An opt-in begins the work of earning engagement

Campaign pressure can push a marketing team to send another blast before anyone checks whether recipients will find it useful. Permission makes the send possible, but leaves a commercial question unanswered: what will the buyer gain from receiving it? After collecting an email address, teams need to apply the buyer’s choices to data collection and messaging so permission can support a valuable relationship.

This gap framed the November 2026 MarTech Conference panel “Beyond the opt-in: Engineering consent for better marketing,” moderated by Jeanne Jennings, founder and chief strategist at Email Optimization Shop. Her marketing-optimization business has a commercial interest in helping organizations improve customer relationships. “Are we optimizing for the number of people who say yes, or are we optimizing for the value of the relationship after they say yes?” Jennings asked. Answering that question requires legal, security, marketing, and IT to agree on how permission should guide their work.

Permission and preferences answer different questions

That agreement starts with separating permission from preferences. Jennings made the distinction explicit: “Consent tells us what we’re allowed to do. Preferences tell us what customers want.” A customer can permit contact without wanting every communication that permission allows. Marketing teams therefore need to assess whether each proposed use of customer data fits the relationship.

To assess that fit, teams must distinguish customer statements from observed behavior. Joss Merckaert, head of consulting operations, North America at fifty-five, emphasized that declared data records an expressed choice, while observed data records an action whose meaning needs interpretation. His firm has a commercial interest in consulting work arising from data-governance and implementation needs. Treating these inputs as interchangeable gives an inferred interest the same authority as an expressed request.

What someone browses also affects how marketers should interpret the action. Zack Meszaros, senior principal solutions engineer at OneTrust, contrasted browsing medical content with researching enterprise software or consumer goods because those activities create different expectations about a business’s response. OneTrust has a commercial interest in organizations investing in consent-management technology. In Meszaros’s account, medical browsing needs particularly sensitive interpretation, while enterprise-software and consumer-goods research also require context before they trigger outreach.

Context matters because a permitted message can still conflict with someone’s preferences or the circumstances behind their browsing. Meszaros warned that premature outreach can damage credibility when marketers act before understanding a signal. So teams need to establish what an action means before letting it determine the next communication. Permission supplies one part of that decision; relevance supplies another.

Okoone experts
LET'S TALK!

A project in mind?
Schedule a 30-minute meeting with us.

Senior experts helping you move faster across product, engineering, cloud & AI.

Please enter a valid business email address.

Contextual requests improve permission collection

Relevance also affects whether customers grant permission in the first place. Sal Tripi, an advisor, recommended explaining the exchange when asking for data: what the prospect will provide, why the brand wants it, and what specific benefit the prospect will receive. These explanations give people clear terms on which to assess the request. Tripi argued that this clarity reduces resistance to opting in.

A clear exchange becomes more useful when its timing matches an immediate need. Meszaros described contextual consent as asking for permission when the customer can readily see its purpose, such as offering timely delivery alerts. Someone considering that request can connect permission to a service they currently need. During the November session, he reported a 5% to 10% lift in high-value channel opt-ins associated with contextual permission requests.

Meszaros’s reported result gives teams designing collection a specific objective: make the value of granting permission clear at the moment of choice. Teams can examine customer touchpoints for requests that serve an immediate purpose, then assess engagement, pipeline, and retention as separate relationship outcomes. They also need to assess the commercial return from governance work separately from collection performance. Once permission is granted, systems must honor it.

Teams need a shared interpretation of customer choices

Honoring permission requires the customer’s selection to reach the systems that act on it. A preference center, the interface through which people manage communication and data-use choices, can record a decision while a sending system continues using older instructions. The customer then expects control over the relationship, but the system delivering communications uses different information. A clear interface therefore creates an obligation that extends beyond the interface itself.

Preference centers were already common among the session’s audience. A poll associated with the November conference found that more than half relied on standard preference centers; a separate finding was that a third operated strictly out of compliance fear. Both standard and modern preference centers depend on choices reaching the systems that collect data and execute marketing. Teams evaluating an upgrade therefore need to test whether changing a selection changes those systems’ subsequent behavior.

Before teams can test that behavior, they need to agree on what each selection means. Merckaert identified misalignment between legal, security, marketing, and IT as an early cause of operational silos. “It breaks before implementation even starts,” Merckaert warned. Without a shared interpretation of the choice and the activity it constrains, engineers start building without a common decision to enforce.

To reach that agreement, Tripi advocated collaborative data governance: shared rules and responsibilities for how the organization handles data. Marketing needs to understand applicable constraints before campaign pressure creates a compliance surprise, while engineering needs clear parameters for implementing them. A governance framework needs enough authority to turn those discussions into decisions that shape daily work. Those decisions then become requirements for the systems collecting data and sending communications.

Customer choices must reach collection and delivery systems

The first technical requirement concerns collection itself. Meszaros called for governing tracking scripts, the code on web properties that collects information about visitors’ activity, so customer choices affect how information is gathered. A preference about data use has limited effect when collection proceeds independently of it. Consent management therefore begins with how the website operates, before an outbound campaign uses the resulting data.

From the website, the relevant choices need to reach the systems that store customer information and execute marketing. Meszaros identified customer data platforms, or CDPs, which bring customer information together; master data management tools, which maintain consistent records across systems; and outbound engines, which send communications. Each has a different purpose, but each needs current instructions relevant to its work. Otherwise, a preference recorded in one place can leave another system acting on an earlier state.

Keeping those instructions synchronized requires reliable connections between applications. Meszaros tied this work to reliable API architecture: application programming interfaces are the connections through which software systems exchange information. Here, the APIs distribute customer choices so the applications collecting data or sending messages can apply them. Teams therefore need to check both delivery of the updated choice and its effect on application behavior.

Checking that effect also requires knowing where the affected data has traveled. Tripi recommended clear parameters for data lineage, the ability to trace data’s origin and movement, and for handling preference revocations, when customers withdraw a previous choice. Lineage helps identify which systems need to respond to a change. Shared governance defines the required response, while synchronization and traceability let teams apply it across those systems.

AI extends consent responsibilities into data and training workflows

The same need to trace and apply choices expands when customer data enters generative AI tools, which produce content from patterns learned from data. Information previously used for campaign delivery may also become an input to an AI workflow. Teams therefore need organized data linked to verifiable consent before that information enters the workflow. A permission check at the point of sending covers only one stage of that use.

Merckaert connected AI’s usefulness to the suitability of its underlying data. “You could have the best models. If you don’t have the right data in the first place, it’s going to be useless,” Merckaert noted. His point makes input selection a prerequisite for useful application: model selection cannot compensate for inappropriate data. Teams need to assess data quality alongside the customer permission governing its intended use.

Because customers can change that permission, the connection between data and consent must remain active. Tripi called for mechanisms that handle preference revocations instantly across messaging platforms and training pipelines, the processes that prepare and supply data for model training. These mechanisms need to respond wherever an active workflow uses the affected customer data. Teams must also distinguish changes to active workflows from actions involving training that has already been completed.

This ongoing responsibility gives data selection a second purpose: keeping useful information connected to a valid customer choice. Jennings emphasized high-signal data, meaning information that helps establish relevance for a particular use. Applying that standard requires teams to decide why they need each input and whether its permission still allows that use. AI makes those decisions more important because an input can influence activity beyond a single campaign.

Audit the consent journey from the buyer’s perspective

To make these requirements concrete, Tripi recommended walking through the consent journey as a buyer. The request should make clear what the person is providing and what they receive in return, without requiring them to infer the exchange. This gives teams a specific customer decision to inspect before tracing its consequences through their systems. It also brings the discussion back to the moment when permission begins.

If the exchange is unclear, Tripi’s next step is to bring the team together and refine the flow. Legal, security, marketing, and IT can use the same customer decision to resolve how the request should read and what the resulting choice should control. Once the request is clear, they need to follow the recorded choice into the marketing and AI workflows that must act on it.

In conclusion

For business leaders, the central decision is whether consent remains a compliance record or becomes an instruction that shapes daily operations. A clear request establishes what the customer agrees to. Shared governance, connected systems, and reliable updates determine whether the business honors that agreement. Investing in collection without funding enforcement leaves the relationship dependent on systems that may hold different instructions.

Leaders can test that operating model by following a customer’s choice from the point of collection through data storage, campaign delivery, and any AI workflows that use the data. Each handoff should have an accountable owner and a defined response when the customer changes their mind. This makes consent a requirement that teams can implement and verify, rather than a policy each function interprets separately.

Success also needs a broader measure than opt-in volume. Collection performance shows whether people accept an exchange; engagement and retention help assess whether the relationship continues to deliver value. Keeping those measures distinct gives leaders a clearer basis for deciding where to invest, and whether the permission they collect supports marketing customers actually want.

Alexander Procter

November 19, 2026

9 Min

Okoone experts
LET'S TALK!

A project in mind?
Schedule a 30-minute meeting with us.

Senior experts helping you move faster across product, engineering, cloud & AI.

Please enter a valid business email address.