Writing

Everything we publish

Primary research we measured, guides for people running an evaluation, comparisons scored on published weights, and a blog for the shorter pieces. What none of it is: opinion posts written to a content calendar, thought leadership, or SEO filler.

Get new work by email

Roughly monthly, only when something is actually published. No sequence and no drip.

1 / 3
No content calendarSample sizes statedCorrections dated

Blog

Shorter pieces written between engagements — what we learned delivering something, or a buyer question that came up often enough to answer properly. Published irregularly, under the same editorial policy as everything else.

Research

Primary research measured from data we had access to. Every paper states its sample size in the first paragraph and names the bias in its sample.

Buyer guides

Practical guidance for people running an evaluation, including the parts that argue against buying anything.

Comparisons

Scored on published weights with no paid placement. We rank fourth on our own ERP directory.

Reference

Definitions and reference material, written to be useful during an evaluation rather than to capture search traffic.

How we decide what goes up

There is a blog at /blog, and it is worth being precise about what it is, because for a long time this page said there would never be one. What we were refusing was the content calendar: output produced on a schedule regardless of whether anything was worth saying. That output is recognisable — eight hundred words restating a well-known idea, a listicle assembled from competitors’ pages, an opinion piece with no evidence behind it.

The refusal stands. What changed is only the container. We publish when we have measured something, when we can explain something we learned delivering an engagement, or when a buyer question comes up often enough to be worth answering properly — and some of those pieces are shorter than a research paper and do not belong in the research index. Those are the posts. There is no schedule, some months produce nothing, and every post is governed by the same editorial policy as the research: state the method, state the sample size, name the bias.

A content calendar guarantees output on a schedule. It guarantees nothing about whether anything was worth saying.

What we will not publish

  • Trend pieces about AI. There is enough of that. Our AI writing is either measured accuracy figures or an explanation of what an agent structurally cannot do.
  • Roundups of competitors assembled from their websites. Our comparisons are scored against published weights by people who have worked with the products.
  • Anything where we cannot state the method. If a figure has no sample size, we do not publish the figure.
  • Guest posts and syndicated filler. Everything here is written by people who build or deliver this.

What we do publish that costs us

An industry guide telling manufacturers to buy Acumatica or NetSuite. A directory ranking us fourth. Migration guides that open by arguing you should stay where you are. A benchmarks page showing 66% on our weakest workflow and a bottom quartile well below every median.

Those are the pages worth reading first, because they are the ones we would have written differently if the goal were traffic.

Corrections

Where a published figure changes, the correction appears in the changelog with the previous value stated, and the page is re-dated. Four corrections are logged there currently, including a restatement of a number we had been using in sales conversations.

Questions

Common follow-ups.

You said there was no blog. Now there is one.
Correct, and the change is noted rather than made silently. What we refused was the content calendar, not the format. The blog carries shorter pieces — a lesson from an engagement, a buyer question worth answering properly — on no schedule, under the same editorial policy. If it ever starts producing weekly posts about AI trends, hold us to this paragraph.
Can we republish your research?
Yes, with attribution and a date. Corrections are noted rather than applied silently, which is what makes a citation stable.
Do you accept guest posts?
No. Everything here is written by people who build or deliver this software.
How often is something published?
Roughly monthly, and irregularly, across the research, the guides and the blog together. Some months nothing is worth publishing and we would rather say nothing.
Where do I start?
The pages that cost us something: the manufacturing guide telling you to buy a competitor, the directory where we rank fourth, and the benchmarks showing our weakest workflow at 66%.

Start with the pages that cost us something.

The manufacturing guide, the directory where we rank fourth, and the 66% benchmark.