Editorial process

How we research, write, and correct what we publish.

GainTrace writes about customer health, churn, and expansion revenue — a subject where a wrong number costs someone a renewal or a budget conversation. This page documents how an article gets made, where our figures come from, what AI does and doesn't do here, and how to tell us when we get something wrong.

A human decides what publishes. Every time.

We use AI the way any modern team does — for research, structure, and editing. It never approves a draft, never verifies a claim, and never publishes. Every article on this blog is read end to end by a named person who is accountable for it, and every product claim, price, and statistic is checked against its original source before it ships.

Our editorial principles

Four rules that decide what runs and what gets held.

Practitioner-written

Every article is written by someone who has run the motion it describes — customer success, RevOps, or the product engineering behind them. No outsourced content mill, no ghostwritten filler.

Sourced and dated

Numbers carry their provenance. A figure is either verified by us against a public source on a stated date, or attributed to the third party that published it. Anything we can't source, we don't print.

Named and accountable

Articles carry a named author with a real role, a publish date, and an "Updated" date once the piece has meaningfully changed. You can always see who wrote it and when it was last checked.

Commercially honest

No paid placements, no sponsored opinions, no bought reviews. When a competitor is a better fit than GainTrace for your situation, the article says so.

Who we write for

Every brief names one of these readers. If it doesn't, it isn't a brief.

Customer success leaders

CSMs and CS managers who own retention, health scoring, and QBRs — and need a system that doesn't require a full-time admin.

RevOps and ops teams

The people wiring product usage, billing, and support data together, and making the routing and alerting actually fire.

Founders and CEOs

Operators at seed-to-Series-B B2B SaaS companies deciding what to buy, what to build, and when a CS platform starts paying for itself.

Sales leaders

Teams working expansion and renewal pipeline who need earlier, better-qualified signals than a renewal-date spreadsheet.

Finance and CFOs

Anyone forecasting NRR, modelling the cost of churn, or defending a retention number to a board.

Buyers doing research

People mid-evaluation comparing platforms, prices, and contracts — the audience our comparison and pricing articles are written for.

How we build an article

Six steps, in order. Nothing skips step five.

  1. Start with a question someone actually asked

    Topics come from three places: questions that come up in demos and support conversations, search demand from people mid-evaluation, and changes in the market — a vendor repricing, a platform shipping something material. If we can't name the reader and the decision they're trying to make, we don't brief it.

  2. Brief it against what already exists

    We read what currently ranks for the question and write down what those pages get wrong, leave out, or leave stale. If we can't add something the reader can't already get — a first-hand number, a tested workflow, a decision framework — we don't publish another version of the same page.

  3. Research and record every source

    Pricing pages and public vendor documentation, published industry research, review platforms, and our own operating experience. Each figure gets recorded with where it came from and the date it was checked, before a draft is written — not reconstructed afterwards.

  4. Draft for the decision, not the word count

    Direct language, specific recommendations, and the tradeoffs stated out loud. Steps a reader can follow. Tables where a table is genuinely clearer than prose. We cut padding rather than pad to hit a length.

  5. Review against five checks

    A human reviews every draft before it publishes — no exceptions, no automated publishing path exists. A piece is held or rewritten if it fails any of the five checks below.

    • Accuracy. Every claim traced back to its source; every price re-checked against the vendor's live page.
    • Depth. Does it go past what the reader could have guessed? Surface-level sections get cut, not expanded.
    • Clarity. Technical and financial concepts explained in plain language, defined on first use.
    • Usefulness. Instructions, formulas, and workflows are walked through before they ship.
    • Honesty. Not too promotional, not self-serving in its rankings, no claim we couldn't defend to the vendor we're writing about.
  6. Publication checks

    Headings in a clean hierarchy, links pointing where they claim to, tables readable on a phone, structured data valid, and the author and dates rendering correctly. Then it goes live and gets submitted for indexing.

How we handle numbers and pricing

Our most-read articles are about what software costs. Every figure in them carries one of three labels internally — and the first two are the only ones that reach the page.

Verified

We checked the figure ourselves against a public source — a vendor's own pricing page, documentation, or a public marketplace — and we state the date we checked it. Prices move; a number without a date is a number you can't trust.

Reported

The figure comes from a third party, and we name that third party in the sentence that uses it. If we can't name the source, the number doesn't run.

Never

No invented statistics, no rounded-up vendor marketing claims presented as research, no unattributed quotes, no borrowed benchmark whose original study we haven't read.

Vendors change pricing without announcements, and published figures go stale fast. If you find a price on this blog that no longer matches the vendor's live page, tell us and we'll re-verify it — see corrections below.

How we rank and compare tools

We publish comparisons and “best tool” lists in a category we compete in. That's a conflict of interest, so it runs under stated rules.

  • The methodology is stated in the article — what we compared, on what criteria, and what we couldn't assess.
  • GainTrace is placed honestly. If we're not the right fit for a segment in the list, the list says which tool is.
  • No vendor pays to appear, to move up, or to stay out. There are no affiliate arrangements behind our rankings.
  • We don't bump the date on a list without changing its substance. An "updated" date means the content was re-checked.
  • Pricing in a comparison follows the same sourcing rules as everywhere else: verified with a date, or attributed.

How we keep articles current

Publishing is the start of an article's life, not the end of it.

ReviewCadenceWhat it covers
Performance reviewQuarterlyIs the article still answering the question people arrive with? Sections that miss the intent get rewritten.
Pricing and market checkOn change, and at minimum twice a yearCompetitor pricing, packaging, and positioning move constantly. Every price we quote gets re-verified and re-dated.
Product and platform updatesOn releaseWhen GainTrace ships something that changes an answer, the articles that gave the old answer get updated.

When a piece changes materially, its “Updated” date changes with it. When nothing changed, the date doesn't move — a refreshed date on this blog means the content was actually re-checked.

Corrections and reader feedback

We get things wrong. When we do, we'd rather hear it from you than leave it up.

What we'll fix

  • A factual error, a misquoted source, or a number that never had one.
  • A price or plan that has changed since we verified it.
  • An unfair or out-of-date characterisation of a competitor's product.
  • A broken link, a dead reference, or missing context that changes the meaning.

How to tell us

Email hello@gaintrace.com with the article URL and what's wrong. Vendors are welcome to write in about their own product; we'll verify and correct on the same terms as anyone else.

Substantive corrections are made to the article itself and the update date is changed. We don't quietly delete pages to make an error disappear.

How we use AI

We build an AI product, so we'll be specific rather than vague about this.

What AI does here

  • Organise research and surface gaps in an outline
  • Suggest structure, tighten wording, and flag unclear passages
  • Help find sources — which are then opened and read by a person
  • Check drafts for readability, heading structure, and internal-link opportunities
  • Assist with routine formatting and metadata work

What AI never does

  • Decide what gets published, or publish anything itself
  • Produce a statistic, price, or quote that a human hasn't verified at its source
  • Replace the practitioner judgement in a recommendation
  • Generate articles at volume — we publish a small number of pieces we can stand behind, not a content farm
  • Write anything that ships without a named human reviewing it end to end

When we mention our own product

GainTrace is a customer success platform, and this blog exists partly to reach people who might buy it. Pretending otherwise would be the least trustworthy thing on the page. So the rules are simple:

  • Product mentions are labelled as what they are. We don't disguise a pitch as neutral advice.
  • Educational articles stand on their own. If you never become a customer, the article still has to have been worth your time — that's the bar every draft is held to.
  • Comparisons explain fit and tradeoffs, including where a competitor is the better choice.
  • We don't make guarantees about results — not for our product, not for anyone else's.

Who's accountable

The people whose names go on the work — each with a profile and everything they've written.

Hold us to it.

Found a number that's moved, a claim that doesn't hold up, or a comparison that isn't fair to your product? Write to us — corrections are the fastest way this blog gets better.