Member System API

Situation before deployment

We have around 10,000 members. Five systems—EC site, billing system, customer support tool, email marketing, and analytics platform—each held member data, synchronized via overnight batches.

When a member updated their address or plan, the change was reflected first on the EC site, but a lag of up to a day appeared in other systems. Support staff receiving inquiries would sometimes reference stale information and issue guidance that had to be corrected later, causing dozens of such cases per month.

Multiple overnight batches ran between systems, and if one failed midway, other systems could be left with stale data. Batch operations monitoring effort and incident response were both rising.

Background and selection reasons

We also compared CDP products and building a member platform from scratch in-house. CDPs lean toward analytics and are overkill as CRUD infrastructure for business systems. In-house build was hard to estimate for both development time and maintenance cost.

We chose NocodilySuite because it fit the requirement: holding member data as a master and providing REST APIs and Webhooks out of the box. The existing five systems could simply reference the Member API, dramatically simplifying inter-system data synchronization.

In the PoC we first switched only read-only access from the EC site and validated response time and stability. After running in parallel with the existing batches and verifying data consistency, we moved to full migration including writes.

Deployment process

Four months in total. Month one covered PoC and API spec design, months two and three migrated the member ledger and implemented 24 endpoints, and month four switched the five systems over. We ran the overnight batches in parallel for one month, checking daily data diffs before full cutover.

The endpoints cover member search, detail retrieval, registration, update, cancellation, plus references for plan changes and payment status—24 total. Write-side operations were designed to fire real-time Webhooks, and each system uniformly updates its own cache when it receives a notification.

We did not eliminate the existing overnight batches; they now run daily as a diff-sync fallback if a Webhook doesn't reach its destination. Webhook delivery success is monitored, and if failures persist, a Slack notification is sent.

Changes after deployment

Discrepancies in member information across systems disappeared, and information corrections during support responses dropped to near zero. Member-initiated changes now reflect immediately across all channels, improving experience consistency.

The number of overnight batches was cut by 85%. What remains is only the diff-sync fallback, greatly reducing the operational monitoring burden. Morning incident response for batch failures also dropped to about 0–1 cases per month.

When adding a new integration target, development now proceeds simply by handing over the API spec and Webhook endpoint. Integration with marketing analytics tools and external services now finishes in about two weeks, letting campaigns launch faster.

Future Plans

In the next phase we plan integration with a CDP and delivery of member behavior data to a real-time analytics platform. The Member API will act as a hub, letting marketing and operations share data.