecast accuracy in the traditional sense. This is because, quite frankly, it’s not possible to predict with a high degree of accuracy the exact quantity and timing of 13 units of sales over a 52 week forecast horizon.

The goal here is not pinpoint accuracy (the logic does start with a random number after all), but reasonableness, consistency and ease of use. It allows for long tail items to have the same multi-echelon planning approach as fast selling items without having separate processes “on the side” to deal with them.

For fast selling items with continuous demand, it is common to forecast in weekly buckets, spread the weekly forecast into days for replenishment using a traffic profile for that location and consume the forecast against actuals to date for the current week:

In the example above, the total forecast for Week 1 is 100 units. By end of day Wednesday, the posted actuals to date totalled 29 units, but the original forecast for those 3 days was 24 units. The difference of -5 units is spread proportionally to the remainder of the week such as to keep the total forecast for the week at 100 units. The assumption being used is that you have higher confidence in the weekly total of 100 units than you have in the exact daily timing as to when those 100 units will actually sell.

For slow moving items, we would not even have confidence in the weekly forecasts, so consuming forecast against actual for a week makes no sense. However, there would still be a need to keep the forecast stable in the very likely event that the timing and magnitude of the actuals don’t match the original forecast. In this case, we would consume forecast against actuals on a less frequent basis:

The logic is the same, but the consumption period is longer to reflect the appropriate level of confidence in the forecast timing.

Controlling Store Inventory – Selective Order Release

Let’s assume for a moment a 1 week lead time from DC to store. In the example below, a shipment is planned in Week 2, which means that in order to get this shipment in Week 2, the store needs to trigger a firm replenishment right now:

Using standard replenishment rules that you would use for fast moving items, this planned shipment would automatically trigger as a store transfer in Week 1 to be delivered in Week 2. But this replenishment requirement is being calculated based on a forecast in Week 2 and as previously mentioned, we do not have confidence that this specific quantity will be sold in this specific week at this specific store.

When that shipment of 1 unit arrives at the store (bringing the on hand up to 3 units), it’s quite possible that you won’t actually sell it for several more weeks. And the overstock situation would be further exacerbated if the order multiple is greater than 1 unit.

This is where having the original decimal forecast is useful. Remember that, as a practical matter, the small decimals represent the probability of a sale in a particular week. This allows us to calculate a tradeoff between firming this shipment now or waiting for the sale to materialize first.

Let’s assume that choosing to forgo the shipment in Week 2 today means that the next opportunity for a shipment is in Week 3. In the example below, we can see that there is a 67.8% chance (0.185 + 0.185 + 0.308) that we will sell 1 unit and drop the on hand below safety stock between now and the next available ship date:

Based on this probability, would you release the shipment or not? The threshold for this decision could be determined based on any number of factors such as product size, cost, etc. For example, if an item is small and cheap, you might use a low probability threshold to trigger a shipment. If another slow selling item is very large and expensive, you might set the threshold very high to ensure that this product is only replenished after a sale drives the on hand below the safety stock.

Remember, the probabilities themselves follow the sales curve, so an order has a higher probability of triggering in a higher selling period than in a lower selling period, which would be the desired behaviour.

The point of all of this is that the same principles of Flowcasting (forecast only at the point of consumption, every item has a 52 week forecast and plan, only order at the lead time, etc.) can still apply to items on the long tail, so long as the planning logic you use incorporates these elements.

Ordinary Observation

OrdinaryObservation

It’s September 28, 1928 in a West London lab. A young physician, Alex, was doing some basic research that had been assigned to him regarding antibacterial agents. He’d been doing the same thing for a number of days when one day he noticed something odd.

What caught his eye and attention that fateful day was that mold had actually killed some bacteria in one of his plates. Usually samples like this are discarded, but instead Alex kept this sample and began to wonder. If this mold could kill this type of bacteria, could it be used to kill destructive bacteria in the human body?

Alexander Fleming would spend the next 14 years working out the kinks and details before “penicillin” was officially used to treat infections. It was an invention that would revolutionize medicine by discovering the world’s first antibiotic.

Dr. Fleming was able to develop this innovation through the simple power of ordinary observation. Sherlock Holmes famously said once to Watson: “You see, but you do not observe. The distinction is clear.” According to psychologist and writer Maria Konnikov…“To observe, you must learn to separate situation from interpretation, yourself from what you are seeing.

Here’s another example of the power of observation. Fast forward to 1955, a relatively unknown and small furniture store in Almhult, Sweden. One day, the founder and owner noticed something odd. An employee had purchased a table to take home to the family. Rather than struggling to try to cram the assembled table into his car, this employee took the legs off and carefully placed them in a box, which, in turn, would fit nicely in his car for delivery home.

As it turned out, the owner of the store, Ingvard Kamprad, would observe this unpacking phenomena regularly. Carefully he observed what his employees were doing and why it was so effective. And, if this concept was better for his employees, it would stand to reason that it would also be better for his customers – and the bottom line.

Soon after, Kamprad would work tirelessly to perfect the idea of selling dis-assembled furniture – changing the customer journey for furniture acquisition forever, and making IKEA synonymous with this brand promise and a worldwide household name. All because of the power of ordinary observation.

A final story about observation and its impact on supply chain planning.

Ken Moser is one of Canada’s top retailers – leading and managing one of Canadian Tire’s best stores in northern Ontario. About 15 years ago, he was visited by a chap who would eventually build the world’s first and, to date, best Flowcasting solution.

This person followed Ken around the store, asking questions and observing how the store operated and how Ken thought – particularly about how to manage the inventory of tens of thousands of items. Rumour has it that when Ken got to a section of the store, he proclaimed something like…”these items are like a set-it-and-forget-it. I have no idea when they’ll sell, and neither do you. All I know is that, like clockwork, they’ll only sell one a month. For others, it’s like one every quarter.”

Our Flowcasting architect was fascinated with this observation and spent time watching/observing customers perusing this section of the store. And like the two examples above, deep observation and reflection would eventually morph into an approach to forecasting and planning slow selling items that is, to date, the only proven solution in retail. All from the awesome power of ordinary observation.

Yogi Berra, the great Yankee catcher and sometimes philosopher, hit the nail on the proverbial head regarding the importance of ordinary observation when he proclaimed…

You can observe a lot, just by watching.

Turns out, you can.

Noise is expensive

Noise

Did you know that the iHome alarm clock, common in many hotels, shows a small PM when the time is after 12 noon?  You wonder how many people fail to note the tiny ‘pm’ isn’t showing when they set the alarm, and miss their planned wake up.  Seems a little complicated and unnecessary, wouldn’t you agree?

Did you also know that most microwaves also depict AM or PM? If you need the clock in the microwave to tell you whether it’s morning or night, somethings a tad wrong.

More data/information isn’t always better. In fact, in many cases, it’s a costly distraction or even provides the opportunity to get the important stuff wrong.

Contrary to current thinking, data isn’t free.

Unnecessary data is actually expensive.

If you’re like me, then your life is being subjected to lots of data and noise…unneeded and unwanted information that just confuses and adds complication.

Just think about shopping now for a moment.  In a recent and instructive study sponsored by Oracle (see below), the disconnect between noise and what consumers really want is startling:

  1. 95% of consumers don’t want to talk or engage with a robot
  2. 86% have no desire for other shiny new technologies like AI or virtual reality
  3. 48% of consumers say that these new technologies will have ZERO impact on whether they visit a store and even worse, only 14% said these things might influence them in their purchasing decisions

From the consumers view what this is telling us, and especially supply chain technology firms, we don’t seem to understand what’s noise and what’s actually relevant. I’d argue we’ve got big time noise issues in supply chain planning, especially when it relates to retail.

I’m talking about forecasting consumer sales at a retail store/webstore or point of consumption.  If you understand retail and analyze actual sales you’ll discover something startling:

  1. 50%+ of product/store sales are less than 20 per year, or about 1 every 2-3 weeks.

Many of the leading supply chain planning companies believe that the answer to forecasting and planning at store level is more data and more variables…in many cases, more noise. You’ll hear many of them proclaim that their solution takes hundreds of variables into account, simultaneously processing hundreds of millions of calculations to arrive at a forecast.  A forecast, apparently, that is cloaked in beauty.

As an example, consider the weather.  According to these companies not only can they forecast the weather, they can also determine the impact the weather forecast has on each store/item forecast.

Now, since you live in the real world with me, here’s a question for you:  How often is the weather forecast (from the weather network that employs weather specialists and very sophisticated weather models) right?  Half the time?  Less?  And that’s just trying to predict the next few days, let alone a long term forecast.  Seems like noise, wouldn’t you agree?

Now, don’t get me wrong.  I’m not saying the weather does not impact sales, especially for specific products.  It does.  What I’m saying is that people claiming to predict it with any degree of accuracy are really just adding noise to the forecast.

Weather.  Facebook posts.  Tweets.  The price of tea in China.  All noise, when trying to forecast sales by product at the retail store.

All this “information” needs to be sourced.  Needs to be processed and interpreted somehow.  And it complicates things for people as it’s difficult to understand how all these variables impact the forecast.

Let’s contrast that with a recent retail implementation of Flowcasting.

Our most recent retail implementation of Flowcasting factors none of these variables into the forecast and resulting plans.  No weather forecasts, social media posts, or sentiment data is factored in at all.

None. Zip. Zilch.  Nada.  Heck, it’s so rudimentary that it doesn’t even use any artificial intelligence – I know, you’re aghast, right?

The secret sauce is an intuitive forecasting solution that produces integer forecasts over varying time periods (monthly, quarterly, semi-annually) and consumes these forecasts against actual sales. So, the forecasts and consumption could be considered like a probability.  Think of it like someone managing a retail store. They can say fairly confidently that “I know this product will sell one this month, I just don’t know what day”!

The solution also includes simple replenishment logic to ensure all dependent plans are sensible and ordering for slow selling products is based on your opinion on how probable you think a sale is likely in the short term (i.e., orders are only triggered for a slow selling item if the probability of making a sale is high).

In addition to the simple, intuitive system capabilities above, the process also employs and utilizes a different kind of intelligence – human.  Planners and category managers, since they are speaking the same language – sales – easily come to consensus for situations like promotions and new product introductions.  Once the system is updated then the solution automatically translates and communicates the impact of these events for all partners.

So, what are the results of using such a simple, intuitive process and solution?

The process is delivering world class results in terms of in-stock, inventory performance and costs.  Better results, from what I can tell, than what’s being promoted today by the more sophisticated solutions.  And, importantly, enormously simpler, for obscenely less cost.

Noise is expensive.

The secret for delivering world class performance (supply chain or otherwise) is deceptively simple…

Strip away the noise.