Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Marketing Technology › The Marketing Stack

Mapping Consent Across the Marketing Stack

How to map consent across marketing tools: each purpose, where permission is captured and stored, which tools read it, how withdrawals travel and the tests.
Tutorialadvanced13 min readUpdated

Each tool in a stack handles its own slice of consent reasonably well. The consent management platform asks visitors about cookies and tags and tells the tag manager the answer. The email platform records who subscribed and honours unsubscribes. The CRM holds a do-not-contact flag. The trouble is that a person is one person across all of them, and their choices are made in one tool and need to be honoured in another.

The marketing stack hub says the stack has to carry consent state from the point of collection to every tool that acts on data. The data flow and integrations guide names what happens when it does not: consent leakage, where data collected under one permission is activated for another because the state did not travel with it. In practice the gaps look ordinary. A newsletter permission given at checkout stays in the shop and never reaches the email platform's segment rules. An objection taken on the phone is logged in the CRM, but the text message tool reads from its own list. A customer list uploaded to an ad platform was built from a report that ignored the suppression flag. The academy exercise on mapping consent and retention across the stack for a wine club practises tracing one member's withdrawal through those tools to the place where it stopped.

Two kinds of permission also get confused. The first is the visitor's choice about cookies and tags, which the consent management platform records and signals; Google documents consent mode as the way default states are set before its tags load and updated when the visitor chooses, with tags adjusting their behaviour to the state. The second is the lawful basis for each processing purpose, which the GDPR expects to be chosen before the processing starts and recorded beside the purpose. For electronic marketing, separate rules apply alongside it, which the ICO's direct marketing guidance explains, including consent for emails and texts in many cases and an opt-out in every message. A consent map keeps both kinds in view, row by row, without mixing them up. The academy checkpoint on consent maps, retention and privacy operations tests what a consent map records, how to trace a withdrawal and which events should reopen the map.

Before you start

Read the data flow and integrations guide if you have not, because the map follows data along the flows it describes. Gather the stack inventory, the integration register if one exists, the consent management platform's category and tag assignments, the wording of every sign-up form and checkout opt-in, and the measurement plan, whose events are the ones consent decides whether to collect.

Book time with whoever owns data protection in the business. They decide the basis for each purpose; the map records it. You also need someone with administrator access to each tool on the list, since the settings that filter on consent are rarely visible to ordinary users. The GDPR glossary entry and the privacy and consent guide are useful shared reading for everyone involved, and the guide to list building and consent covers what an email consent record should contain.

The steps

  1. List every purpose the stack processes personal data for

    Write one row per purpose, not per tool. Typical rows are site analytics, advertising measurement and remarketing, email marketing, text message marketing, customer list uploads to ad platforms, on-site or in-email personalisation, and sales follow-up of marketing leads. A purpose is what the data is used for; the same email address can appear in several rows. Read the stack inventory and the integration register to make sure every tool that receives personal data appears in at least one row. If a tool appears in none, either a purpose is missing or the tool should not be receiving the data.

  2. Record where the permission or other basis for each is captured

    For each row, record what the business relies on and where it comes from: the cookie banner's analytics or advertising category, a newsletter checkbox at checkout, a preference centre, a sign-up form, or another lawful basis the data protection owner has decided applies. Record the wording people saw and where the evidence is kept. The GDPR describes consent as freely given, specific, informed and unambiguous, and as easy to withdraw as to give, as the lawful basis entry explains, so a row relying on consent should be able to point to the moment and words that produced it. Where the row relies on another basis, record that it was chosen and by whom. Do not decide the basis yourself; leave the cell open and flag it until the owner has.

  3. Name the system that holds the record for each purpose

    Pick one system as the holder of each row's record: the consent management platform for cookie and tag choices, the CRM or the email platform for marketing permissions, a central preference store where the stack has one. Every other tool reads from that system or receives updates from it. Where two systems both claim to hold the record, decide which wins, as the guide to syncing the CRM and marketing platform does: for consent, the most restrictive value wins, so an older permission never overwrites a newer withdrawal. Record permissions per channel, with source and date, as the list building guide asks.

  4. List the tools that read each record, and how

    For each row, list every tool that acts on the data and say how it learns the consent state: it reads the holding system directly, it receives a sync on a schedule, it receives a file someone builds by hand, or it does not learn it at all. The last answer is a finding. The integration register's consent filter column is the place to check what each connection actually excludes. Each integration register row also names the connection's owner, the person to ask when its filter is missing or unclear. For customer list uploads, the playbook for customer lists across ad platforms asks for a recorded basis and the removal of people who asked not to be marketed to before any file leaves the CRM; record where in the build of the file that check happens.

  5. Trace a withdrawal and an objection through every tool

    Take each row and follow a change of mind through the stack. Someone unsubscribes in the email platform: which tools hear about it, by which route and how quickly? Someone tells customer service they object to direct marketing: where is it logged, and does it reach email, text messages, customer list uploads and personalisation? Someone replies STOP to a text: the text tool captures it first, so does it travel back to the CRM that holds the record, and on to the tools that read it? Trace both directions for every channel that captures choices of its own. The GDPR's right to object to direct marketing is absolute, as the GDPR entry states, and the data operations and taxonomies guide says an objection must be applied in every channel it covers and that suppression lists kept separately in each tool drift apart. A suppression list every sending tool checks is the usual way to make that hold. Write the route for each row; where the route is a person remembering, mark it. When a request arrives by phone, reply or in person rather than through a link, the marketing opt-out request SOP runs it along these routes one person at a time.

  6. Check how tags behave under each consent choice

    For rows that depend on tags, record what each tag does when the visitor accepts, rejects and has not yet answered. The consent mode guide explains how Google's tags read the default and updated states, and the tag governance guide sets the rule that every tag in the inventory has a consent category. Tie each row to the measurement plan's events, so it is clear which events consent decides whether to collect. If tags fire before the visitor answers, the troubleshooting page on tags firing before consent traces each early request to its cause.

  7. Test each row with marked test records

    Write one test per row and run it. For tags, open the site in a clean browser, reject, and confirm that the tags in that category hold back; the consent banner QA checklist holds the banner checks. For marketing permissions, create a marked test contact that has unsubscribed or objected and confirm it is excluded from the next send, the next sync and the next customer list file. The test contact's name and email domain should make it unmistakable, so nobody treats it as a customer. Record the date and the result beside the row. A row whose test fails is fixed before anything else on the map.

  8. Name an owner per row and the changes that trigger an update to the map

    Give each row one owner: the person who answers if that purpose's consent fails. Then write down the changes that require the map to be updated before they go live: a new tool or integration, a new sign-up form or opt-in wording, a change to the banner or its categories, a new customer list upload, a new channel, and any change the data protection owner makes to a basis. Make the map one of the documents the stack review reads, with each owner confirming their row and its last test result. A consent map kept this way describes how the stack behaves now, not how it behaved on the day the map was drawn.

The map is a single table. The example below shows the columns with a sample of rows written in general terms; the business's own map replaces every cell with its actual systems, decisions and tests.

PurposePermission or basis, and where capturedSystem holding the recordTools that read it, and howHow a withdrawal or objection travelsTestOwner
Site analyticsVisitor's analytics choice in the bannerConsent management platformAnalytics tags, through the tag manager and consent modeNew choice updates the state on the next pageReject in a clean browser; analytics storage stays deniedAnalytics lead
Advertising measurementVisitor's advertising choice in the bannerConsent management platformAd tags, through the tag manager and consent modeAs aboveReject; ad tags hold backPaid media lead
Email marketingOpt-in at sign-up or checkout, with wording and dateEmail platform or CRM, as decidedEmail platform; CRM by syncUnsubscribe syncs to the CRM and the global suppression listUnsubscribed test contact excluded from next sendEmail lead
Text messagesSeparate opt-in for texts, with wording and dateCRMText message tool, by syncObjection syncs from the CRM to the text tool; a STOP reply captured in the text tool syncs back to the CRMObjected test contact excluded from next text; test STOP reply appears in the CRMCRM owner
Customer list uploadsBasis recorded by the data protection ownerCRMUpload file built from the CRMSuppression and objection filters applied when the file is builtSuppressed test contact absent from the filePaid media lead
PersonalisationBasis recorded by the data protection ownerCRM or customer data storeSite and email personalisation rulesObjection removes the person from personalised segmentsObjected test contact sees default contentLifecycle lead
Illustrative example — not a real result

Plinzarro Kitchenware draws its consent map

Plinzarro Kitchenware, an invented online kitchenware retailer, maps 6 purposes across 9 tools in two working sessions with its operations manager, who owns data protection. Four rows hold up. Two do not.

The first gap is the customer list upload. Newsletter consent given at checkout is not the problem: around 40 percent of new subscribers join that way, and the permission is stored in the e-commerce platform, which passes it to the email platform correctly. But the customer list the paid media lead uploads to ad platforms every month is built from an e-commerce export that ignores the suppression list and the objection flag, so people who unsubscribed and people who told customer service they object to marketing are both included. The newsletter checkbox covers email, not an ad audience, so the upload row also has no recorded basis until the operations manager decides one.

The second gap is objections. Customer service logs a "no marketing" request in the CRM, about 25 a month, and the email platform reads the CRM's flag nightly. The text message tool, added the year before, reads only its own opt-out list, so an objection made by phone never reaches it. The route fails in the other direction too: a STOP reply is recorded in the text tool's own list and never reaches the CRM, so the record that should hold each person's choices is incomplete.

The fix has the same shape in both cases: one record that every tool reads. The upload file is now built from the CRM, with the suppression and objection filters applied before export, and the operations manager has recorded the upload's basis in its row. The text message tool syncs its suppression from the CRM's flag, and its STOP replies sync back to the CRM. Each row gets a test: a marked test customer who has objected is added, and the team checks that it is absent from the next upload file, the next email send and the next text campaign; a test STOP reply is checked for in the CRM. All the tests pass on the second run.

Verify it worked

  • Read each row. It should name the purpose, basis, holding system, readers, route, test and owner. Flag any undecided basis.
  • Check every tool in the stack inventory and the integration register against the map. Each tool should appear in at least one row.
  • Run the marked test contact who has unsubscribed or objected. It should be absent from the next send, the next sync and the next customer list file.
  • Send a test STOP reply to the text tool. It should appear in the system that holds the record.
  • Reject in a clean browser. Tags in the rejected category should hold back, with the test date and result beside the row.

If it goes wrong

  • A test contact still appears in a send or file. The tool never learns the consent state, or the file came from an export that ignores the flags. Filter where the file is built.
  • A STOP reply never reaches the record. The text tool captures it first and nothing syncs back. Add the route back to the holding system.
  • An old permission overwrites a newer withdrawal. The sync lets the latest edit win. Change the rule so the most restrictive value wins.
  • Tags still fire after a rejection. A tag may lack a consent category, or a default state may be wrong. Check both, then re-run the test.
  • A basis cell stays empty. Flag the row and ask whoever owns data protection. Do not choose the basis yourself.

Next steps

Common mistakes

  • Mapping by tool instead of by purpose. The same address serves several purposes, and each needs its own row, basis and test.
  • Treating the banner as the whole of consent. Cookie and tag choices are one kind of permission; marketing permissions and objections live in other systems and need their own rows.
  • Letting each tool keep its own suppression list. Separate lists drift, and an objection recorded in one tool is ignored by the next.
  • "Most recent edit wins" for consent. An old permission synced late overwrites a new withdrawal. For consent, the most restrictive value wins.
  • Deciding the lawful basis in the map. The map records the data protection owner's decision; it does not make it.
  • Building upload files from exports that ignore suppression and objection flags. Filter at the point the file is built, from the system that holds the record.
  • No test per row. Without a marked test record checked against every send, sync and upload, a broken route is found by a complaint.

Frequently asked questions

Is the consent banner not enough to manage consent across the stack?

The banner, and the consent management platform behind it, records a visitor's choices about cookies and tags and signals them to the tags on the page. It does not hold the permission someone gave to receive marketing emails at checkout, the objection they made to customer service or the decision to include them in a customer list upload. Those live in the CRM, the email platform and other tools, and the consent map is where both kinds of permission are drawn together, purpose by purpose.

Should every tool hold its own copy of consent?

Most tools need to read consent, and many keep a copy because that is how they filter. The problem is not copies; it is copies that are updated independently. The map names one system that holds the record for each purpose and makes every other tool read from it or receive updates from it, so a withdrawal recorded anywhere reaches everywhere. Suppression lists kept separately in each tool are the classic case of copies that drift apart.

Does this page tell us which lawful basis to use?

No. The GDPR lists the lawful bases and expects one to be chosen for each purpose before processing starts, and the ICO's direct marketing guidance explains when electronic marketing needs consent. Which basis applies to a given purpose in a given business is a decision for whoever owns data protection there, with legal advice where needed. The map records that decision and makes the stack act on it; it does not make it.

Terms used on this page

Frameworks

Where this page is used

Part of these learning paths

Sources

  1. Regulation (EU) 2016/679 (GDPR) — EUR-Lex
  2. Direct marketing guidance — Information Commissioner's Office (UK)
  3. Consent mode — Google Tag Platform

Related