Consent gaps
ECC has no consent model. It holds identity (KUNNR, address, e-mail), communication flags and blocking, but none of what a consent check needs. These are the 12 gaps, with the ECC fields that come closest. The CRM Sync gap fill chapter shows what fills each one.
The rule that follows from the table: keep consent out of ECC. ECC stays the source of identity and blocking; consent lives in an append-only ledger in Xano that ECC and batch jobs may read but never write.
| Gap | What ECC has | Closest ECC fields | What’s missing |
|---|---|---|---|
| Consent per purpose (analytics, ads, sale / share, email, SMS) | At most a custom ZZ_ flag on KNA1. ADR6-FLG_NOUSE means “don’t use this address”, not “no consent”. | KNA1-ZZ_MKT_OPTINADR6-FLG_NOUSE | A record per purpose: one flag can’t say “email yes, ads no”. |
| Who decided, how and with which banner | Change documents, but only for fields flagged for change documents; custom ZZ_ fields usually aren’t, so a flip leaves no trace. A batch user shows up as the author. | CDHDR-USERNAMECDHDR-TCODECDPOS-VALUE_OLDCDPOS-VALUE_NEW | The source, banner version and actor stored with each decision. |
| An exact, comparable time | DATS dates have no time zone; CDHDR UDATE / UTIME are server local time. | CDHDR-UDATECDHDR-UTIME | UTC timestamps and a sequence number for events in the same second, so the newest row reliably wins. |
| Where consent was given, and under which regime | LAND1 / REGIO are the billing address, not where the shopper was when they decided. | KNA1-LAND1KNA1-REGIO | The region (ISO 3166-2) and regime (US opt-out with GPC, EU / UK opt-in, KR PIPA opt-in) at decision time. |
| Consent before there is a customer | Nothing: a visitor has no KUNNR until they order. | (none) | Consent keyed by a visitor ID and linked to the customer later. |
| Consent for one-time customers | One-time customer accounts share a single KUNNR across many real people. | KNA1-KUNNR | A key for the person, not the shared account. |
| Consent for B2B contact persons | Contact persons (KNVK) hang off the company record. | (none) | Consent per person, not per company. |
| A withdrawal that sticks | A DEBMAS IDoc or a 15-minute middleware job rewrites whole segments, so a stale ZZ_MKT_OPTIN = X overwrites a withdrawal. | KNA1-ZZ_MKT_OPTIN | An append-only ledger; batch jobs read consent and never write it; withdrawals pushed as events. |
| Data subject requests: access, deletion, opt-out of sale | Later enhancement packs add end-of-purpose blocking and ILM deletion, but no request tracking or receipts. | KNA1-LOEVM | A request record with a due date and a receipt from every system it fans out to. |
| Age or minor status | No age field. | (none) | A minor flag where the law requires one. |
| Retention per purpose | Retention is per object (ILM), not per purpose. | (none) | An expiry per purpose. |
| Permission to use a person for ads and audiences | Nothing ties a customer to ads permission. | (none) | ad_user_data and ad_personalization checked before anyone is exported to an audience. |
On a real system, check first whether any custom consent field is flagged for change documents (its data element’s change-document setting). If it isn’t, there is no record today of who changed it or when.
The /ACME/CONSENT table in this book is the shape the ledger takes if it must live in ECC; the Xano ledger in the next chapter adds the columns ECC can’t supply.
Does S/4HANA close the gaps?
Partly. An upgrade to S/4HANA improves the data-protection mechanics, but consent and the global state at decision time still sit outside the ERP core. SAP does sell consent tooling; it is a separate product the ERP receives by integration, which adds one more copy to reconcile rather than one record of truth.
What S/4HANA improves
| ECC gap | In S/4HANA |
|---|---|
| Several customer masters (KNA1, KNVK contact persons) | The Business Partner is mandatory: one record per person or organisation, with relationships for B2B contacts. This narrows the contact person gap. |
| Data subject requests | ILM blocking and deletion at end of purpose, Read Access Logging, and information retrieval to collect one person’s data across objects. |
| Batch-only interfaces | APIs and events exist, though most estates still run IDocs and middleware batches. |
What is still missing
| Gap | Why S/4HANA doesn’t close it |
|---|---|
| Consent per purpose | No per-purpose, append-only ledger in the core: still a flag or a custom field on the Business Partner. SAP’s consent products keep it in a separate system. |
| Region and regime at decision time | The Business Partner holds an address, not where the shopper was, under which regime and with which banner. ERP is not on the path of a browser session. |
| A withdrawal that sticks | If middleware still rewrites whole master-data segments on a schedule, a stale opt-in still overwrites a withdrawal. A new ERP doesn’t change the batch. |
| Retention per purpose | ILM retention follows the business object and its legal hold (tax, commercial), not the purpose of consent. |
| Consent before there is a customer | A Business Partner must exist first; visitors and pre-login consent are out of scope. |
| Audiences | Nothing in ERP checks ad_user_data and ad_personalization before an audience or a BigQuery segment is exported. |
| Residency across regimes | One global instance in one region doesn’t give per-regime storage for EU, UK or KR personal data. |
Why enterprises struggle with compliance. Consent ends up copied across systems that each own something else:
Consent banner / CMP (session) ─┐
Commerce platform ──────────────┤ each holds a copy of "consent",
Marketing cloud / CDP ──────────┼── refreshed on its own schedule,
S/4HANA Business Partner flags ─┤ none on the path of every decision
Warehouse and ad exports ───────┘
- No record wins. When the banner, the CDP and the ERP flag disagree, no one can show which was in force when an export ran.
- Batch delay is a compliance gap. A withdrawal that takes hours to reach an audience job is out of compliance for those hours.
- Ownership is split. Legal owns the policy, marketing the CMP, IT the ERP, data teams the warehouse; no one owns the decision.
- Several ERPs and regions, after acquisitions, multiply every gap above.
- Upgrades treat consent as migration scope. The fields are moved; the decision logic is never designed.
The rule holds through the upgrade: ERP is the source of identity and blocking; consent lives in an append-only ledger beside it, and every export checks it live through the release gate. That is also why the gaps can be closed without a full upgrade: see An alternative to a full S/4HANA upgrade.
The S/4HANA features named here vary by release and by which SAP products are licensed. Check them against the system in front of you: where a consent product is in place, the first gap becomes “a consent store outside ERP that must be reconciled”, not “no consent store”.