Wide Angle Analytics Review (2026)

Wide Angle Analytics shut down on 31 August 2026; this profile is kept for reference and the service can no longer be bought. It was the strictest German privacy-first option in this directory. Where Plausible runs Hetzner Germany with a Slovenian CDN and Fathom multi-regions through AWS EU + US, Input Objects GmbH planted its entire stack on OVHcloud's EU Public Cloud and refused third-country transfers — which made it one of the cleanest Schrems II answers you could buy without going self-host.

The cookieless mechanism was also unusually well-documented. Most rivals say "we hash visitors" and stop there. Wide Angle published the literal formula: SHA-256(site + IP + User-Agent + Daily-Salt) — and explained that the daily salt lived in transient memory and rotated every 24 hours, which meant yesterday's session ID could not be reconstructed even by them. That kind of transparency is still what a DPO wants to read, and it is the part worth copying from this product.

An external DPO from dpo-consulting.com signed off on the legitimate-interest legal theory, which was rare at the $15/mo entry tier and mattered when the "no banner" decision had to be defended to procurement. The per-plan site allowances were generous: 10 sites on Starter, 20 on Plus, 100 on Advanced — agencies and portfolio operators got more sites for less than per-site rivals.

The team was tiny, the source was closed, and the price floor was high — and the shutdown on 31 August 2026 is the bus factor arriving. Input Objects GmbH did not publish a team page, a Crunchbase profile, or any funding announcements — the support page openly admitted "you may chat with the CEO, developer, or designer," which was delightful when it worked and a single point of failure when it did not. There was no public status page; outages were documented after the fact on the blog.

The product was closed-source. The github.com/inputobjects organization is empty; only thin SDK wrappers (Nuxt, Vue) were ever published. If your compliance team requires auditable code, Matomo, Umami, GoatCounter, or Plausible (AGPL) all win here — and code you can run yourself does not disappear when a vendor stops trading.

There was no free tier — Starter was $15/mo, the highest entry price among the major privacy-first names ($9 Plausible, $20 Simple Analytics, free GoatCounter or self-hosted Umami). Funnel reports, GSC connector, Slack webhooks, public dashboards, email digests, and a published tracker script weight were all either missing or undisclosed. The canonical sub-processor list sat behind the DPA — only the marketing-site iubenda policy was public.

Closed on 31 August 2026 — this section describes who it suited while it was trading, and what to look for in a replacement.

It suited: EU small and mid-size businesses and agencies with a German or French DPO who had to rule out third-country transfers entirely; portfolio operators running 5–50 sites who disliked per-site billing; teams that already trusted OVH for hosting.

The shape of the sweet spot: the $40/mo Plus plan carried 1M events, 20 sites, 24-month retention, 10 users and live chat. If you are replacing it, that combination — generous site count on one bill, EU-only hosting, long retention — is the thing to match, and it is what most per-site rivals price badly.

It never suited: US-only customers (no US hosting); hobbyists wanting a free tier (there was none); engineering teams needing self-host or open source (closed); marketers needing funnel reports, GSC integration or Slack alerts (never shipped).

Nothing here is actionable any more — the service closed on 31 August 2026. Kept because the approach is worth borrowing.

Install was a single async-defer tag — there was no SDK to bundle. WordPress, Wix, Squarespace, Webflow, Ghost and Carrd were documented as one-click integrations.

The interesting setup option was the first-party proxy via custom domain — you pointed a subdomain (e.g. analytics.yoursite.com) at Wide Angle's tracker host so the script and beacon hit your own origin, which side-stepped ad-blockers and content-security-policies. Several live vendors offer the same thing today, and it is the setting worth looking for when you replace this product.

Per-script tracking flags were configurable inline: data-waa-fingerprint='true' enabled canvas-fingerprint mode (off by default), data-waa-cid='...' injected a logged-in customer ID, data-waa-inc-params='id,session' whitelisted URL params, data-waa-exc-paths='/wp-admin/*' excluded paths. All of it was server-validated against the per-site config in the dashboard.

The migration that matters now is off Wide Angle, not onto it: the service stopped operating on 31 August 2026 and the vendor issued a downloadable snapshot of customer data rather than an in-app export. If you still hold one of those snapshots, keep it — there is no way to pull the data again.

Moving in never had a GA4 importer. Migration was re-tagging-only: drop the GA4 tag, drop in the Wide Angle script, restart from zero. There was no historical-data ingest path, which is the same posture as Plausible and Fathom.

The banner-removal argument was the real draw: because the product shipped under legitimate interest with no personal data collected, EU sites could typically retire the analytics-purpose cookie banner entirely (marketing and advertising cookies stayed a separate question). Any replacement you pick has to carry the same legal reasoning if you want to keep the banner off.

Custom-event mapping required planning. GA4 has thousands of automatic events; Wide Angle shipped pageviews, outbound clicks, downloads and named "Custom Actions", wired through waa('track', 'event_name', {...}). UTM tracking was automatic.