All insights

Marketing & Visibility

Email marketing for clinics: separate care messages from promotional follow-up

8 min read

Audit the purpose of each clinic email, document the correct marketing permission and test suppression before turning follow-up into an automated campaign.

Marketing.clinic mascot sorting appointment service messages and permission-based promotional emails into separate trays.

The short answer: classify the message before choosing the audience

Keep necessary care and administrative emails separate from promotional follow-up. Being a patient is not blanket permission to receive marketing. Before activating an email sequence, identify its purpose, the permission supporting it and the system that stops it when someone opts out.

Start with the messages already leaving your clinic: booking confirmations, appointment reminders, clinician-approved follow-up, newsletters and invitations to return. The practical task is to give each template an owner and a clear route through approval, sending and withdrawal.

This guide concerns UK clinic communications to individuals, not corporate prospecting. Official guidance was checked on 6 November 2026. It is operational guidance, not legal advice or a compliance guarantee; have your data protection lead review uncertain classifications and health-data uses.

Read the whole email, including the footer

The ICO distinguishes neutral service communications from direct marketing. Appointment confirmations and reminders can be service messages. Adding promotional content can make the message direct marketing even when administration remains its main purpose. Ordinary branding alone does not do that.

A factual time change and a discount on another service therefore belong in different workflows. Do not decide from the template name: a message labelled aftercare may still contain an upsell. Have the clinical owner assess genuine care content and the marketing owner identify invitations to purchase.

Review subject lines, preview text, banners, linked buttons and automatically inserted signatures. Shared footer software can quietly put a promotional campaign into every receptionist's reply. Save an actual delivered test email, not just the editor preview, because different sending systems may add different content.

Create a small template register

Use a spreadsheet with one row per template. Record the sending system, trigger, purpose, reviewer, permitted audience, withdrawal behaviour and last test date. Include manual reception templates as well as automated campaigns; otherwise the same promotional message can bypass the controls through an ordinary inbox.

Give ambiguous templates a paused status. Reception should not have to interpret advertising and privacy law during a busy shift. A named manager can gather the exact wording and intended recipients for review, then record the decision beside the template.

TemplateReview questionOperational control
Appointment reminderIs every element necessary and non-promotional?Lock the approved version against campaign banners.
Return invitationIs this agreed care or encouragement to buy again?Pause until purpose and audience are approved.
General newsletterWhich documented permission covers this recipient?Check eligibility and suppression before sending.
Clinician follow-upWho approves the content and delivery channel?Keep clinical review separate from marketing approval.

Check every products-and-services soft opt-in condition

For unsolicited marketing emails to individual subscribers, PECR generally requires consent unless an applicable exception is satisfied. The products-and-services soft opt-in is not a general existing-patient exemption. All the following conditions must be met.

Do not assume a historic database qualifies because addresses came from appointments. If the collection-stage opt-out is missing or cannot be evidenced, an unsubscribe link added today does not repair it. Exclude anyone who opted out. Do not send an unpermitted promotional email merely to ask for marketing consent; obtain an appropriate choice through a lawful interaction instead.

  • Your organisation obtained the contact details directly from the individual.
  • You collected them during a sale or genuine negotiations for a sale of a product or service.
  • You market your own similar products or services, within reasonable expectations in that context, not another organisation's offers.
  • You offered a clear, simple opportunity to refuse marketing when collecting the details.
  • Every subsequent marketing message offers a simple opt-out, free apart from any ordinary transmission cost.

Make consent evidence specific and health-data use separate

Valid consent must be freely given, specific, informed and unambiguous. Use a separate, unticked choice that names the clinic and email marketing, with a clear withdrawal route. Agreement to a privacy notice or treatment is not the same choice. Keep the wording version, source, timestamp and affirmative action.

Health information has additional protection. The ICO says special-category direct marketing needs a lawful basis and a special-category condition; explicit consent is generally needed because another condition is unlikely to apply. An email subscription does not by itself authorise targeting based on diagnoses or treatment history.

Our recommended boundary is to keep clinical records out of campaign tools and never upload patient or health-data audiences to advertising platforms. Do not send patient identifiers, clinical details or revealing audience labels to analytics. Hashing an email address is not a reason to relax that boundary.

Make stopping a campaign part of the design

The ICO recommends minimal suppression records so objections are respected rather than forgotten when a list is imported again. Stop the relevant marketing promptly and honour the scope of the request. Keep only what is needed to prevent further unwanted contact; suppression is not an audience for another campaign.

Map every place that can send: practice software, booking provider, newsletter platform, reception mailbox and agency tool. Give one person responsibility for checking that preference changes reach them. A working unsubscribe page is insufficient if tomorrow's scheduled export restores the recipient.

Test with invented contacts controlled by the team. Unsubscribe one, submit a broad objection for another, then run the normal import and scheduled-send process. Confirm the blocked messages never enter the delivery queue. Keep screenshots of the settings and the test outcome without including real patient records.

Review what the campaign promotes, not just who receives it

Permission to email does not make every advertisement permissible. ASA guidance prohibits advertising prescription-only medicines to the public; indirect references can also promote them. Do not use a patient newsletter to circulate public POM promotions or assume removing a brand name solves the problem.

Give the reviewer the complete message and destination page together. Ask them to check claims, images, prices, exclusions and the action the reader is encouraged to take. Keep a dated approval record for the final version, including any changes introduced by the email platform.

A hypothetical clinic with one overworked email sequence

Imagine a two-site clinic whose booking software sends a confirmation, a practical reminder and a later invitation to buy another service. This is a hypothetical example, not a client case study. The manager discovers that all three templates use the same audience setting and promotional footer.

The team pauses the promotional stage, removes the offer banner from the administrative templates and records who approves each message. It checks permission evidence before considering any marketing restart. Records with unclear eligibility remain excluded; nobody estimates consent from how recently a person attended.

Next, staff use dummy records to test withdrawal across both sites. They also confirm that genuine necessary communications follow their separately approved process. The outcome is a documented workflow, not an invented uplift in bookings. Success can first mean fewer ambiguous templates and no failed suppression tests.

Make a permitted newsletter useful enough to keep

Once the audience is eligible, choose one useful topic rather than a catalogue of offers. A general clinic newsletter might explain how consultations work, introduce a verified practitioner profile or clarify how to contact the right branch. The content still needs advertising review; an educational tone does not make a promotional message exempt.

Use a descriptive subject, one main answer in ordinary text and a link to the relevant maintained page. Do not use a misleading 'Re:' subject or imply that a newsletter contains a personal clinical update. Keep the monitored reply address, sending identity and unsubscribe route obvious on a phone.

Review aggregate delivery failures, complaints, unsubscribes and useful destination visits before expanding. Open rates are affected by image blocking, privacy features and automated activity, so do not treat an open as proof that someone read or valued the message. Test rendering, links and suppression with team-controlled addresses first. A smaller relevant list is preferable to a larger list whose permission or purpose nobody can explain.

The manager's pre-send checklist

Run this check when introducing a sequence, changing a form or replacing a supplier. Record the approver and unresolved issues. Do not quietly turn an unresolved item into an assumption because the campaign has a launch date.

  • The delivered message has an agreed purpose, owner and current approved version.
  • Recipient eligibility can be evidenced without guessing from patient status.
  • The sending identity is clear and the marketing unsubscribe route works without a login.
  • Imports, scheduled messages and agency systems respect the latest suppression record.
  • The export contains only the minimum approved fields, with no clinical notes or audience uploads to ad platforms.
  • The linked page matches the approved message and has been checked for advertising concerns.
  • Someone monitors delivery failures, complaints and replies after launch and can pause the sequence.

FAQ: Can we email everyone who has attended the clinic?

Not simply because they attended. Keep uncertain records out of promotional sends until the relevant permission evidence is reviewed. A smaller, explainable audience is a more manageable starting point than a large historic export with unknown collection settings.

FAQ: Is a recall message always a care message?

No. Its purpose, content and context matter. Ask the clinical owner whether it implements an agreed care process, and inspect any promotional additions. Document the classification instead of giving every template called recall an automatic administrative exemption.

FAQ: Should unsubscribing stop appointment confirmations?

A marketing unsubscribe should not automatically disable separately justified, necessary non-promotional communications. Design separate controls and have the care team review contact preferences and appropriate delivery channels. Do not use that distinction to keep selling through a service-message template.

FAQ: What should we ask our website supplier to change first?

Ask them to show the live form wording, where each choice is stored and how it reaches the sending platform. Then test withdrawal end to end. Our website services can help review that journey; bring the approved communication rules so implementation follows a documented decision.

Need help putting this into practice?

Talk to our team about your website, enquiry journey and the practical changes your business needs. We can help define the next step and who should own it.

Get expert website advice

Free HxAi Website Snapshot

Find out what may be costing your website enquiries.

Enter your website and we will prepare an initial snapshot covering visibility, performance, structure, conversion pathways and AI search readiness. A human specialist will review the analysis before your report is emailed.

  • No automated sales report
  • Reviewed by a human specialist
  • No obligation
  • Your details are never sold
HxAi Snapshot RequestReady

Takes approximately 30 seconds. Your reviewed report will follow by email.