Marketing automation for lead capture, follow-up and campaign attribution

A CRM stores your contacts, deals and sales activity, while marketing automation acts on those records by capturing leads, sending follow-up and moving contacts between stages when something happens. The CRM is the system of record; automation keeps it current and keeps conversations moving. Many CRMs include basic email automation, but it rarely covers WhatsApp, several lead sources and ad-platform feedback together, which is where a separate orchestration layer such as n8n is useful.

No. n8n is an orchestration tool rather than a complete marketing platform, so it usually works alongside one. It does not design emails, host your contact lists or provide sending infrastructure of its own. What it does well is connect systems: pulling leads from several sources, applying routing and timing rules, triggering WhatsApp messages and keeping the CRM current. Many organisations keep a CRM or email platform for lists and campaign emails, and use n8n for the logic those tools cannot express or only offer on higher plans.

Usually not, as long as your current CRM offers an API or webhooks on the plan you pay for. Before quoting, we check whether the API allows the reads and writes the workflows need, what call limits apply on your edition, and whether pipeline stages and fields are tidy enough to automate against. If stages are unclear or the CRM is really a shared spreadsheet, we fix the structure first, because automation repeats whatever process it is given, including its gaps.

No. WhatsApp's business policies require a contact's opt-in before you message them, and any conversation your business starts must use a template Meta has approved. We collect opt-in on forms and in chat, store when and where it was given, and only start WhatsApp sequences for contacts with a valid record. Free-form replies are allowed within 24 hours of the customer's last message. Meta charges per template message at rates that vary by category and country, so we keep WhatsApp for the touches that matter and use email for the rest.

There is no fixed number: a sequence should run only while each message gives the person something useful, and stop the moment they reply, book, buy or opt out. We usually start with a short sequence where every message answers a common question or shows a relevant example, rather than repeating a reminder. Messages per contact are capped across all sequences, quiet hours and time zones are respected, and reply and unsubscribe rates are reviewed so weak messages are rewritten or removed.

Lead scoring adds points for signals that suggest a sale and removes them for signals that do not, and your sales and marketing leads agree those rules with us. Typical inputs are the service requested, budget or timescale answers, location, pages visited and whether the person replied or booked. We start with simple, explainable rules, test them against past leads that did and did not convert, and adjust. A model-based score only makes sense once you hold enough historical outcomes to train and test it.

We authenticate your sending domain, keep lists clean and send only to people who expect to hear from you. In practice that means SPF, DKIM and DMARC records, a separate subdomain for marketing mail, one-click unsubscribe and automatic removal of addresses that bounce. Gmail and Yahoo require these basics from bulk senders and watch complaint rates closely. New sending domains start with small volumes, and contacts who never open are moved to a slower track or removed, which protects delivery for everyone else.

We record first touch and last touch on every contact and keep every touch in between, so you can compare models instead of committing to one. First touch shows which campaigns introduce new people; last touch shows what prompted the enquiry. Because the full history sits in the event store, a weighted or multi-touch view can be added later without collecting the data again. Earlier visits can only be captured with the visitor's cookie consent, so some leads will only ever have a last touch, and reports say so.

Sometimes, but it depends on who they are and how you obtained their details, so the rule is set with your advisers rather than assumed. Under PECR, emailing individuals usually needs consent unless the soft opt-in applies: details gathered during a sale or quote, marketing of similar products and a clear chance to refuse each time. Addresses at limited companies follow different rules, though UK GDPR still applies. We encode the rules you approve, record each contact's basis and honour opt-outs across email and WhatsApp straight away.

Yes. Routing rules read the service requested, the location given or implied by the phone code, the language the person wrote in and the campaign they came from, then assign an owner and start the matching sequence. Rules can share leads in turn across a team, respect working hours in each country and reassign a lead nobody has touched within an agreed time. Sequences can run in English, Urdu or Arabic, and Meta approves each WhatsApp template language separately.

It is logged, acknowledged and assigned as soon as it arrives, even though nobody is at a desk. The person receives a reply on the channel they used, with an answer to their question where it is a common one and a link to book a call or visit. The assigned owner is notified, and the lead sits at the top of their list when the working day starts. If nobody acts on it by an agreed time, the workflow alerts a manager or passes it to a colleague.

A first release that captures leads from one or two sources, syncs them to your CRM and sends an acknowledgement with an owner alert is often live two to four weeks after discovery, as an indicative range. The main drivers are the number of lead sources, how clean the CRM data is, WhatsApp Business account verification and template approval, and how many sequences need writing. Scoring, further sequences and ad-platform feedback then follow in short releases you review every one to two weeks.

Costs depend on scope, so we give a fixed quote for the build after discovery and an estimate of monthly running costs before you commit. Build cost rises with the number of lead sources, integrations, sequences to write and reports. Running costs cover hosting for n8n and its database, your email sending service, Meta's WhatsApp template charges and any AI model usage. Unlike platforms priced by the number of contacts, the self-hosted community edition of n8n has no per-contact or per-execution licence fee, so running costs follow message volume rather than list size.