Safety Norms: Digital Marketing and Online Advertising

1Purpose, Scope, and Safety Principles

These norms apply to anyone who plans, creates, approves, launches, or measures paid, owned, or earned marketing that reaches people online. Covered channels include search, social media, display, video, email, messaging, influencer and affiliate activity, and retargeting. Covered people include employees, contractors, agencies, creators, and volunteers acting for the organization.

The purpose is to prevent harm to audiences, to people whose data is used, to creators, to staff, and to the organization itself. Four principles guide every decision: respect the audience, tell the truth, collect and use only the personal data needed, and escalate when unsure. Statements marked must are mandatory. Statements marked should are recommendations that a manager may adapt only with written reasons. Advertising codes, privacy laws, copyright statutes, and platform policies vary by jurisdiction and change, so verify current versions with competent authorities, platform documentation, or qualified legal counsel.

2Roles, Competence, and Accountability

The accountable marketing lead approves each campaign, including its objectives, budget, audience definitions, and public claims. A privacy steward reviews tracking tags, audience uploads, vendor data sharing, and any use of sensitive information. A legal or compliance reviewer must check campaigns in regulated or sensitive categories, such as health, finance, alcohol, gambling, political messaging, and anything directed at children. Agencies and creators are bound by contract to these norms and to the approval path.

Competence is required before access. Only people who have completed approved training on platform advertising policies, consent tools, and data handling may build tracking, upload customer lists, or launch regulated campaigns. Anyone without that authorization must request support rather than work around the gap. Managers must remove ad account access within one business day when a person changes roles or leaves.

• Ad accounts must use named individual logins with multi-factor authentication. Shared passwords are prohibited.
• Each campaign file must record who approved each claim, audience, and budget change.
• Contractors may not hold owner-level access to customer data unless the privacy steward approves it in writing.

3Course-Specific Hazards and Risk Assessment

Digital marketing hazards are mostly harms delivered through information, targeting, and data rather than physical force. Before launch, the campaign owner must complete a written risk assessment that names the audience, data used, claims made, channels, and worst plausible outcome. Likelihood and severity are rated low, medium, or high. Any high rating requires legal or privacy review and a named senior approver before launch. Ratings must be revisited whenever the audience, offer, or channel changes.

• Misleading claims and hidden sponsorship (high): exaggerated results, fabricated reviews, undisclosed paid endorsements, or buried affiliate links can deceive buyers and breach advertising rules.
• Unlawful data collection (high): tracking pixels, session recorders, or uploaded lists gathered without a valid basis or notice can trigger enforcement and loss of trust.
• Harmful targeting (high): ads aimed at minors, people in financial distress, or groups defined by sensitive health or identity traits can cause real harm.
• Copyright and trademark infringement (medium to high): unlicensed music, stock images, fonts, or logos can lead to takedowns and claims.
• Platform enforcement (medium): disapproved ads or suspended accounts can stop revenue and erase performance history.
• Creator harassment and doxxing (medium): public-facing creators may face abuse, threats, or attempts to reveal personal information.
• Brand safety failures (medium to high): ads placed beside hateful or violent content, or a tone-deaf message, can spread quickly.
• Physical strain (low to medium): long screen sessions cause eye, neck, shoulder, and wrist discomfort.

4Preventive Controls and Safe Work Practices

Preventive controls are built into the workflow before anything goes live. Claims are checked against substantiation on file. Creative is checked against a rights register. Data flows are mapped before tags are deployed. Launch is blocked until required sign-offs are recorded in the campaign file. Controls must be verified in a pre-launch review, not assumed.

Safe work practices protect the people doing the work. Staff must not engage in coordinated fake engagement, post from personal accounts without disclosure, or accept requests to attack competitors or individuals. Creators must not film in unsafe locations or perform stunts for content, and must follow manufacturer instructions for any equipment used.

• Every factual or comparative claim must have supporting evidence on file before launch. Evidence must be kept for the retention period set by the legal reviewer.
• Sponsored, affiliate, or incentivized content must carry disclosure in the ad or post itself, placed where the audience will see it before acting. A profile bio alone is not sufficient.
• Reviews and testimonials must come from real customers who consented to publication. Fake, purchased, or edited-to-mislead reviews are prohibited.
• Customer lists may be uploaded to ad platforms only when the source permits that use, notice was given, and any required choice was honored.
• Tracking tags may not be placed on pages that collect health, financial, or precise location details without privacy steward approval.
• Behavioral advertising and data collection aimed at children are prohibited unless legal and privacy reviewers approve a specific, compliant design.
• Music, images, footage, fonts, and logos must be original or licensed for the channel and territory. License records are filed with the campaign.
• Synthetic images, cloned voices, or likenesses of real people require written consent and may not imply an endorsement the person did not give.
• Creator agreements must require disclosure, prohibit false claims, and grant the brand the right to request edits or takedown.
• Complaints must be logged and reviewed within one business day. Any claim that a campaign is misleading triggers a pause pending review.
• Recommendation: use frequency caps and exclusion lists to limit repeated exposure to distressing or irrelevant ads.
• Recommendation: preview each ad on desktop and mobile, and test every link, disclosure, and form.
• Recommendation: launch small and monitor complaints, refunds, and disapprovals before scaling budget.

5Equipment, Environment, Information, and Protective Safeguards

Workspace safeguards reduce strain and errors. Screens should sit about an arm's length away with the top edge at or slightly below eye level. Short breaks at regular intervals, such as every hour, are a recommendation that supports comfort and focus. Damaged chargers, cables, or power strips must be removed from use and reported to facilities.

Information safeguards protect accounts and data. Use an approved password manager and multi-factor authentication. Lock screens when leaving a device. Do not open ad accounts or customer exports on public networks unless an approved secure connection is used. Do not display dashboards containing personal data in public spaces or on shared screens. Export files must be stored in approved locations and deleted according to the retention schedule.

Production equipment such as lights, tripods, audio rigs, and drones must be operated only as the manufacturer directs. Drone and location filming may require permits that depend on jurisdiction, so verify requirements with the relevant authority before shooting. Campaigns involving physical products or samples must follow local product safety and labeling rules, confirmed with competent authorities or qualified professionals.

6Incident Reporting, Emergency Response, and Escalation

An incident is any event that could cause harm, including a misleading live ad, a privacy complaint, a suspected data breach, a copyright claim, a platform suspension, or abuse directed at a creator. A near miss is an event caught before harm occurred, such as a broken link found in preview. Both must be reported to the campaign owner and privacy steward on the same working day as discovery, using the incident log.

Immediate actions are required. Pause or unpublish the affected ad or post. Capture evidence by saving screenshots, URLs, timestamps, and audience counts. Do not delete comments or messages without logging them first. Do not reply to threatening individuals or engage in public disputes. If a person faces immediate danger or a threat of violence, contact local emergency services first, then notify the manager. Staff must never attempt to confront or locate the person themselves.

Data incidents follow a separate path. The privacy steward determines whether personal data was exposed, assesses any notification duties, and engages counsel. Notification deadlines and recipients differ by jurisdiction and must be verified with competent authorities or qualified counsel. Platform enforcement should be appealed through the official process using the accountable account holder, not personal accounts.

7Records, Training, Inspection, and Review

Every live campaign needs a complete campaign file. It must contain the approval record, risk assessment, claim substantiation, consent records, license records, data processing notes, screenshots of final live ads, and the incident log. Files are retained according to the schedule set by legal and privacy reviewers, and access is limited to authorized staff.

Training is mandatory before ad account access and must be refreshed at least once each year and after major platform or regulatory changes. Training covers ad policies, disclosure, consent, data minimization, copyright basics, creator safety, and incident reporting. The manager keeps completion records.

Inspection keeps controls working. Each month, a reviewer samples live campaigns and checks disclosures, claims, tracking, and license records. Metrics include complaint volume, disapproval rate, takedown requests, near misses, and time to close incidents. Norms are reviewed at least once each year and after any significant incident, enforcement action, or platform policy change. Recommendations that prove unworkable should be revised rather than ignored.

8Conclusion and Next Steps

These norms are a starting point, not a finished rulebook. Each team should adapt them into locally approved procedures that match its markets, platforms, product categories, and staffing. Verify current advertising codes, privacy laws, copyright rules, and platform policies with competent authorities and qualified professionals before relying on any specific requirement. Clear rules, honest claims, respect for personal data, and prompt escalation protect audiences, creators, staff, and the organization.

Implementation checklist

• Confirm each claim has substantiation on file before launch.
• Verify disclosure wording appears within the ad or post itself.
• Check that customer data uploads and tracking tags have privacy steward approval.
• Confirm every image, track, font, and logo has a license record.
• Obtain written consent before using a real person's likeness or voice.
• Complete a risk assessment and rate likelihood and severity.
• Preview ads on desktop and mobile and test all links.
• Log complaints and pause any campaign under misleading-claim review.
• Report incidents and near misses on the same working day.
• Keep ad accounts under named logins with multi-factor authentication.