A QA specialist checks a tablet beside a blister packaging line in a bright pharmaceutical packaging hall.
S/4HANA · Pharma and biotech

What a pharma site asks in the first 30 days of S/4HANA

Pharma S/4HANA programs usually pass validation on schedule. The first month after go-live then depends on a few hundred practical questions from QA, the lab and the warehouse, and on whether someone approved the answers in advance.

Every pharma company still on SAP ECC has the same date in its plans. Mainstream maintenance for ECC 6.0, enhancement packages 6 to 8, ends on 31 December 2027, and extended maintenance to the end of 2030 costs extra. Over the next three years, plants, labs and distribution centers will move to S/4HANA in waves, one site after another, on a global template.

In a GMP company each wave runs on two plans. The validation plan proves the system does what it should, and auditors read it line by line. The people plan gets the QA specialist, the lab analyst and the warehouse lead through their first shift in the new system. In most programs it consists of a training matrix and a hypercare hotline.

The questions employees ask in the first month show where the people plan works and where it doesn't.

How long ECC is still maintainedSAP ECC 6.0, enhancement packages 6 to 8. The black line marks October 2026.

What you will read

Readiness

“Trained” has two meanings in a GMP go-live

GMP rules put training first. The US rule for drug manufacturing asks for training in the operations each person actually performs, and EU GMP Chapter 2 expects the same.

“Training shall be in the particular operations that the employee performs.”21 CFR 211.25(a)

Every S/4HANA wave therefore has a release gate: nobody gets access to GMP transactions in the new system until their training on the new SOPs is on record. That record is easy to produce. A read-and-understood task, a recorded demo and a signature, and on cutover day the learning system shows 100%.

Whether people can do the work on the first shift is a separate question. Can the warehouse move a pallet to the line? Can QA see why a batch is blocked? Can the lab find the inspection lot the system created by itself? The answers decide whether trucks wait, batches sit in QA and deviations go up in the first month, and few programs measure them.

Trained

A training record exists for every person and every new SOP.

Checked by
Auditors and the release gate
Measured by
Learning-system completion
On go-live day
Usually 100%

Ready

The person can do the task in the new system on the first shift.

Checked by
The first night shift
Measured by
Questions, tickets, workarounds and deviations, per group and shift
On go-live day
Usually unknown

Reading by group

Five groups ask most of the questions

Change does not land evenly across a site. In a typical pharma S/4HANA wave, five groups carry most of it, and each asks a different kind of question in its first week.

QA, batch release and deviations

What changes

Batch status, usage decisions and release checks move to new screens and roles.

“Which status means this batch can ship?”

Batches wait in QA and release dates slip.

QC laboratory

What changes

Inspection lots appear automatically at goods receipt, and results go into new screens.

“Why is there an inspection lot for this material now?”

A testing backlog, and material blocked for the line.

Warehouse and logistics

What changes

New stock types, storage bins and handling units, and different postings.

“Why can't I move this pallet to the line?”

Trucks and lines wait, and workarounds appear.

Production supervisors and operators

What changes

Process orders, confirmations and goods issue work differently.

“Do I still confirm each phase, or only at the end?”

Wrong consumption, corrected weeks later.

Master data and site finance

What changes

New material types, plant settings and cost objects, with new owners for fields.

“Who owns this field now, the site or global?”

Errors that travel to the next sites.

A lab analyst in a white coat reads a message on her phone by a laboratory window.

Typical pattern at a manufacturing site on a global template. Groups and questions differ from site to site.

The questions are concrete: which status, which screen, who decides. They also repeat. The same few dozen come back across three shifts and two sites, and when the answers sit in a ticket queue, the second site asks all of them again six weeks after the first.

A cleanroom operator in a gown works alone at night in a pharmaceutical plant.

The questions nobody types into a ticket

“Will the new system show every mistake I make?”
“Will the warehouse need fewer people once it runs?”

People ask these in the canteen and on the night shift, rarely on the hotline. They need an approved answer too, written before go-live and worded the same way at every site.

In their words

What people say in the first week

I could release a batch in the old system with my eyes closed. Now I check three screens and still call the key user.
Claire, QA specialistPackaging site
The training was a video and a long SOP. On the night shift there is nobody to ask.
Marek, warehouse team leadDistribution center
Nobody told us the inspection lot would appear by itself. We thought it was an error and raised a ticket.
Anna, QC analystSterile manufacturing site

Composite quotes, built from questions typical in the first week of a pharma S/4HANA go-live. Names and roles are fictional.

One situation, told twice

Day three after go-live, 05:40, at a packaging site

A night-shift warehouse lead checks his tablet next to a shrink-wrapped pallet of printed cartons.
Changewiser AI assistantIllustrative
Why can't I move the cartons for line 4?
The cartons are in quality inspection stock until the lab records the usage decision. You can check the status in the stock overview. For urgent line supply, call the QC on-call number in section 4.
From work instruction WH-014, approved by QA
Site lead · this week
14warehouse questions about stock status, mostly from the night shift

To avoid

The question goes to a queue

  1. Line 4 needs a pallet of printed cartons at 06:00. The system won't let Marek move it, because the stock is in quality inspection.
  2. He calls the hypercare hotline. It opens at 07:00, so he leaves a ticket.
  3. A supervisor moves the cartons with a workaround so the line can start.
  4. QA opens a deviation.

Two more shifts ask the same question that week. Each becomes a separate ticket, and the pattern shows up in the weekly hypercare report.

A better approach

The question gets an approved answer

  1. Marek asks in the chat tool on the shift tablet why he can't move the cartons.
  2. The answer comes from the approved work instruction and names it. Urgent line supply goes through the QC on-call number.
  3. The on-call analyst releases the material and the line starts, without a workaround or a deviation.
  4. The site lead sees 14 warehouse questions about stock status this week and asks the key users to brief all three shifts.

The work instruction gets a clearer first page through change control.

Illustrative scenario. Names are fictional.

Our view

Give the people plan the same discipline as the validation plan

In a GMP company, every answer an employee gets about a GMP task should come from an approved document, and the company should be able to show which one. Answers improvised by whoever is on shift fail that test, and so do general AI chatbots. Fast answers pass it when they come from approved content.

When the questions are collected in one place, the program sees which site, group and process is struggling while there is still time to act, and the next wave starts with its answers approved.

How Changewiser works on a site

  1. Employees askIn their chat tool, on any shift.
  2. Answers come from approved contentEach answer names its source document, so the SOP stays the reference.
  3. Patterns show up, without namesQuestions are grouped by site, group and topic. Anything not covered goes to the team without the person's name.
  4. Documents get betterWork instructions improve through change control, and every approval stays in a history that can't be quietly changed.

Three things to set up before the first site goes live

  1. Write the first-week questions before cutoverFor each of the five groups, list the 20 to 30 questions they will ask in week one, including the ones nobody will type into a ticket. Have the answers approved together with the SOPs.
  2. Measure readiness next to trainingTrack questions, tickets, workarounds and deviations by group and shift for 30 days, next to learning-system completion. Use both in the go-live decision for the next wave.
  3. Carry the answers to the next siteEvery question answered at the first site goes into the approved answer library for the second, through the same change control as the SOPs.

In brief

2027

Mainstream maintenance for SAP ECC 6.0, enhancement packages 6 to 8, ends on 31 December. Extended maintenance to 2030 costs extra.

Source: SAP

7×

Projects with excellent change management are about seven times as likely to meet their objectives as those with poor change management, 88% against 13%.

Benchmark: Prosci, 2023

5 groups

QA, the lab, the warehouse, production and master data ask most of the day-one questions at a typical pharma site.

Typical pattern, varies by site

Hayk Derdzakyan

Planning an S/4HANA wave at a GMP site?

In a 30-minute session we map the first-week questions for one of your sites with you.

Hayk Derdzakyan, founder of Changewiser

Sources

  1. SAP, maintenance commitment for SAP S/4HANA and SAP Business Suite 7, announced 4 February 2020. Summary: CIO, “SAP promises to support S/4HANA through 2040”.
  2. US Code of Federal Regulations, 21 CFR 211.25, Personnel qualifications.
  3. European Commission, EudraLex Volume 4, EU GMP Guidelines, Chapter 2: Personnel.
  4. Prosci, Best Practices in Change Management, 12th edition (2023). Summary: “The Correlation Between Change Management and Project Success”.

All articles