Marketing Automation Platform Development

Visual drag-and-drop journey canvas where marketers build multi-step sequences without engineering support. Multi-channel orchestration covers email sends via SMTP or AWS SES, SMS sends via Twilio, push notifications via APNs (iOS) and FCM (Android), and in-app messages delivered through your product's notification layer -- all managed from the same journey canvas rather than across separate tools with no shared state.

Event-based triggers ingest from form submissions, page visits, purchase events, CRM status updates, and event stream sources (Segment or RudderStack event track calls). Conditional branching on contact attributes, RFM segment membership, cohort assignment, or lead score tier lets journeys follow different paths for different audience groups without requiring a separate journey per segment. A/B and multivariate test nodes support configurable split ratios with Bayesian inference for winner selection -- the platform calculates the posterior probability that each variant is the best performer rather than waiting for a fixed sample size, which means winner promotion happens faster. Delay and time-based send windows, wait-for-condition steps, and frequency capping prevent over-messaging contacts who are in multiple active journeys simultaneously.

Drag-and-drop template builder with reusable content blocks and brand-locked design constraints, sending via SMTP relay or AWS SES for cost-effective high-volume delivery. Dynamic content personalisation inserts contact attributes, behavioural history (last product viewed, last purchase category, days since last open), and cohort-specific content blocks -- different body copy for high-value customers versus first-time buyers in the same send, without requiring two separate campaigns. Dynamic subject line and preview text personalisation improves open rates beyond generic batch-and-blast campaigns.

GDPR/CAN-SPAM/CASL compliance is built into the send infrastructure: unsubscribe processing within the legally required timeframes (CAN-SPAM requires within 10 business days; CASL is more stringent), suppression list management that prevents re-mailing unsubscribes and hard bounces, and consent record storage with timestamp and channel for audit purposes. Deliverability management includes dedicated sending domain setup, SPF and DKIM authentication, DMARC policy configuration, sending domain warmup schedules for new IP addresses, and real-time monitoring of bounce rates and spam complaint rates against ISP feedback loops (Google Postmaster Tools, Microsoft SNDS). Campaign performance analytics cover opens, clicks, conversions, and revenue attribution linked to purchase events.

Configurable lead scoring combining behavioural signals -- page visits weighted by URL intent tier, email opens and clicks, content downloads, demo requests, pricing page views -- with demographic and firmographic fit attributes from your CRM or enrichment data (Clearbit, Apollo, ZoomInfo). Two scoring approaches are supported: a rules-based additive model where each signal is assigned a point value by the marketing team, and a logistic regression model trained on historical conversion data that assigns weights based on which signals actually predicted closed-won deals in your data rather than what a marketer assumes should matter.

RFM (Recency, Frequency, Monetary) scoring segments your contact database into tiers based on last engagement date, engagement frequency, and revenue contribution -- useful for retention and re-engagement campaign targeting beyond the initial lead qualification use case. Score thresholds trigger sales handoff notifications (Slack message, CRM task creation, or email alert) and CRM status updates automatically when a contact crosses the qualification threshold. Score decay reduces a contact's score at a configurable rate when they stop engaging, so the score remains a live signal rather than a historical high-water mark. Disqualification rules remove contacts from scoring (and optionally suppress them from all outbound) when they match known non-buyer patterns -- competitor domain, free email domain lists, or specific demographic exclusions defined by your sales team.

SMS send via Twilio or AWS SNS with opt-in keyword capture, opt-out (STOP) processing, and carrier compliance (TCPA for US, CTIA messaging principles, carrier short-code and toll-free number requirements). Double opt-in flows for SMS are supported where required by regulation or carrier policy. Delivery status webhooks from Twilio track delivered, undelivered, and failed statuses per message so the platform can suppress non-deliverable numbers, retry on transient failures, and report accurate delivery rates rather than sent counts.

Push notification delivery to iOS via APNs and Android via FCM (Firebase Cloud Messaging) with token management to handle expired device tokens without manual cleanup. Web push via the Web Push Protocol covers desktop and Android Chrome subscribers. Delivery status tracking and retry logic handle APNs and FCM feedback service responses for expired or invalid tokens. GDPR-compliant consent management stores push permission grant timestamp, browser/device, and IP for audit purposes. Channel preference management ensures contacts only receive communications on channels they have explicitly opted into -- a contact who has opted out of SMS but not email doesn't receive SMS sends from any journey, enforced at the send layer rather than relying on journey logic checks. All three channels -- email, SMS, push -- are managed from the same journey canvas.

Open, click, and conversion tracking across email (via pixel tracking and click-through URL wrapping), SMS (via short link click tracking), and push channels (via notification action tracking). Conversion funnel analytics with drop-off rate at each journey step identifies exactly where contacts exit a nurture sequence so the specific underperforming step can be addressed rather than guessing. Cohort analysis groups contacts by journey entry date or segment membership to compare engagement and conversion rates across cohorts entering the same journey at different times.

Marketing attribution modelling supports five models applied to revenue events from your ecommerce platform or CRM: first-touch (100% credit to first interaction), last-touch (100% credit to final pre-conversion interaction), linear (equal credit across all interactions), time-decay (more credit to interactions closer to conversion), and data-driven attribution using Shapley values from cooperative game theory to allocate credit proportional to each channel's marginal contribution to conversions. Shapley value attribution is computed from historical multi-touch path data and is the most accurate model for understanding channel contribution, but it requires sufficient conversion volume to produce stable estimates. CDP (Customer Data Platform) architecture with event stream ingestion from Segment or RudderStack feeds the attribution model with a complete interaction record across channels, including web, mobile, email, and paid media touchpoints from the same unified timeline. Attribution data exports to your data warehouse (Snowflake, BigQuery, Redshift) for cross-channel reporting alongside paid media spend.

Bidirectional contact and activity sync with Salesforce (via the Salesforce REST API and Bulk API 2.0 for large record sets), HubSpot (via HubSpot CRM API v3), or your custom CRM via a configurable webhook or REST endpoint. Contact records in the marketing platform update when CRM lifecycle stage, deal stage, or custom fields change; CRM records update when contact behaviour in the marketing platform crosses configured thresholds (lead score change, journey completion, form submission). Activity data -- email opens, link clicks, journey entry and exit, SMS delivery status -- writes back to the CRM contact activity timeline so sales has the engagement history without switching tools.

CDP architecture with event stream ingestion from Segment or RudderStack ingests product events (feature usage, session data, in-app actions) alongside CRM and email engagement data into a unified contact profile. This enables the marketing platform to trigger journeys based on product behaviour signals -- a contact who hasn't used a key feature in 14 days triggers a re-engagement journey, not because a salesperson noticed it but because the event stream fired the trigger. GDPR and CASL compliance is built into the contact data layer: consent records store the channel, timestamp, consent basis, and version of consent language for each contact, with suppression list management that prevents messaging contacts who have withdrawn consent. Webhook support handles custom integrations with other systems in your stack -- billing platforms, customer support tools, product databases -- with data mapping configured to your field structure.