The Agentic Revenue CompanyRevenue as a Service
Blog

Consent management in sales: double opt-in, blocks and channel checks as system rules

Consent management belongs in the system as a hard rule: no outreach without a permitted channel, opt-out per contact and a record of every consent.

Some time ago, a contact told your sales back office on the phone that they no longer want to receive emails. The employee noted it properly – in a call note. Months later, a new outreach starts. The note is still there, but nothing in the system stops the email. This is exactly where it is decided whether consent management is a good intention or a rule that works.

As long as people write every email individually, many companies rely on attentiveness: whoever knows the file knows whom not to write to. With agentic sales – AI agents in sales – this shifts. An agent cannot be relied on to read between the lines of free-text fields. It acts on what is in the system as a rule, and it does so at a breadth no team reaches by hand.

That is not an argument against agents. It is an argument for recording consent and contact blocks where they reliably reach every outreach: as a hard rule that is checked before every send. For your customers, this is about more than formalities. Anyone who has asked not to be contacted again and still receives an email starts to doubt the whole business relationship.

What the law says – and what we do not decide for you

When you may contact someone through which channel depends on the legal situation in your market, on the channel and on the nature of the business relationship. Terms such as consent, double opt-in or objection describe different things: agreement to be contacted, a procedure that confirms and records this agreement, and the right to object to being contacted.

This article is not legal advice, and a system alone does not make outreach legally compliant. Which rules apply to your customers and channels is something you clarify with your legal advisers. In the mandate, we record who is responsible for this review. Our job is to make sure the system applies the agreed rules consistently.

Channel checks: no permitted channel, no case

In the systems we build and run, the check does not start at the moment of sending. Without a permitted channel, no case is created in the first place. So an agent does not plan any outreach that would then have to fail at the last hurdle.

Three rules carry this:

  • “Unknown” does not count as “allowed”. If information on consent is missing, the system chooses the cautious reading.
  • A shared address is not clearance for a contact. The fact that a company lists a general email address does not mean a specific contact may be written to.
  • Blocks also apply to the first email. A contact block does not take effect from the second message onwards, but for every outreach.

Blocks already maintained in the connected ERP system are taken over by the system during the data sync. So nobody has to keep track of the same information in two places. Why well-maintained data carries so much weight for agents is described in CRM data quality.

Even the editorially approved first outreach, which goes out without individual approval, passes rule, channel and address checks beforehand. How these checks apply when emails are sent is shown in the feature Send e-mails according to defined rules.

Consent management has two sides: capturing consent and respecting objections. The system has fixed paths for both:

  • Double opt-in with a record: when a contact confirms their consent, this confirmation is recorded. It can be evidenced later instead of merely being assumed.
  • Opt-out and preference page per contact: every contact can unsubscribe themselves or change their preferences. The opt-out applies to that person, not across the whole company.
  • Emergency stop: if something is wrong, sending can be halted without first having to adjust every single rule.

The system now distinguishes two levels here: the channel – such as e-mail or phone – remains the hard upper limit. Within a permitted channel, the decision is made per contact and purpose, for example e-mails about an ongoing case, advertising or a newsletter. On the preference page, each person switches every purpose on and off individually. A withdrawal therefore affects exactly what was meant: someone who no longer wants the newsletter still receives the question about their open order. How channel, purpose, proof and emergency stop work together is shown on our topic page Consent management in marketing and sales.

The opt-out per contact matters particularly in B2B. Within the same company, one person may no longer want emails while their colleague is happy to continue working together. How agents coordinate several contacts at one customer is covered in One customer, one voice.

From mandate to system rule

Consent is not an add-on you build in afterwards. It is part of what we call “Governed by Design”: rights, limits and records are part of the design from the start.

In AI-first coaching, you find out whether and how AI agents can secure and grow your revenue. If the answer is yes, the mandate defines which customers are to be contacted through which channels and who is responsible for the legal review. During implementation, we translate these decisions with your functional leads into rules the system enforces. In operation, we keep them up to date when your requirements change. How agents act within approved scopes is described in Agents: AI agents as the new workforce in sales.

This is particularly noticeable in customer reactivation: when you contact dormant customers again, you often come across contact data whose consent status nobody knows precisely any more. What a first outreach can look like there is shown in The reactivation email.

Your next step

Where do blocks and opt-outs sit in your company today – as a rule in the system or as a note in a file? If you want to clarify how AI agents can contact your customers without commitments getting lost along the way, check your REVAGO potential.


← All articles