Independence is a process, not a claim. This is the bar every issue must clear.
Tool feature claims link to the vendor's documentation as of the publish date. Pricing claims link to the pricing page with retrieval date. Industry statistics cite the original report — not a derivative blog summary. We don't invent numbers.
Karbon Magazine covers the Karbon ecosystem; AICPA covers the profession at large; Going Concern covers Big-4. We cover the competitor tools and edge-of-stack software those publications structurally can't. The angle isn't "we're unbiased" — it's "we cover what others can't."
Every sponsored placement carries a visible SPONSORED label and a distinguishing visual border. Sponsor copy is sponsor-written. Editorial coverage and sponsorship are decoupled — sponsoring Materiality buys placement, not editorial favour.
Every issue must contain at least one specific, actionable insight — not just description. "AI is transforming accounting" is not a claim we'll make. "Tool X works for these two scenarios but breaks on this third one" is.
When a tool falls short, we say so — with sourced evidence. When AI doesn't help, we say so. When something we recommended later disappoints, we update. The reader-correction loop catches what we miss.
At least one piece per month involves us testing a tool on real workflows, with the tool's free trial or our paid access. Synthesised reviews are valuable; tested reviews are the moat.
Every issue ends with "Spot something we got wrong? Reply and we'll publish the correction next issue." Corrections are not buried — they go at the top of the next issue, prominently. The reader-correction loop is the mechanism by which an editorial newsletter stays honest.
Any claim about a tool's pricing, feature behaviour, or accounting workflow specifics is checked against the vendor's docs as of the publish date — with a cited URL. From Month 4 onwards, accounting-specific technical content may route through a credentialed reviewer (CPA / CA / ACA). We don't pretend to be CPAs ourselves.
Some pages and issues contain affiliate links. We may earn a commission when readers sign up for a tool through our links — at no extra cost to the reader. Affiliate relationships do not influence editorial coverage. When we cover a tool we have an affiliate relationship with, the relationship is disclosed at the top of the piece in addition to per-link disclosure.
Affiliate programs we currently participate in (or have applied for):
Updated as relationships are established or end.
Materiality is built using Claude (Anthropic) for research synthesis, drafting, and verification — alongside human editorial supervision. The publisher reads and approves every issue before send. From Month 4 onwards, our masthead may include a fractional credentialed reviewer (CPA, CA, or ACA) who provides technical sign-off on accounting-specific content.
We don't pretend Claude isn't involved. We also don't pretend the system runs without human judgement. The honest framing: AI does the synthesis-heavy lifting, the publisher does the editorial-judgement work, and the result is reviewed for accuracy before send.
This is the pattern most professional B2B publications now use. The trust-deficit research suggests transparency about it earns more trust than hiding it. We agree.
If you find a factual error in any issue:
What gets corrected: factual errors, mis-stated pricing or features, mis-attributed quotes, outdated info presented as current.
What doesn't get a correction (just a clarification): editorial judgement someone disagrees with — we'll publish counter-views if substantive, but a difference of opinion isn't an error.
Sponsorship does not buy editorial coverage. When we cover a sponsor's tool, the relationship is disclosed at the top of the review.
We will decline a sponsorship if we're actively running a critical review of the sponsor's tool that issue. We will pause a sponsorship if a sponsor's behaviour makes the relationship reputationally untenable.
Full sponsorship terms, including the founding-sponsor program, are at /sponsor.
We will get things wrong. The standards above don't promise we won't — they promise that when we do, the correction is visible, prompt, and structural. That's the only sustainable answer to the AI-content trust problem in B2B media: be wrong sometimes, be honest about it always, and let readers see the process.
If something here ever fails — if you spot a buried correction, an undisclosed sponsorship, a claim without sourcing — that's a process failure. Email us.