THANK YOU FOR SUBSCRIBING
As long as I can rememberworking in Supply Chain, BI tools have been my best friends. Withoutthem, impossible to performspecificanalysis on new issues, add a KPI, specialreport to the Directors etc. Actually, in all the industrialcompanies I have been working for, weweredependent on BI to run dailyoperations. Whyisthat ? I have 2 examples in mind.
For one warehouse in charge of preparingthousands of orders per day, weweregivenprecise instructions by the VP of merchandising to manage productshortages (particularlyduringproductlaunches), takingintoaccount country, margins, store size, store ownership… There wasabsolutely no waywecouldexecutethat in the WMS withoutmakingmistakes. So, with the sharpminds in the team wehaddevelop analgorithm to extractorderslinesfrom the ERP wheretheywereplaced, downloadthem in PowerQuery, apply the priorityrulesweweregivenbefore re-injectingthem in the WMS. Dangerous ? Absolutely. Wecould have made hugemistakes. However, wewere able withthismethod to execute business priorities in servingourcustomers and own stores.
In charge of supplyingrawmaterialsfor severalfactories, I didn’twant to use standard MRP functions in SAP, generatingtoomanyalerts and changes. I decided to use automaticreplenishmentbased on thresholds. And guesswhat, to set up the thresholds I built a alrogithm in Excel (based on pastvariability of consumption, delivery lead times, MOQ etc..). And hereagain, extracting data from SAP into Excel, running the model, tweakingit to adjust the targets, and re-importingitinto SAP became a bi-monthly routine. Withthis I reducedinventoryfrom 9 to 2,5 months of coverage and neverwent out of stock.
I can rememberdozens of situations like this, when the Supply Chain teams had to be smart, and under the radar of IT and finance used BI or desktop tools to significantlyimproved Service or inventorylevels. It evenstartedin 1993, when I startedmycareer at Renault. Renault wascleverenough to have developped a BI environmentwhere production data couldbeaccessed by the SC teams to build reports and KPIs. I realizedthenthatlearning SQL could help me tremedously in mywork. With the years Ilearned how to use MS Access, then Pivot tables, thenPowerQueries, etc and then I was not cleverenough and delegatedthesetasks to ourprecious BI teams, workingwith more and more advancedtools.
Now start the questions :
Whycan’twe have thesequeries and optimizationpossibilities in our APS ? APS are done to optimizesupplychains, right ?
To whomshould the BI team report to?To IT or in Supply Chain ?
Does the rise of Data Science change what I have experienced for more than 25 years ? If yes, in which direction, and whoshouldactually master Data Science techniques ?
Here are mytentativeanswers to these 3 questions
What about thesefancies software ? I believeany one whoalreadywentthrough an APS implementation process has seentheir limitations. What are those ?
Speed. 12 to 18 monthsis a minimum for implementingan APS. Duringthis time, chances are thatyou business has changed. The resultyougetis not coveringyour new needs… not even sure theywereactuallyoveringyour initial needs.
Resources. Duringthese 12 to 18 months and of course afterimplementationisfinished, you have to mobilizeyourmostexperienced and brightest people. Theirtasks : understand the software, describe the actual process, constraints… in a wayit can bemodeled by the software, extract and clean data, tests, retests, train, understand the limits of the software. In otherwords, youneed to mobilizeyour teams on new problems… beforestartingsolving the problemsyouhadbefore. Instead of workingon optimizingyourSupply Chain, theyoptimize the model, the software.
Costs. For implementingan APSyouneed to spend money on the the licences and support / maintenance for the software, on the consultants helpingyou to overcome the complexity of the project, on the IS/IT integration, on training… Don’t forget3otherthings : opportunitycosts (whatyoucould have donewith the same money, on BI solutions for instance…), the time and efforts to actuallyget the fundingfromyour CFO and general management… and the time youspendprovingthatyouactually do betterwith the new tool.
Yes, extremdifficulties to provethatyou are actuallybetter off with the systemsthanwithoutit… especially if youconsideryoucould have spent all this money differently.
Fixed solution : APS vendorswill tell youtheir solution issuper flexible and can perfectlydescribeyourSupply Chain and help youoptimizeit.Frommyexperience, all SupplyChains are unique.Whatmakesyourcompany’soffer unique isalsowhatmakesit’sSupply Chain unique. Integration of the value chain, diversity of suppliers, markets, customers, constraints, opportunities… make a complex system whichisabsolutely unique.
Data : most APS are quiterigid in the waytheyrequire data. Don’t forgetthatyourSupply Chain is visible onlythroughyour data. How manyerrors and approximations are yougoing to maketrying to fit your data into the data model required by the APS ?Only part of the data youprovideisexactlywhat the APS needs. As a result, the model of your SC (digital twin as it’scallednow) can beuncomplete and fairlywrong.
Flexibility : Once yougetyour APS implemented, don’tyouthinkyou’llneed to update it ? Tweakit ? Adjustit ? How do you run continuousimprovementthrough a fixedIT solution ? Yes, exactly, you use BI solutions on top of it, becauseyou are probablysick and tiredalready to go through the change request process to close the gap between the toolyou have and the toolyouneed.
Whereshouldyour BI team sit ?
The first question isalready : do youneed an BI team at all ? Or do youneed « BI trained » Supply Chain experts ? I have seenbothwork. The criticalsuccess factor isthatthese people nowyourSupply Chain. And as I explaineditearlier, a large part of yoursupplychainisactually the data coming out of your ERP, orderingsystems, WMS, TMS, APS… It meansthat the team / the experts need to perfectly know how and wheneachpiece of data isgenerated. Whatdoesthis date mean ? This quantity ? A IT expertise isdefinitelyneeded in building the back-end of the data cube, lake… Data must be made quicklyavailable in a semi-structuredformallowing fast queries. At the time youbuildthis cube, youdon’t know whichquerieswillbedeveloppedlater on… becauseoptimizationneedswill come onlylater on. There isactually no limitas youdon’t know whichideasyou’ll have in 6 months, whichrequestyou’llgetin 12 monthsfrom top management to improveyourprocurement, production or sales. That’s the all point. There isonly a short list of predefinedqueriesthatyou can run everydayfor the next 3 years. Most queriesgetoutdated super fast as your business changes. I’m sure Consultants can come up with 5 maturity stages on BI solutions and teams set up(and you’llprobablysitbetweenmaturity stages 1 and 2 sothatthey call bill youhours to help youreach the next stages !). I’malso sure you’llneed to fightwith IT to keep BI competencies in your SC teams, or in direct control of your team.
And by the way, how do BI solutions compare to APS ? In myview, the do betteron all aspects : costs, speed of implementation, adaptation to the uniqueness of your SC, focus on actuall issues and opportunities vs creating new issues, and mostimportantlythey support continuousimprovement, as long as they are run in an « agile » way.
How does Data Science change the picture ?
MyviewisthatprogressData Science makes APS evenless relevant.
Data Science bringstwobenefitsto SC professionnalson top of the current BI strengths.
a/ Possibility to manipulateevenlargerunstructured data sets
b/ Sophisticatedstatistics. I willdevelopthisbenefit, thatmost APS actuallydon’tbring.Mastering a Supply Chain consisst in reducing, anticipating and absorbingvariability. The waywe have doneit for the last 25 years has been quite « empirical ». I will not developfurther the limits of thoseapproaches (one is for instance the used and abusedassumptionsthatmostvariabilityis « normally » distributed, which, we all know, isquitewrong) but willinsist on how modern « data science » actuallyfits to SC needs in the sensethatData Science tools are masteringanalysis of variability.