This release is about more, longer, sharper. More companies you can track at once. Longer observation windows on each one. And a sharper bar on what we surface as a signal. Plus the first real credit system, top-up packs that roll over with bulk pricing baked in.
Five times the monitoring
We bumped the per-user monitoring cap from 20 slots to 100. That's the largest single capacity change since the product launched, and it lands without a price increase on Pro. The infrastructure can already carry the load. The limit was a deliberate safety valve while we tuned the re-crawl scheduler. The valve is now wide open.
100 is the new floor on Pro. We're already working through what it takes to push that to 200, then 500, while keeping the same crawl freshness guarantees we have today. The next bump is queued.
A real credit system, finally
Until this release Surfacer only billed by subscription, with a fixed daily allowance baked into each plan. If you hit it on a Tuesday, you waited for Wednesday's bucket. There was no way to buy extra credits, no way to add monitoring slots on top, and nothing rolled over. That stops now. Top-up packs land as the first credit-on-demand option, and the credits they buy never expire.
Rollover credits stack on top of your monthly Pro allowance and only get drawn after the monthly bucket runs dry, so committing once buys you a long runway. Bigger packs are deliberately cheaper per credit, which means committing rewards you instead of just pre-paying.
More signal families, observed over time
The shape of our product is shifting from "tell me about a company once" to "watch a company every day and tell me what changed." That requires a deeper observation cadence and a wider net of signal families re-evaluated on a daily rhythm. This release extends nightly re-evaluation to several categories that used to update less frequently:
- Pricing-page drift: semantic diff against last night's snapshot, surfaces any real pricing-page change as a signal.
- Hiring intent: open positions tracked daily, with role types categorised so a "VP Sales" lands harder than a "Junior Backend Engineer."
- CMS migrations: fingerprint changes on the homepage trigger a re-tagging pass.
- Leadership moves: managing-director and board-member changes detected via Impressum + Handelsregister diffs.
A monitored company that does nothing for two weeks gives you a quiet feed and a stale row. A monitored company that hires a VP Sales, redesigns its pricing page, and announces a funding round in the same fortnight gives you a top-of-feed call card with three reasons attached. The product is finally good at telling the difference.
Quality keeps compounding
More monitoring and more signal families would be a noise machine if the underlying scorer didn't keep getting sharper. The verifier pass that landed this release tightens the precision band on hiring-spike detection. We're now correctly flagging directional hiring (a new team forming, not backfill churn) on roughly 88% of cases in the validation set, up from 74% last month. Funding-round detection precision moved similarly.
Also in this release
- New public pages. /platform shows the dashboard, feed, monitoring, and company-detail surfaces in HTML/CSS mockups; /about lays out the manifesto and the team; this page is now itself public so prospects can read the changelog before signing up.
- Shared marketing nav + footer. Landing, pricing, platform, about, and what's-new all share the same chrome now.
- Black primary CTAs. Accent-fill buttons on the marketing pages flipped to the dark
.btn-primarystyle so primary actions are visually consistent everywhere. - Amber by default. The public accent picker is gone; amber is the new global default. Existing users who picked green or purple keep their choice.
What's next
Quality is our key metric we are aiming for next. The hiring-spike verifier in the chart above shows what a compounding precision pass looks like, and the next release applies the same approach to funding detection, leadership-move classification, and pricing-drift semantic diffing. Each one trades a few weeks of verifier work for fewer false positives in your morning call list.
Alongside the precision work, two capacity pushes are queued. Faster-than-daily re-crawls on the highest-signal monitored companies, so a real change shows up in your feed within hours, not the next morning. And a path to lift the per-user monitoring cap well past 100 once the scheduler holds up at higher fan-out.