Every integration monitored by outcome
Not "did the request return 200" but "did a lead arrive in the CRM in the last hour". Alerting on the business outcome catches the failures that a status-code check sails past.
Web Development
API & Systems Integration
We connect the systems that are not talking to each other: website to CRM, CRM to ERP, booking to calendar, payments to accounting. Built with retries, alerting and a runbook.
Free proposal · No contracts over 30 days · Reply within one business day
- Integrations shipped with outcome-level alerting
- 100%Integrations shipped with outcome-level alerting
- Duplicate records from retry logic in production
- 0Duplicate records from retry logic in production
- Typical detection time for a broken integration
- <1 hrTypical detection time for a broken integration
The problem
The integration that quietly stops working is the expensive one. A token expires, an endpoint changes, and leads stop reaching the CRM — but nothing errors visibly, so the first sign is a sales team wondering why the month is quiet.
How we do it
What the work actually involves
Idempotency and replay
Retries that cannot create duplicate records, and a dead-letter queue you can replay once the far end is fixed. Data lost during an outage should be recoverable, not gone.
Field mapping agreed before the build
A written map of which field goes where, what happens on conflict, and what a required-but-missing value does. Most integration disputes are really disagreements about this document not existing.
Included
What you actually receive
- Systems and field-mapping document
- Integration build with retries, timeouts and idempotency
- Dead-letter handling with replay
- Outcome-level monitoring and alerting
- Runbook covering credentials, rotation and failure response
Tools we use
- Node.js
- HubSpot
- Salesforce
- Zapier
- Make
- Stripe
- Twilio
Common failures
Where this usually goes wrong
The three we are called in to fix most often. If any of these describe your current setup, that is the first thing we would change.
Assuming the third party is up
Every external call fails eventually. Without timeouts, retries and a fallback, one slow vendor takes your checkout down and the logs point at your own server.
Syncing everything on a schedule
A nightly full sync is simple until the data set grows and the window does not. Event-driven updates are more work up front and stop being a problem at scale.
Credentials in the codebase
API keys in a repository are a breach waiting for the repository to be shared, forked or made public. They belong in environment configuration, rotated on a schedule.
Measurement
How you will know it is working
Agreed before we start, so progress is visible in the months before it shows up in revenue.
- Integration error rate
- Failed calls as a share of attempts, per vendor. It shows which dependency is actually costing you.
- Ongoing
- Data freshness
- Lag between a change at the source and the change appearing in your system. Agreed as a target before build.
- Pre-launch
- Recovery without intervention
- Share of failures the system retries successfully on its own. If it is low, someone is being paged for something the code should handle.
- Month 1
Questions
Any question in your mind?
The questions we get asked most, answered without the sales gloss. If yours is not here, ask us directly — you will get a straight answer either way.
Still have questions?
Book 30 minutes with a strategist. No deck, no pitch — just answers about your situation.
Book a callOften, and we will recommend it when it fits — low volume, standard connectors, and a workflow your team wants to change themselves. Custom code earns its place at high volume, with unusual field logic, or where per-task pricing has started to cost more than the build would.
Looking for pricing, contract terms or timelines?
Read all FAQsKeep reading
More web development work
- Next.js DevelopmentProduction Next.js builds with server rendering, image optimisation and Core Web Vitals budgets enforced in CI.Read
- Node.js DevelopmentAPIs, integrations and back-end services in Node — built to be handed over, monitored and maintained by your team.Read
- React DevelopmentComponent-driven front ends and web applications in React, built on a design system rather than a pile of one-off components.Read
- Website RedesignA redesign that keeps the traffic you already have and fixes the reason the old site was not converting.Read
- Website Speed OptimisationCore Web Vitals work with the diagnosis written down first — so you can see what each fix is worth before paying for it.Read
- Headless CMS DevelopmentSanity, Payload or headless WordPress wired to a modern front end — with an editing experience built around your team.Read
- Web Application DevelopmentPortals, dashboards and internal tools — scoped tightly, shipped in phases, handed over with documentation.Read
- UI/UX DesignInterface design grounded in how people actually use the thing — research, prototypes, and a system that survives handover.Read
- Landing Page DesignCampaign landing pages built to convert paid traffic, with message match to the ad and a test plan from day one.Read
- Marketing AutomationLifecycle flows — welcome, abandonment, onboarding, win-back — that run without anyone pressing send.Read
Ready when you are
Ready to Build Your Next Big Project?
Tell us what you are trying to grow. You will get a straight answer on whether we can help, and what it would cost, within one business day.