Heap Alternative: Autocapture Without the Enterprise Bill
Heap pioneered autocapture. What that idea looks like in a lightweight, cookieless tool — clicks, pageviews, and errors captured automatically.
Heap's founding insight was correct and ahead of its time: nobody knows in advance which interactions will matter, so capture everything and define events retroactively. The insight outlived the pricing model — Heap went enterprise (and into the Contentsquare family), and the teams who loved autocapture's promise now face quotes sized for companies with analytics committees. The good news: the insight itself got cheap.
What autocapture got right
- Zero instrumentation lag: the data exists before the question does.
- No 'we forgot to track it' regret when a new question arrives.
- Engineers stop being a bottleneck for marketing's curiosity.
What it got wrong — worth admitting
- Weight: capturing every DOM interaction costs script size and main-thread time — a real Core Web Vitals tax.
- Noise: retroactive event definitions on raw interaction soup produce fragile, selector-based 'virtual events' that break with every redesign.
- Privacy surface: indiscriminate capture hoovers up form text and PII unless carefully configured — the opposite of data minimization.
The middle path: curated autocapture
Clycyo's answer is to autocapture the layer that is always meaningful and never sensitive: pageviews (including SPA routes), clicks, page-load time, JavaScript errors, Web Vitals, referrers, and UTM context — automatically, in a 1.1 KB script, cookieless. Then the handful of business-defined moments (signup, activation, payment) are explicit one-line track() events, named by humans, stable across redesigns.
You keep autocapture's core promise — when a new question arrives, the click and navigation history is already there on each visitor's timeline — without the noise pile or the consent problem. And the explicit events give reports a vocabulary that selector-scraping never achieves.
Side by side
| Dimension | Heap | Clycyo |
|---|---|---|
| Clicks & navigation captured automatically | Yes (everything) | Yes (curated layer) |
| Retroactive event definitions | Yes — its signature | No; explicit events instead |
| Cookieless / no consent banner | No | Yes |
| Per-visit performance + errors | No | Yes |
| Revenue attribution via webhook | Indirect | Built in |
| Pricing model | Enterprise sales | Free tier + self-serve |
Who should still pick Heap
Large organizations whose analysts genuinely mine raw interaction data and who have governance to manage the privacy surface. If instead you adopted Heap so that 'we'd have the data when we need it' — the curated layer plus explicit events delivers that promise at startup weight and price. Watch it on live traffic at /open: every click and pageview you see there arrived with zero instrumentation.