December 2025: a year-in-review release (and what 2026 looks like).

December is the only month of the year we don't ship features. We freeze the codebase on the 15th, the team takes time off, and the only release activity is bug fixes via on-call. So instead of a feature post, here's the actual look back: what landed in 2025, what we learned from it, and what 2026 is shaped like.

Four things that shipped in 2025

1. Fixed-cost lifecycle dates (November, v2.0)

The single biggest shift in how nouz models a small business this year. Before November, a fixed cost was a single recurring amount with no concept of when it started or ended. After November, every fixed cost has a start_date and optional end_date, and the daily slice only applies if the date falls in that range. This made mid-cycle rent increases, contract switches, and short-term staff additions actually representable. See fixed-cost lifecycle dates and the November release notes.

2. The Statistics tab (June, v1.5)

nouz spent its first year as a pure "today's number" tool. The Statistics tab in June was the first real time-series layer — trend charts, peak-day detection, break-even visualisation. We rebuilt most of it in February 2026 (next release), but the original shipping of the tab was the moment nouz stopped being a logger and started being something owners checked even on quiet weeks.

3. Mobile entry, properly (March, v1.4)

The first nine months of nouz worked on mobile in the sense that the pages rendered. They didn't really work in the sense that owners would close their day on a phone. The March release rebuilt every form for thumb-reach and tap targets, added the mobile-first home screen, and got daily entry-on-mobile up from 12% of sessions to 51%. Still room to grow — the April 2026 release brings native camera invoice scan.

4. The product card and COGS snapshot (January, v1.3)

The decision that every product sale should capture its COGS at the moment of sale — not look it up live — was made in January. It's the reason editing a product today doesn't rewrite yesterday's P&L, and the reason archived products still appear correctly in historical reports. Reworked again in January 2026 with the rewritten product card; the underlying snapshot model is unchanged.

Four things planned for 2026

1. Multi-location reporting (May)

Growth and Pro have supported multiple locations all year, but cross-location reporting has been thin — basically a location switcher and nothing else. The May release brings side-by-side compare, a roll-up monthly P&L for the whole business, and a per-location contribution view. For multi-currency businesses (now supported as of February 2026), the roll-up converts to a single reporting currency using ECB rates.

2. Bank and processor auto-import (June/July)

The single most common manual entry today is "card revenue from Stripe / Mollie / SumUp." June ships a direct connector for Stripe and Mollie that auto-imports settlement-date revenue, net of fees, into the daily P&L. July extends to SumUp, Zettle and Adyen. The keystroke this saves for high-volume shops is huge. For the why-it-matters context, see BAKED. Berlin, who logs 300 card transactions a day.

3. Mobile widget for today's EBIT (July)

iOS and Android home-screen widgets showing today's running EBIT, no app open required. The most-requested feature on the public roadmap, and one we kept punting because the system-level work was higher than expected. v2.7 makes it real.

4. Daily-close email digest (August)

A short end-of-day email with the day's EBIT, the headline mover (best product, weakest day-of-week comparison, biggest expense), and any insight tag that fired. Optional, off by default, designed for owners who want one pull-style update rather than opening the app. Replaces the morning-after-yesterday accounting habit with same-day numbers — see how the P&L settles.

What we got wrong (the year edition)

Two larger misreads from 2025.

• We launched without fixed-cost dates. The first 11 months of nouz treated fixed costs as "current and forever." Anyone who changed rent, switched insurance, or hired/lost staff mid-year had to manually correct historical slices. We knew this on launch day and shipped it anyway. It cost us about a dozen churns we could have prevented. Fixed in November.
• The Statistics tab launched as a wall of metrics. Time-on-tab in the first three months after the June launch was 22 seconds. We treated it as a discovery problem and added tooltips. It wasn't a discovery problem — it was a story problem. The February 2026 rebuild around five owner questions (instead of a grid of charts) fixed it. We should have done it that way the first time.

What's next (January 2026)

January's release is a rewrite of the product card — the screen you open when you click a single SKU. New layout around three sections (identity, performance, snapshot history), faster COGS snapshot writes, and a cleaner archive flow. Smaller in scope than the year highlights above; bigger in code than it looks.

For anyone who joined nouz this year and hasn't poked through the whole product yet, the product overview is the cleanest tour. Thanks for shipping a real year with us.

FAQ