Marketing automation
Marketing Automation for SMEs: A Practical Guide
What marketing automation is, the five automations an SME should build first, what Swiss law (the Unfair Competition Act and the FADP) requires for email and SMS marketing, how to choose the tools, and how to measure results.
Key takeaways
- Marketing automation runs a defined sequence when a trigger occurs, such as a form submission, a purchase or a period of silence, using data from your CRM.
- Start with the automations closest to revenue: every lead into the CRM with its source, then quote or cart follow-up. Add welcome, post-purchase and re-engagement flows one at a time.
- Swiss law requires prior consent, a correct sender and a free opt-out for email and SMS advertising (UCA Art. 3 para. 1 let. o), with an exception for existing customers and similar products.
- Choose between an all-in-one platform, connecting the tools you already use, or a custom build based on your process and data, not on feature lists.
- Measure each flow against a baseline on business outcomes such as response time, quote-to-order rate or repeat purchases. Open rates are unreliable.
Marketing automation is software that sends a message or creates a task when a customer does something, such as submitting a form, buying, leaving a cart or going quiet, using the data in your CRM. For an SME the sensible start is a handful of automations tied directly to revenue, built on clean data and documented consent, not a large platform project. This guide covers the five automations to build first, what the Unfair Competition Act and the revised data protection act require for email and SMS, how to choose between a platform, a custom build and your existing tools, how to measure results, and the mistakes that cost the most.
What is marketing automation?
Marketing automation runs a defined sequence when a trigger occurs: a form submission, a purchase, a date or a period of inactivity. Each automation has a trigger, conditions, actions across channels such as email, SMS or a task for sales, and a stop rule. The CRM holds the data it runs on. Without clean data, automation only sends mistakes faster.
A typical example: a visitor requests a quote on your website. The form creates a contact and a deal in the CRM, with the page, campaign and consent status attached. The visitor receives a confirmation within a minute, in the language of the form, and a salesperson gets a task. If the quote has had no reply after three working days, the CRM sends a short follow-up and, two days later, reminds the salesperson to call.
This differs from a newsletter, which goes to a list at a time you choose, and from transactional messages such as order confirmations, which confirm something the customer has just done. Where marketing automation ends and sales automation begins is covered in our guide to CRM automation.
Which automations should an SME set up first?
Start with the two closest to revenue: getting every lead into the CRM with its source, and following up quotes or abandoned carts. Then add a welcome sequence, a post-purchase flow and re-engagement of inactive customers. Build them one at a time and measure each before adding the next, because every flow also adds maintenance.
| # | Automation | Trigger | What it does | What to measure |
|---|---|---|---|---|
| 1 | Lead capture to CRM | Website form, chat, booking | Creates or updates the contact and deal with source, campaign, language and consent; assigns an owner; confirms receipt | Time to first response; share of leads with a known source |
| 2 | Quote or cart follow-up | Quote sent with no reply after a set number of days; cart left without an order | Sends a reminder with the quote or cart link and creates a task; stops on reply or order | Quote-to-order rate; recovered orders |
| 3 | Welcome or onboarding | New subscriber, new customer or first order | Three or four messages over about two weeks: what to expect, how to get started, whom to contact | Second purchase or activation; unsubscribes |
| 4 | Post-purchase | Delivery confirmed or project closed | Usage tips, a review request, a related product | Repeat purchases; reviews collected |
| 5 | Re-engagement | No purchase or interaction for 6 to 12 months | One or two win-back messages, then stops mailing contacts who don’t respond | Reactivated customers; list health |
Lead capture comes first
If leads arrive in a shared inbox, every other automation lacks the data it needs. Capture the source page, the campaign parameters, the language and the consent status with each submission, and check for duplicates before creating a new contact. The first measurable win is usually speed: how many minutes pass between a form submission and the first human reply.
Follow-up needs stop rules
A reminder sent after the customer has already replied or ordered does more harm than no reminder. Every follow-up needs a check before each step: has the person replied, bought, unsubscribed or complained? Whether a cart reminder counts as advertising depends on its content, so the safe default is to send it only to people who consented or who fall under the existing-customer rule described below.
Re-engagement should end with a decision
In an affiliate win-back system we built for a SaaS company’s partner program, each step re-checks the partner’s recent activity before anything is sent. A partner who becomes active again leaves the sequence automatically, and any human reply stops it and becomes a task for a manager. Messages go out in the partner’s own language, chosen from their profile. The same rules apply to customers: re-check before each message, stop on any reply, and stop mailing those who stay silent.
What do Swiss rules say about consent for email and SMS marketing?
Under the Unfair Competition Act (UCA, Art. 3 para. 1 let. o), mass advertising by telecommunication, which includes email and SMS, needs the recipient’s prior consent, the correct sender and a simple, free way to refuse. The exception: customers who gave you their contact details when buying, and were told then that they could refuse, may receive advertising for your own similar products.
The FDPIC summarizes the same rule as opt-in for email advertising, with opt-out for existing customers and similar products. OFCOM adds the practical side: the sender must be clearly identifiable, and recipients need a working way to unsubscribe, such as a “remove me” link. In practice:
- Record consent. Store the date, the form or source, the wording shown and the language for each contact. A double opt-in, where the person confirms by clicking a link, is not required by the UCA, but it gives you evidence.
- Use the customer exception narrowly. Tell customers at the point of sale that they can refuse advertising, and limit messages to your own similar products.
- Make opting out easy everywhere. An unsubscribe link in every marketing email, a working opt-out for SMS, and opt-outs synced between your CRM, email tool and SMS tool.
- Keep transactional and marketing messages apart. An order confirmation is not the place for a promotion to someone who has not consented.
The revised Federal Act on Data Protection (FADP) has been in force since 1 September 2023. It requires you to inform people when you collect their data: at least who you are and how to contact you, the purpose, the recipients and, if data goes abroad, the countries involved (Art. 19). Many automation tools store data in the United States. Since 15 September 2024, Switzerland recognizes US companies certified under the Swiss-US Data Privacy Framework as offering adequate protection; for US providers without that certification you need other safeguards. Wilfully failing to provide the required information can lead to fines of up to CHF 250,000 for the private individuals responsible (Art. 60). If you market to people in the EU, the GDPR may apply as well. This section is an overview, not legal advice.
Should you buy a platform, build custom, or connect the tools you already have?
Buy an all-in-one platform when your journeys are standard and nobody on the team will maintain integrations. Connect your existing tools when your CRM, shop or accounting system already holds the data. Build custom when the process or the data doesn’t fit a platform, when data location is strict, or when volumes make per-contact pricing expensive. Many SMEs end up with a mix.
| Approach | Fits when | Watch out for | Examples |
|---|---|---|---|
| All-in-one platform | Small team, standard flows (newsletter, welcome, cart), marketing owns the tool | Pricing that grows with the contact list; a second CRM next to the one sales uses; data location; exporting later | HubSpot, Brevo, ActiveCampaign, Zoho |
| Connect existing tools | CRM, ERP or shop already in place, for example bexio, Shopify or WooCommerce | No-code chains nobody monitors; missing error handling; accounts owned by a former employee | Zapier, Make, n8n, direct APIs and webhooks |
| Custom build | Multi-step B2B quotes, partner programs, data from several systems, strict data location, high volumes | Needs a developer for maintenance; needs a clear specification and tests | Your own services calling email and SMS provider APIs |
Whichever route you take, check five things before you sign: how consent and opt-outs are stored and synced, whether each contact has a language field (German, French, Italian, English), where the data is stored and whether a processor agreement is available, whether you can export contacts and history, and whether failed runs are logged and alerted. Our comparison of custom and off-the-shelf software goes deeper into the build-or-buy decision.
How do you measure whether marketing automation works?
Measure each automation against a baseline taken before launch, on a business outcome: response time, quote-to-order rate, recovered carts, repeat purchases or reactivated customers. Where volumes allow, keep a small control group that does not receive the flow, so you see what the automation adds rather than what would have happened anyway.
- Don’t steer by open rates. Apple’s Mail Privacy Protection downloads remote content, such as the pixels that count opens, in the background by default, whether or not the recipient engages with the email. Many recorded opens are not reads. Use clicks, replies and conversions.
- Track the cost side. Unsubscribes and spam complaints per flow show when a sequence is too long or too frequent.
- Monitor the plumbing. Failed webhooks, sync delays between CRM and email tool, and bounces break flows silently. Set alerts, not just dashboards.
- Review monthly. Look at each flow’s outcome, adjust one variable at a time, and switch off flows that don’t earn their keep.
What are the most common mistakes?
Most failures come from order and ownership, not software: buying a platform before mapping the process, automating messy data, and flows that keep running after the customer has replied or bought. The rest are missing consent records and messages in the wrong language.
- Tool first, process second. Draw the journey on one page before choosing software.
- No single source of truth. Contacts duplicated across the CRM, the shop and the email tool, each with a different consent status.
- Missing stop rules. Every flow must stop on reply, purchase, unsubscribe and complaint.
- The wrong language. A German reminder to a customer in Lausanne. Store the language per contact and build templates for each.
- No frequency cap. Three flows firing at the same person in one week.
- No owner. Flows that nobody checks after launch keep running on outdated offers.
Where to start
Map your lead and customer journey on one page, list where each piece of data sits today, and check how consent has been recorded so far. Then pick the first two automations from the table above, build them, and measure for a few weeks before adding the next. If you would like help with the mapping or the build, our marketing automation service covers the journey design, and API integrations covers the connections between your CRM, shop and email tools.
Sources
- Federal Act on Unfair Competition (UCA), Art. 3 para. 1 let. o (English translation, Fedlex)
- FDPIC: Advertising and marketing
- OFCOM: Spam
- Federal Act on Data Protection (FADP), Art. 19 and Art. 60 (English translation, Fedlex)
- Federal Office of Justice: New data protection legislation (in force since 1 September 2023)
- Federal Council: Swiss-US Data Privacy Framework, certified US companies offer adequate protection (14 August 2024)
- Apple: Mail Privacy Protection and privacy