Guides

Written from the projects that went wrong

These guides come out of rescue engagements rather than out of marketing. That means they are specific, occasionally unflattering to us, and willing to tell you the answer is to buy nothing — which is the answer a meaningful share of the time.

Get a second opinion

Send your shortlist, requirements, or a vendor quote. A person reads it and replies in writing.

1 / 3
Vendor-neutralSample sizes statedIncludes when not to buy

The guides

Two published, six in progress.

We publish a guide when we have enough engagement evidence to say something specific, rather than restating what every vendor blog already says.

ERP buyer's guide

How to run a selection that survives implementation — the four decisions, the eight questions, and when to buy nothing.

Read the guide →

Why ERP projects fail

Primary cause across 41 stalled deployments. Only four percent were the software.

Read the guide →

ERP pricing explained

Licence, implementation, integration, and the internal cost nobody books. What each vendor actually charges for.

In progress

Implementation cost & timeline

What a mid-market implementation really costs and how long it really takes, by system.

In progress

Total cost of ownership

Modelling three years properly, including the lines vendors leave out of a comparison.

In progress

When to leave QuickBooks

The three signals that predict it, and the two fixes that do not require replacing your ledger.

In progress

ERP requirements checklist

A requirements document that produces useful demos rather than a feature bingo card.

In progress

ERP selection process

Eleven to fourteen weeks, phase by phase, with what to do in each.

In progress

Why most ERP content is useless

Search almost any ERP question and you get the same article rewritten forty times: a definition, a benefits list, a feature comparison that discriminates between nothing, and a call to book a demo. It is produced at volume because it ranks, and it is useless because it was never written to be read.

The tell is that it never says anything that costs the author something. No vendor blog will tell you the product is the cause of failure four percent of the time, that you should ask for a reference who left, or that a meaningful share of ERP searches should end with buying nothing. Those are all true and all commercially inconvenient.

If a guide never says anything that costs its author a sale, it was not written to help you decide.

What we try to do instead

  • State the sample. Where a number comes from our engagements, we say how many and over what period. Forty-one is a small sample and reads as one.
  • Disclose the interest. We sell an ERP platform and implementation services. That is on every guide that touches a buying decision.
  • Include the case against us. Every comparison names where the other product wins, and the industry guides name where we are the wrong answer.
  • Be specific enough to be wrong. A guide that cannot be disagreed with is not making a claim. If we say a QuickBooks Online conversion takes three to four weeks, you can hold us to it.

Where these come from

Most of the material originates in diagnostics — the two days we spend inside a stalled or unhappy deployment before quoting. That is an unusually good vantage point, because you see what actually happened rather than what the project plan said would happen, and you see it across many companies with different systems.

It also biases the material toward failure modes, which is worth naming. A company that bought well and implemented smoothly never calls us, so our view of this category is skewed toward the projects that struggled. Read the guides as field notes from that vantage point rather than as a survey.

Corrections welcome

If something here is wrong, tell us. Factual errors get fixed and dated with a note saying what changed. We do not silently edit.

Questions

About these guides.

Are these gated?
No. Everything renders in full without an email. We ask for one only if you want something sent to you.
Can we republish or cite them?
Yes, with attribution and a link. The underlying directory data is published under a permissive licence too.
How current are they?
Reviewed quarterly, with a last-reviewed date on pages where currency matters. Anything drawn from engagement data states the period it covers.
Why so few guides?
Because we publish when we have something specific to say. Filling out the set with generic articles would rank and would undermine the reason to read any of them.

Have a question these do not answer?

Send it. A person reads it and replies in writing, whether or not there is a deal in it.