Date: 12.10.2025 (2025-10-12)
Paul Brandstaetter (12.10.2025)
• What data do they have to usually pull? and from which systems? What are the challenges? Who delivers the data?
• They make data request -> form -> specs what data is required -> either it is centralized or decentralised
• Client: they sent the requests to all 56 sites -> sites exported the data ERP
• Form: not super complex
• E.g. all invoices, all employee data, salaries
• Normally standard data requests - later one or two more requests based on insights during process
• Challenges:
• Wrong format is a big challenge -> they sent a template
• Delay in transfer of data
• The company has to do
• Other client: they had a central system -> they were able to deliver all data
• Invoices are one line in Excel -> structured -> no original PDFs
• Sometimes they have to categorise
• Currently they want to analyse colours -> they need to understand which code means what -> employees have to do it
• They get a lot of Excel -> they merge it into one Excel file and then do the analysis in Excel mainly - they also have Power BI
• “SpendCube”:
• Inhouse GenAI tool to categorise invoices (works for 80%)
• E.g. Head of Spend -> sanity check with stakeholders -> sometimes new categories -> aim for 90 to 95% accuracy (at Client: 3 to 4 people for a few weeks did that at Client -> they had to go in line and review 1000s of customers to recategorise -> a lot of sloppy mistakes -> Schlampigkeitsfehler because multiple people work on an Excel files)
• Line item base -> e.g. PO every line -> Invoice is a summary of a month e.g.
• Data from ERP are for the analysis - apart from that rather qualitative data - ERP has most of the data that is relevant
• Supply chain data comes from other software systems -> tool that visualises the supply chain (all that SW is built by Data & insights team) -> all these tools are not really good - not 100 people that are focussed on one product - rather project
• They have a longtail of mini supplier
• What software tools are they using? (e.g. Celonis, Palantir)
• He never used an external tool -> its not common
• Leapfrogs -> they can use external tools / its allowed
• Which internal tools do they have?
• SpendCube
• Can be bought by customer after the project
• They benchmark categories with other companies (spend, # of employees)
• APQC und BME are used
• Own database / know how as well
• Based on that data they decide what to focus on
• Then more data and discussion with the customer
• SpendGuard
• Checks which invoices are needed -> if not / in doubt escalation
• Can be bought by customer
• Hosted on their server
• Can be deployed in a way where the customer can create criteria when an invoice has to be approved by someone higher up
• Process:
• They talk to the site managers who already shares hints
• Then they do benchmarks with apply (tool)
• Insert: route, day, volume > what are costs
• Customer tracks deliveries and delays (they have that in SW)
• Often logistics has really good data in online tools (e.g. Maersk)
• Then world map Datenströme
• Tool calculates where to put a warehouse and where bottlenecks are (very complex -> he never did it himself) and then when you want to build a warehouse you have to manually calculate the costs etc.
• Client often already knows where problems are but do not have the data to prove and to fire people (enablement is a big topic)
• zum SCM is eben ah nu viel abhängig vom qualitativen feedback von so Supply chain manager
• owa glaub do gibts ah scho so richtig häftige Big data tools wos da des richtig mathematisch lösen usw
• hom ma owa ned
• Contract Screener -> summarises contract and you can query
• Invoice Screener -> summarises an invoice and you can query
• SupplyChainRadar -> you add data and then looks how supply chain can be optimised
• What data are they mostly working with? How do they manage raw data (e.g. invoices) if the data is distributed not properly structured? e.g. invoices are not categorized.
• See above
• Generic: what is the biggest pain point for them in projects?
• SpendCube -> categorisation of spend takes a lot of time
• Zwischenmenschliche Themen: people think they are replaced
• Implementation:
• FTE: Identify too many people in an department -> stakeholder gives OK -> send surveys to manager -> they have to fill out what are tasks of every employee -> they analyse -> they decide which employee is redundant
• They do it in waves -> SG&A (not Pauls team)
• Colours: supplier is too expensive -> RFQ / tender -> negotiate with suppliers
• BCG does it - they have pro negotiators
• For smaller topics they have tutorials -> enable employees
• Supply chain: you see the routes / bottlenecks -> other route or new warehouse or source from somewhere else (street vs. ships)
•
• All these topics are discussed in steer committee
• Paul does Supply Chain (Inverto), the others are BCG: Sales does topline, Ops does processes / assets / machines, SG&A
• A lot overlap: e.g. colour simplification is Supply Chain / Sales / Ops
• Coordination of who gets the savings assigned
• Which processes / topics at customers are the biggest challenge for them to do their work? (e.g. data situation, etc.)
• See above
• Track: API as a Service
• N/A
• How to enable consultants to create recurring SaaS revenue?
• It would require a continuous synch or invoices for SpendGuard
• Why do they not use AI yet?
• N/A
• Hypothesis: when it comes to process automation thinking in functions makes more sense vs. thinking in industries
• He thinks: 100% agree -> departments use the same tools independent of the industry
• Product ideas:
• Enable to constantly copy data from legacy system in certain format to other software
• See above
• It is important that employees pflegen the data
• Invoice → PO → Contract product
• Get data from Sharepoint
• He thinks that invoices come by email but he is not sure -> then they have to put it on sharepoint and then assistant has to put it into system (he is not sure)
• Other:
• They make a lot of saving via renegotiations with suppliers / tenders - he does not know what the other teams mostly focus on
Maersk Insight:
• Integrated Supply Chain Engine
• Tracking of their containers - rail road binnengewässer -> other companies take over - other startups tracked the data (e.g. Project44 -> data aggregator) - Maersk aggregated the aggregators plus own data
• Goals: live update data cross the world
• Then they wanted with data adapt if there are crisis / accident etc. -> customer can adapt route -> rerouting incl analysis of costs, time and CO2 footprint
• Challenges:
• a lot of SW which is old; nobody changes SW; a lot of Excels; prone to error; key person sick -> SC resilience