How to get business metrics into OPTIMAZE
A unit cost needs two numbers, and only one of them arrives on an invoice. Attributed spend is the numerator. The denominator is business activity, customers served, transactions processed, calls made, and that number lives in systems finance already trusts. OPTIMAZE takes it in through Zapier, from any of the applications Zapier connects to.
This is the step most technology cost programmes skip, and skipping it is why so many of them stall one measurement short of being useful.
Why the denominator is the hard half
The soft side of a unit cost is not the activity number. Customers served, transactions processed and orders shipped are counted by systems that already have to be correct for billing and revenue recognition, which makes them the most reliable figures in the business. Nobody argues about them.
The difficulty is that they are somewhere else. They sit in a billing platform, a CRM, a data warehouse, a product analytics tool, and the cost platform has no view of any of it. So the ratio gets assembled by hand: export the activity figure, paste it next to the cost figure, calculate, put it in the deck.
That works once. What it does not do is produce a tracked figure, and a unit economic that has to be rebuilt every month is not a metric, it is a recurring task that competes with everything else on somebody's list.
A ratio you rebuild is a ratio you eventually stop rebuilding.
What Zapier does here
Zapier connects to the applications a business already runs, and OPTIMAZE accepts business metric data pushed in from any of them. Whatever counts the activity you want in your denominator, if Zapier can reach it, the figure can arrive in OPTIMAZE without anybody exporting anything.
Once it is in, it appears under Business Metrics and becomes available as the denominator of a unit economic, which means the ratio updates as both sides of it move rather than when somebody remembers to update it.
The practical consequence is that the choice of denominator stops being constrained by what is easy to export. If cost per active customer is the right measure and the active customer count lives in a product analytics tool nobody in finance has access to, that is now a connection rather than a negotiation.
Which metrics are worth bringing in
The test is not whether a number is interesting. It is whether a decision turns on the ratio it would produce.
Start with the measure the business is already judged on. If the board looks at customers, cost per customer is the ratio that matters and nothing else needs connecting first. If the business runs on transaction volume, cost per transaction is the one. A single well-chosen denominator against properly attributed spend produces a figure a CFO can defend, and three poorly chosen ones produce a dashboard nobody reads.
Then add the measures a specific decision needs. Cost per feature for a product owner defending a roadmap slot. Cost per tenant for a pricing conversation. Cost per inference for an AI feature whose usage is growing faster than anyone forecast.
What not to connect is anything you cannot explain the provenance of. A denominator nobody owns produces a ratio nobody trusts, and the ratio inherits every argument about the underlying number.
Where this sits in the sequence
Business metrics come third, and the order is not arbitrary.
Attribution comes first, because a cost that cannot be traced to the product or team that caused it cannot be assessed for return, and a ratio built on untraced spend has a numerator nobody can check. Business metrics come next, because that is the denominator. Targets come last, set against the ratio once it exists and tracked as it moves.
Connecting business metrics before attribution is done produces a ratio with a solid denominator and a numerator that is still an estimate, which is the more dangerous of the two failures because the figure looks finished.
What you need before you start
An account that can authorise a Zapier connection, and a decision about which activity measure you want first. That is the whole prerequisite list. No tagging standard, no data warehouse project, no agreement across engineering about a labelling convention.
The same route handles operational data more broadly, not only the headline activity count. Anything a connected application records and a ratio could use can be pushed in the same way.
Common questions
What business metrics can OPTIMAZE accept?
Any metric that can be pushed from an application Zapier connects to. The constraint is what your systems count, not what OPTIMAZE will take.
Do I need a data warehouse first?
No. The metric is pushed from wherever it already lives. A warehouse is one possible source rather than a prerequisite.
Why does the denominator matter so much?
Because a unit cost is only as defensible as both halves of it. Attributed spend divided by a number nobody owns produces a ratio that collapses the first time somebody questions the activity figure.
Can the same metric feed a target?
Yes. Once a business metric is available, a unit economic built on it can carry a target that updates as the underlying cost and usage data changes, rather than being recalculated by hand each period.
Read next
How to get unit economics for technology spend
How to attribute AI and LLM spend to features
