Social Media Management Tool

01The challenge:

A scheduler that dropped posts at the busiest hour of the week

BuzzMetrics began inside a small social media agency as a tool for its own clients, and 600 small businesses had signed up for the free trial. The prototype published everything from a single cron job. On Monday mornings it fell behind, and when a platform token expired, the post never went out.

Trial users noticed before the founders did. Almost none converted to paid, support answered the same missed-post complaint every week, and the analytics page pulled live from each network's API, so it took twenty seconds to open.

02What we built:

Every post is a queued job, and every chart reads from a cache

The app is Next.js on the front and Node.js behind it, with accounts, posts and results stored in MongoDB. Scheduled posts are jobs in a Redis-backed queue that retries on failure, spreads the Monday peak across workers and flags an expiring token days before it can break a post.

Analytics are pulled from each network overnight and on demand, then cached in Redis, so dashboards open in under a second. OpenAI drafts caption options from a customer's best-performing posts and brand notes, and the customer picks, edits or ignores them before anything is scheduled.

03What changed:

From a leaky free trial to 10,000 paying customers

BuzzMetrics passed 10,000 paying customers about eighteen months after the paid launch. The queue now publishes around 1.4 million posts a month, and 99.9% go out within a minute of their scheduled time. Missed-post tickets, once the top support category, fell out of the top ten.

BuzzMetrics' product manager sets the roadmap. We remain the engineering team behind it, shipping weekly and watching queue depth and API error rates across the networks, and we added a second on-call engineer once the customer count passed five thousand.