The old paradigm and the new...

in this post I'd like to write about historic development of MRP systems and most planner's pursuit of “The Holy Grail” to Improve Forecast Accuracy.
From Orlicky’s early MRP system through today’s Advanced Planning Systems there were lots of advancements from a technical point of view. With new databases like HANA we have now so much computing power that an explosion and net requirements calculation of a Bill of Material – no matter how deep or wide – is a non-issue. Integration was also accomplished. First, through MRP II’s addition of capacity and resources leveling as well as a material availability check. And with the dawn of ERP systems, notably SAP’s R/3 the assimilation of HR, Sales, Finance and much more was accomplished. Eventually, the Advanced planning Systems (in SAP that was APO) took center stage and promised full automation.  However, it was forgotten that complex, non-linear supply chains afforded much more from organizations that had to stay competitive. Noise-laden, constantly expedited production programs wrecked havoc on purchased parts requirements and materials planners still today are chasing forecasts and ever changing demands with loads of variability in the hope that they can expedite and manual correct infeasible plans. So... Over the years the tools have developed and grown technologically but not really delivered results
Additionally, demand fluctuations being propagated from consumer behavior upstream through the supply chain define the bullwhip effect. In fact, the increase of fluctuation is quite large as we go ‘backwards’ towards the source. Research indicates a fluctuation in point-of-sale demand of +/- five percent will be interpreted by supply chain participants as a change in demand of up to +/- forty percent. Much like cracking a whip, a small flick of the wrist (a shift in point of sale demand) can cause a large motion at the end of the whip (manufacturer’s response)

This phenomenon still holds true today in most supply chains. Companies driven by MRP, MRPII, ERP or APS systems and their inherent deterministic planning place pressure on planners. These are faced with sheer insurmountable tasks to keep high service levels to the production lines with low inventory holdings in the face of variability. The only remedy seems to improve forecast accuracy which is nearly impossible as the future still can’t be predicted with certainty. It is time to rethink and revive this very important department. Years of negligence due to excitement about new technologies have taken a toll on its performance and effectiveness. The Bullwhip Effect is getting the better of us and, in many cases, destroys cash flow and profitability

The new paradigm centers on variability and its detrimental impact on the bottom line. Variability is anticipated and part of the plan. Only then can you plan strategies to either reduce or absorb the enemy number one of any value chain. A reduction of variability directly translates into a reduction of required inventory levels, a lift of availability and with it the service levels and shorter cycle times

Instead of ‘single order related replenishment’ – which is a prevalent planning method in many of today’s organizations – our method is mostly demand neutral and plans with buffers to fulfill any actual requirements… within its defined performance boundaries and set service levels. This will make your system liberated and independent from demand swings and cuts out the uncertainty in forecasting.

Like a seawall holds of surges of water, the new paradigm provides protection from the bull whip effect and ensures noise-reduced and leveled replenishment and inventory management of your purchased parts, ingredients or materials. What we suggest is... Instead of creating a detailed plan for a single situation that will never occur… 

...create a set of policies that work for a range of situation, using the buffers time, inventory and capacity. The buffers Time, Capacity and Inventory are an integral and important part of our system of Effective Materials Planning as they serve as the basis for the construction of replenishment policies. As described in the award winning book Factory Physics® or the teachings of Demand Driven MRP, the right combination of buffers to counter variability is key. It reduces noise, increases transparency and automation, allows for integration and helps to align the plan with a company’s supply chain strategy. It even helps to drive awareness for the need of a strategy and the definition of performance boundaries and general user or planner guidelines.

Buffer Strategies help to standardize your Planning and Execution and provide standard operating procedures to help planners, buyers and schedulers to align their activities to achieve a large increase in performance and profitability.

How do you construct these buffers into SAP? That answer requires a bit of an elaborate discussion but in a nutshell: for the inventory buffer... instead of the static safety stock (which drives a bunch of dead stock) your using a dynamic safety stock with a range of cover profile. For the time buffer, you fix the availability checking rules and work with safety time (On MRP2) and to use capacity buffers you must set up your value stream and production scheduling methods in a way that your bottleneck work center runs below 100% utilization.

thoughts from SAPPHIRENOW 2016

I had the pleasure again to visit SAPPHIRENOW again... not sure how many times I have been there yet, but the first one I attended was at the Embassy Suites in Orlando in 1992. 300 people were there... a stark contrast to the roughly 30,000 this year.

There was the usual show floor with all the partner exhibits, lots of presentations, all the ASUG seminars and keynotes from Bill McDermott and Steve Wozniak. But my personal highlight - as it is every year - was Hasso Plattner. He still has is, so I think, the driving force behind SAP and to a very large degree it is his genius that makes the SAP story an absolute awesome success.

HANA is his latest achievement and with it he is probably breaking any barrier there is. HANA has been around for a while but his speech made me finally understand what the possibilities are. They are enormous and I see an incredible amount of excitement for it. In a few years everybody will be running on HANA - first as a database for the suite of software available today, then on the new suite S/4 HANA.

It looks like a bright future for business software and there will be incredibly sophisticated apps and features running businesses more effectively and fast... if ! and that is still the crux of the matter: you fix the process first and use the tool as it was intended. Planners still have to understand how to move through the planning horizons and when to use rate based planning or takt-based scheduling... It won`t help you to move your production scheduling to S/4 HANA if you run your lines by access database clculations today or your buyers replenish with spreadsheet lists.

Fix and clean up your old ERP first and then use the cut-over transaction Hasso presented in his speech. It´s a fantastic way to go to HANA if you´re ready for it.

Hasso is ready and covered all the bases... did you?

Make To Stock Planning done right - are you doing it?

The topic of MTS versus MTO comes up often and people have their own opinions about the definition, but it's actually a very straight forward thing... isn't it? Let's review.

The definition that makes most sense to me goes as the following:

In a Make To Stock environment production is triggered BEFORE an actual sales order comes in and any receipt from the production line is put into inventory from where customer orders are fulfilled.

unfortunately, this does not make everything clear and concise. What if I get a customer order three months before I have to deliver to it? In that case you can enter the SO and it will show up in (SAP transaction) MD04. But we're still not producing to it when we're operating in MTS. It simply reduces the forecast that we produce to and gives us additional information about what we'll actually have to deliver, but it does not trigger production.

What if I get a large order today and do not have enough inventory to fulfill it? you'll have to wait until the next period produces enough product and tell the customer to wait. Because... what is produced this period is for customer orders requesting product in the next period. This is an important point to make... look at the following example:

in the graphical display of MD04 we can see that there is a forecast for the next five months. We can also see that the MRP Run has created orders to fulfill those forecasts. Not that MRP generates orders with their latest possible delivery date standing just before the forecast is due (it does assume unlimited capacity).

As we all know there is a another step necessary before we can start producing: the orders (standing all on top of each other at the latest delivery date) will have to be sequenced, leveled and scheduled into the previous period, so that we can build up inventory for next months' fulfillment of sales orders

This situation becomes visible in the next graphic. Orders were distributed and scheduled (according to the capacity situation) for the next two months. What is not displayed here is how the inventory is consumed by incoming sales orders. that is because we're looking into the future here and the system doesn't know how the sales orders come in (if you had sales orders in the system for the second months, it actually would be displayed by a red bar going down and the green line (inventory) going down as well)

Now what's really important here is that when all inventory forecasted for a period is exhausted and an additional sales order comes in (over and above what was forecasted), that sales order can only be fulfilled by safety stock or the customer will have to wait until the next period.

Do not change the production schedule to fulfill the order if the order is part of a Make To Stock strategy !

and this is where many processes and setups fail. A sales order for an MTS part triggers changes in the production schedule and the whole plan falls apart. Set a strategy periodically, know it and stick with it!

Do you know transaction MFS0? Do you perform mid-term planning?

As was discussed in another one of my blog posts, I deem it extremely important that the planning horizons are respected and specific activities are carried out in specific zones. Along those lines I'd like to refer to the mid-term planning horizon where one can execute stimulative MRP. SAP calls the simulative planning module LTP (Long Term Planning, which is a bit confusing as I would only use it for mid term planning).

With SAP-ERP's LTP you can create planning scenarios and attach different versions of a demand program to it. That way you're holding true to still only managing demand in the mid-term and once you find a fitting demand program, you can release it into operative MRP.

To plan these different versions of a demand program or forecast, a simulative MRP Run (the LTP Run with transaction MD02) is carried out and the result can be evaluated and planned in transaction MFS0. You probably never heard about about transaction MFS0 as it is hard to find any mention of it. But it is the mirror transaction to MF50 in operative MRP.

MF50 is often misunderstood and most people think it can only be used in conjunction with repetitive manufacturing. Nothing could be further away from the truth. MF50 is a tabular planning table, perfectly suited to balance supply with demand for any period you choose to look at. It can do so for detailed planning, rate-based planning (which makes it perfect to deal with in the mid term) and even sequencing. Additionally, there is a graphical planning screen available too (like in CM25)

As you are simulating in SAP-LTP you will find that everything is attached to a planning scenario (that's why you will find that object in MS04, MS07 (mirror transactions to MD04 and MD07) and many other LTP transactions. And when you look to plan as in MF50, the planning scenario with its associated demand version, it can be done with transaction MFS0.

Give it a try and have a look... MFS0 is the perfect planning transaction for rate-based, mid-term planning... but only if you use it in the context of an overall system of planning with planning horizons.

The Difference between using a Reorder Point Procedure or a PD with Safety Stock

Have you ever wondered why planners tend to rather use a safety stock with MRP Type 'PD' to buffer variability, as opposed to a reorder procedure? Oftentimes PD is used across the board and no one wants to bother with another MRP type. However, with SAPs standard configuration the safety stock doesn't serve as a planning buffer and the additional inventory ends up being dead stock.

Consider the following scenario: a safety stock of 30 pieces is set and a production order asks for 10 pieces on April 15. A deterministic planned material (PD) would replenish the inventory up to 40 pieces just before April 15. If, for some reason, the production order would increase the required quantity to 15 pieces two or three weeks before its start date, the MRP run (which excludes safety stock in the net requirements calculation) would generate an additional order proposal for 5 pieces. In that case, the inventory on April 14 would be 45 pieces... even though 40 would be plenty (25 more than needed). Simply speaking: safety stock is taken out of the planning efforts and the entire safety stock remains unused (...becomes dead stock).

A far better solution, in my opinion, is the use of a reorder point method. If we're taking the same example, we'd set the reorder point to 30 pieces and thge MRP run would only replenish if the inventory level is lower than 30. During the replenishment lead time these 30 pieces serve as a true buffer and are consumed by regular consumption. Any safety stock that is set, is only used as a signal (exceptional message) when inventory gets really low. As you can see in below graphic, this method is a true buffering strategy.

Switching your policy from a PD with safety stock to a reorder point procedure will instantly reduce your dead stock and bring average inventory holdings down. If you're concerned about stock-outs, increase the reorder point. That is much better than never dipping into safety stock.

(note: if you're adamant about using PD. you can customize SAP so that at least a portion of the safety stock is considered as a buffer for planning)

Repetitive Flow Manufacturing in a Discrete Environment

Did you perceive an oxymoron in the headline? Many people do and that is one of the reasons why so often software implementations (especially SAP) for the production environment do not deliver the expected results. The wrong manufacturing type is applied.

Repetitive flow manufacturing is an approach to discrete manufacturing that contrasts with batch production. In its core, repetitive manufacturing strives to introduce flow onto production lines or cells and tries to avoid waste (overproduction, cycle time, scrap etc.) that often can be associated with batch production. 

Just imagine the production of shower heads for bathrooms. It’s a discrete manufacturing style and you may either produce in batches – drill the holes for 1000 spray plates first, then attach cover plates to these 1000 spray plates and then the hoses – or repetitively in that exactly one spray plate gets drilled and then it ‘flows’ to the cover plate attachment and then flows to the hose installation (while at the same time another spray plate is drilled). In the first case you’ll end up with a discrete production order to drill 1000 spray plates, another production order for the attachment of a thousand cover plates and yet another production order for a thousand hose installations.

In the latter case, however, you’ll end up with one! order to run and manufacture 1000 shower heads whereas spray plate, cover plate and hose installations are executed in a flow-like manner, continuously over a specific period of time.

There is much more detail to this but it becomes obvious very fast that repetitive execution (and planning) in a discrete environment is a much better choice and brings about many benefits (product costing instead of discrete order costing, order management and associated reduced steps, reduction of waste according to lean principles, enhanced transparency etc.)

But let’s not get too excited. Repetitive is not for every discrete manufacturer. If, for example, you are manufacturing custom railroad turnouts and every turn needs to be engineered from scratch, you will have a hard time flowing these products through a line. But that is quite alright! You want to treat and cost each one of these orders from the customer separately.

Do you really want to do that with 1000 shower heads?

Scheduling Levels for Integrated Long, Mid and Short Term Planning

Sometimes I wonder how much flexibility, great thought and opportunity for business improvement was packed into SAP-ERP. Yes, it's not the new, flashy poster child of SAP, but it has an enormous amount of functionality and features. the key is to find this stuff and apply it correctly. Just last week I was experiencing one of those magic moments when a certain function, which I didn't give the attention it deserves, makes perfect sense to solve a problem most people think SAP is not good at solving. The function helps with planning in the long, middle and short term as I describe in a previous blog Respecting your planning horizons. 

SAP ERP provides you with the ability to plan, sequence, level and schedule on three different levels. These levels somewhat correspond to the level of detail used in the horizons long-term, mid-term and short-term. For whatever reason, these scheduling levels are not very often used to their full potential but they provide excellent features to move your plan through the periods and fine-tune it along the way. You can, as an example, use different task lists for the long term as what you’re using for the short term. Also, you may decide to perform mid and long term scheduling period and rate based,  whereas you’re planning your capacity in the short term to specific dates and hours, even minutes.

This poses some interesting opportunities for your planning efficiency. As most people stick with discrete routings throughout all planning horizons, there is the great possibility to use a rough-cut planning profile for the long term, rate routings in the mid-term and for detailed planning in the short term we can determine exact, planned execution times with a routing or recipe.

If you do this, you’re effectively planning rates and periods (like 20 pieces for August and 5 pieces for week 25) for when specific, date based planning is too far out and you are planning your order’s capacity load onto a very specific point in time with its specific output quantity for the next – let’s say – four weeks. Remember that the short term planning horizon does not equal the frozen zone. In the short term you’re still planning with planned orders, whereas the frozen zone only contains released production orders.

Task lists are assigned to a planning horizon in the production version of a material as shown in below screen shot

In above example a discrete routing (with a group counter 1 out of routing group 500000002) is assigned to short term, detailed planning. Interestingly, a rate routing has been assigned to rate-based planning which effectively describes the mid-term. Often, one comes across the notion that rate routings are only used in repetitive manufacturing. However, that is not true as rate routings provide an exquisite instrument for period and rate-based planning. They describe a production process with an output rate from the operation (quantity per time) other than a discrete routing which uses the opposite: the time it takes to produce (time per quantity) a lot size. Most people would agree that planning period and rate based in the mid-term, more closely reflects the actual business process and allows for better planning results and manageability.

The settings in above figure also suggest that for the long term (rough-cut planning) a rough-cut planning profile is used. Rough-cut planning profiles are, similar to rate routings, also meant to plan for rates in periods. But that may be the only similarity.

You define your scheduling levels, and what happens to them, in customizing for planned order scheduling. This customizing table has its own transaction code – OPU5. An example of how the scheduling levels can be set up you can see in the screen shot above.

In this example, planned orders (order type LA) in plant 1000 and maintained with production supervisor 101 (production scheduler in the ‘work scheduling’ screen of the material master), will be scheduled and loaded with capacity requirements  for the short term (Detailed Scheduling) and the mid term (Rate Based Scheduling). Note that it is a choice that we use detailed scheduling for the short term and rate based scheduling for the mid term. The decision is done by the assignment of a rate or discrete routing and the type of planning table we will be using (period or date based planning table.

No scheduling or capacity records are generated for the long term (rough-cut planning)

.

You can see the results of these choices (settings) after you run the MRP Run with lead time scheduling. In the generated planned order you will see separate tabs for each scheduling level that was planned. These scheduling records can then be used in the respective planning horizons to plan capacity with the appropriate detail and time frame.

So what's the point in all of this? I think that using scheduling levels can elevate your planning to the next level. Given that you are respecting your planning horizons and you operate in a long, in a middle and in a short term planning horizon (and you also execute in a frozen zone), you can now control the level of detail you're planning with. But more importantly, using a rough-cut planning profile for the long term and a rate routing for the mid-term - and using the appropriate planning table, you can plan periods and rates where, in the past, you might have killed yourself planning exact dates and quantities. 

As an example... at my current client we use rate routings for the mid-term and the resulting rate based orders are scheduled and capacity leveled in the tabular or graphical planning table. In fact we are doing this in Long Term Planning in a simulative mode using transaction MS05. That way we can look at periods and move the rate based orders without having to deal with all the unnecessary detail of a detailed, discrete routing. According to the need to respect the planning horizons (as described in a previous blog post), we manage demand in the mid-term and capacity level simulative planned orders until we find the perfect demand program (as defined by the Planning Scenario in LTP). Once done, we hand over that demand program to MRP so that, by use of a discrete routing and the respective detail scheduling level, exact dates, times and quantities are determined in the short term.

The beauty is, that because of our upfront work in the mid term, the demand program should fall pretty well into the available capacity profile and there shouldn't be too much work in terms of detailed scheduling to do anymore. 

Using scheduling levels might not get you to the perfect system of planning for your capacity and customer delivery service, but it may be a great step forward to a more integrated system of handling and managing your orders (using standard SAP-ERP software functionality). 

In any case, you will have to be clear about where you are planning, what you are planning for and it what timeframe you'd like to use what level of detail first...

...or like Yogi Berra said: "If you don't know where you are going, you might wind up someplace else" 

Materials Planning with SAP... available 11/30 on amazon.com

Maybe you read my blog here or maybe our paths crossed at some point in time. In any case, if we ever discussed the subject of SAP in general or Materials Planning in particular, you know that I have a strong opinion on the way people implement and use it.

More often than not an organization drops functionality and transactions on to the planner, scheduler and buyer - providing minimal (transactional) training - only to expect far better results than ever before ("hey, we spent all this money for you to have these great transactions"). I do not believe that this is working out well and think it might cause even more frustration than there was before.

That is why I wouldn't rest until I got the chance to put my thoughts down in form of a book. Thanks to Rheinwerk Publishing (formerly Galileo), Kathy Spencer, Hareem Shafi and specially Emily Nicholls I was given the great opportunity.

Because I did not want to write a technical 'transaction' book, this might be a bit different from what you expect from a typical SAPPRESS publication.

The book starts shipping November 30 and you can pre-order at amazon.com (http://www.amazon.com/Materials-Planning-SAP-Uwe-Goehring/dp/1493211978/ref=sr_1_1?ie=UTF8&qid=1444479768&sr=8-1&keywords=uwe+goehring)
Critique and suggestions are most welcome and I deeply hope that it can make a small difference to the better in your own organization...

Are you respecting your Planning Horizons?

SAP software provides excellent planning functions and capabilities. But before you can effectively use them one must define the planning horizons... some people, as I have seen often, do not necessarily mind if they're in the long, middle or short term for what transactions and tasks they perform.

As an example, it might happen that a scheduler turns a planned order into a production order several weeks or even months before production of that order starts. Or the forecast is worked by the MRP Run (and subsequently planned orders are generated) way beyond the short or even middle term.

At my current client, we have worked on a definition and rule set for planning, which I would like to share with you, so you may use it as a frame of reference if you find it useful.

Below graphic summarizes the concept and it must be said that there are two major rules valid in any planning system:

Rule of Planning #1: "There is a point in time after which planning activities end". This point in time is not today! It comes before today, exactly at the point where the frozen zone begins. Once you're in the frozen zone you are working with production orders. And production orders are supply elements that we are not planning with anymore. we're expediting on them, reschedule them, re-route operations to different work stations and react to exception messages they received from the MRP Run when actual results differ from the plan. If you find yourself looking for a tool to automatically re-assign and re-schedule within the frozen zone you either don't have a frozen zone or you're under the wrong assumption that the planning system should not only 'plan' but also fix deviations from the plan. These deviations are due to variability. And variability can only be buffered but not be planned. Especially not after it occurs.

Rule of Planning #2: "To plan your resources in the mid and long term level and manage demand - To plan your resources and sequence in the short term level, schedule and manage supply". It doesn't make sense to reshuffle planned orders in the long term. You shouldn't have any in the first place. In the long term you are working in SOP and therefore with a planning hierarchy, monthly demand figures and SOP orders that cause rough capacity requirements (and not detailed capacity requirements). In SOP you move the demand so that capacity violations are resolved. In the mid-term you should work with Long Term Planning (LTP) and its Planning Scenarios. A Planning Scenario contains a demand program which you can simulate (with simulated planned orders) for mid-term capacity planning. Here you should also work with the demand program until you find one that generates simulated planned orders which fit into your available capacity program. That is then the demand program (Planning Scenario) that you activate into the short term. Now you can run MRP on those Planned Independent Requirements and the resulting planned orders can be sequenced, leveled and scheduled in capacity planning. The latter activity was 'managing and scheduling supply' whereas the previous activities were concerned with 'leveling demand'

Those two rules we have persistently respected and followed in our efforts to build a standardized and integrated planning system everyone in the organization is using and looking at for continuous improvements.

The sales planners enter their forecast into product groups within a planning hierarchy. SOP Orders, statistical work centers and rough cut planning profiles are used to analyze the capacity situation for a horizon of 18 months out up to 5 years. Should we encounter a problem, we're moving the entire product group demand into a previous, less capacity constrained period or fill out a request for more capital expenditure to increase capacity and meet increasing customer demand.

When the leveled demand is dis-aggregated from the product group level to the actual product, we then transfer the demand profile into Long Term Planning (LTP) where we simulate various demand programs and generate stimulative supply. Requirements are determined for long lead time items and the procurement process is started if necessary. During Phase 1 of the mid term - 12 months to 18 months out in this example - we let demand changes from the SOP flow in and integrate these into the demand programs. In Phase 2 of the mid term, we perform detailed capacity planning with stimulative planned orders and find the best demand program that fits into our available capacity.

The activation of the demand program (transfer of Planned Independent Requirements into MRP) indicates the move from the mid term to the short term planning horizon. After MRP is run, planned orders are generated which we can sequence, level and schedule within available capacity on the bottleneck work center. 

The last planning activity then is to take all leveled and sequenced planned orders of, say, one week and perform a collective material availability check. Now you have ensured that all materials and the capacity is available for all the orders to be executed and you are ready to move these orders into the frozen zone. This is done by collectively converting the planned orders into production orders for the next week. From then on - within the frozen zone and into backorder scheduling - all planning has stopped. Anything that happens against the plan needs to be adjusted manually. If a work center is down, an alternative work center will have to be found and the sequence in the work order changed manually. 

This last point is especially important as I often get asked if one can automate this function. But in my personal opinion this is an impossible proposition. To actually do so one would have to build all possible cures to an exception into the basic data (production versions, alternative BoMs or routings, etc) and you can not possible do that. It is much better to have someone who is close to the exceptional situation pick an alternative and just change it into the order. 

...after all, that is the whole reason why we';re comparing 'actuals' to the 'schedule' and the 'plan'

trip report (and some personal views) from SAPPHIRENOW, Orlando May 2015

This past week I attended SAPPHIRENOW and ASUG, SAPs annual users conference in Orlando, FL By now quite an impressive event, it used to be a much more subdued meeting place for customers, partners, users and consultants. I have no idea how many people were there this year but it must have been close to the record attendance of some 20,000 in 2013. That's in stark contrast to the 500 or so that I mingled with at the Embassy Suites in the early 90's.

SAPPHIRENOW is an absolute mega event with a humongous show floor, demo labs, larger than life video screens and a demo area within which you could see how an airport or a harbour will manage their traffic in the future. All with gadget gallore... Apple iWatch, multi touch screens, huge displays etc etc. There is an abundance on demo booths, partner displays and HANA, HANA, HANA and HANA. It seems like everything is possible with HANA. As most of you know HANA is in memory computing and makes everything much, much faster and therefore opens possibilities never thought of before.

To me it looks like SAP's strategy is to move everything (and every customer) to HANA so that everyone is equipped with a basis on which modern information technology is limitless.

I like it... but what's in it for YOU? the customer! Hasso's number one priority... remember that?

Yes, there is S/4 HANA, the new ERP. Bill McDermott tells every employee, every partner and every consultant that the customers need to go to S/4 HANA. The promise is great, but sofar S/4 HANA only includes Simple Finance. I can't tell my customers to go and do that. What will ABB do with Simple Finance? And yes again, there will be development and sooner than later S/4 HANA is supposed to completely replace ERP. However, SAP already claims that the functionality will not be the same as in ERP. Its limited and there is much less standard functionality. "Are we moving backwards?", I ask. "No", is the reply, "most of the functionality will be developed by third party consultancies who provide the customers with perfectly customized functionality and solutions.

SAP as a platform instead of a standard software package? Maybe so, but when I think about that I am wondering how expensive that is going to be for the end user, the ultimate client (or has that changed? is the ultimate client still the end user??). First you pay for the software (platform) license and then you have to pay for the development, customization, process engineering. "So what's different from paying for the costly implementation?", you might ask. Think about it... All these complex transactions, processes and routines are not there any more (or to a limited degree). Do you really think that all those third party consultants (who weren't able to perfectly implement SAP software during the past 30 years) will now develop better solutions for you than SAP has done in the past? You think that those people who never understood how takt-based scheduling works in ERP will come up with a leaner approach to your repetitive lines? Do you believe that technologists who understand HANA will be able to help a materials planner develop effective replenishment policies AND develop the transactions that are necessary to monitor exceptions?

Only time will tell, but what's worrying me is that the customer does not really have a lot of options to choose from and Bill McDermott has different priorities today than Hasso Plattner did have in the late 90s.

Maybe I am now one of those old people that I thought didn't understand what was happening when I was young... maybe those HANA people will change the world. I only hope its different from what facebook, uber, whatsapp and google glass or apple watch consider valuable and revolutionary.