You have a number. Here is what we would do with it.
The Step Audit estimates from the run counts you gave it. An assessment replaces those estimates with your real execution history, finds what is actually driving the bill, and hands back a fix list in priority order — before the billing change, not after the first invoice.
Certified Atlassian Partner · 25+ years of DevOps delivery · Westminster, CO
What the assessment is
A fixed-scope review of one Jira or Confluence Automation estate. It is a piece of work with an end, not an open engagement, and the output is a document you can act on without us.
We measure what actually runs. Your automation audit log over a representative month, joined to your rules export. Not estimates — the real count of how often each rule fired and what it did when it got there.
We rank every rule by what it spends. Steps per run multiplied by real runs per month, against your pooled allowance and seat count. That ordering is usually the surprise: the rules people worry about are rarely the expensive ones.
We find the structural causes. Triggers too broad for their conditions, branches with no bound, rules quietly triggering each other, work being done before the condition that would have stopped it.
We write it down in priority order. Each finding with the specific change, the steps it recovers, and what it risks. Ordered so the first few items pay for most of the reduction.
What you get
A written findings report — every rule ranked by real monthly spend, with the estate's total against your allowance.
A prioritised remediation list — each item naming the rule, the change, and the steps it recovers.
A projection of where the estate lands after the changes, on the same numbers.
A walkthrough with whoever owns the rules, so the list leaves with someone who can act on it.
You can do all of this yourself. The Step Audit gives you the ranking for free, and the FAQ on it explains every pattern and its fix in full. This exists for teams who would rather not spend the week, or who want the numbers checked by somebody who has done it before.
What we need from you
Less than most people expect. Two exports and one conversation.
Your automation rules export — the same automation-rules-*.json the Step Audit reads.
Your automation audit log for a representative month. Avoid a month containing a migration or a bulk edit; it makes an outlier look like a norm.
Your plan tier and licensed seat count for Jira and Confluence.
Half an hour with whoever owns the rules, so findings land with context rather than as a list of rule names.
No access to your instance is required to produce the findings. If you would rather we worked in your site directly, we can — but it is not the default and it is not necessary.
Why before December 3
Nothing about the change breaks your automations. The same rules keep running; they simply start carrying a line on the invoice once they pass your pooled allowance, at Atlassian's published rate of $0.50 per 1,000 steps.
The reason to look now is that every fix is the same edit whether you make it in October or in January — but in January you are making it with a number attached, under time pressure, and having already paid for the month that taught you about it.
Start with a conversation
Tell us your seat count and roughly what the Step Audit showed you. We will tell you whether an assessment is worth your money before you commit to one — sometimes the answer is that three rules explain the whole bill and you should just go and fix them.
