Company

What we believe

Every company has a values page and most of them are interchangeable. These are positions rather than values — each one is falsifiable, each one shapes a product decision, and several of them cost us business.

Disagree with any of these?

Tell us which one and why. Several of these have been revised on customer pushback.

1 / 3
Positions, not valuesEach one costs us somethingFalsifiable by our own pages

Company

Eight positions.

If you disagree with three or more of these, we are probably not the right vendor for you, and that is a useful thing to establish on a page rather than in month four.

The ledger is not negotiable

Append-only, double-entry, corrections as reversals. Nothing edits a posted transaction, including us. Every convenience we could offer by relaxing this is a convenience that makes an audit trail meaningless.

Authority must be granted, never assumed

Automation is only safe when someone decided exactly what it may do. An agent that could do more than you explicitly permitted is a liability regardless of how well it performs.

Some things stay manual forever

Payment release, vendor banking, permission grants, period close, statutory filing. We will not enable these for an agent at any price, and that is the one decision we take away from customers.

Measurement beats assertion

Publish the accuracy, the spread, and the bottom quartile. A number without a method is marketing, and a median without a spread sets an expectation half of customers will not meet.

Your data is yours, testably

Export on a schedule you set, to storage you own, in open formats, working while everything is fine. An export you can only get by asking is a negotiation that starts at the worst moment.

Saying no is part of the product

Manufacturing, warehouse management, multi-country statutory. We publish an industry guide telling manufacturers to buy a competitor, because a hedged version wastes your time and ends badly.

Integration usually beats migration

Most companies do not need a new ledger. They need eight systems to agree. That engagement is a fraction of what a migration would earn us and it is the right first answer for most buyers.

The goal is not needing us

A customer who raises how-to tickets six months in is a customer we failed to enable. Consultancies with a different model call that recurring revenue.

What each of these costs

A position that costs nothing is a preference. These are the specific prices we pay, which is the part that makes them checkable.

  • The append-only ledger means more storage, slower writes, and messier-looking accounts when something is corrected. Customers occasionally ask us to just fix an entry and we decline.
  • Bounded agent authority means we lose deals to vendors implying broader automation. We cannot match a claim of full autonomy because we will not build it.
  • Publishing the spread costs us against competitors quoting a single higher figure with no method behind it. Our worst published workflow is 66%.
  • Telling manufacturers to leave forgoes an entire market segment we could have taken demos from.
  • Recommending integration first is a fraction of migration revenue, and a meaningful share of those customers never migrate at all.
  • The three-closed-months rule has delayed cutovers by a quarter, which delays revenue recognition, and we have enforced it anyway.
  • No vendor money means declining affiliate programmes on pages that would earn from them.
  • Enabling customers properly means fewer support hours to bill and a faster path to them not needing us.
A position that costs nothing is a preference. These are the invoices we do not send because of them.

What would change our minds

These are not permanent. Several have already been revised — the export scheduling capability exists because a customer pointed out that an on-request export is not a real exit, and our benchmark reporting includes the bottom quartile because a prospect said a median alone was useless for planning.

The ones least likely to move are the ledger and the agent authority ceiling, because both are structural. Relaxing either would mean rebuilding the guarantees that everything else depends on, and a system where those guarantees are configurable does not have them.

If you think one of these is wrong, say so. We would rather revise a position on evidence than defend it because it is on a page.

What we are not claiming

That we are better than the incumbents. We rank fourth on our own ERP directory and seventh on our own inventory directory, and both rankings are accurate.

That these positions make us safe to buy. We are a young company with less operating history than anyone we compete with, no SOC 2 Type II report, and a partner network of almost nobody. Positions do not offset that; a tested data export does, partially.

That we always live up to them. We will get things wrong, and our error policy commits us to telling you when we do — including errors you would probably never have noticed.

How to check any of this

Do not take the page at face value. Open our ERP directory and see where we rank. Read the manufacturing guide that names two competitors to buy instead. Look at the 66% figure on our benchmarks. Every claim on this page has a page behind it that would be embarrassing if the claim were false.

Questions

Common follow-ups.

Is this just a values page?
These are positions rather than values — each is falsifiable, each shapes a product decision, and each has a stated cost. A value nobody could disagree with is not a position.
Which of these would you never change?
The append-only ledger and the agent authority ceiling. Both are structural, and a system where those guarantees are configurable does not have them.
Have any been revised?
Yes. Scheduled export exists because a customer said an on-request export is not a real exit. Bottom-quartile reporting exists because a prospect said a median alone was useless.
Do these make you a safe purchase?
No. We are young, have no SOC 2 Type II, and almost no partner network. Positions do not offset that — a tested data export partially does.
What if we disagree with several?
Then we are probably not the right vendor, and establishing that here is better than establishing it in month four.

Check the claims rather than believing them.

Every position here has a page behind it that would be embarrassing if the position were false.