Budget and bidding guides

The budget is how much money is allowed to leave. It does not say what you will pay for a result and it does not enter the auction. A draft states it as a mode and an amount in minor units — daily or lifetime — and the two are different money: a daily fifty and a lifetime fifty are not comparable, which is why nothing in Tempio treats a mode change as a decrease.

The bid is what you are willing to pay, and it is the number that actually competes. Depending on the platform and the strategy it may be an amount, a target cost per acquisition, a target return, or nothing at all because the platform is choosing it for you. A bid payload carries whichever of those apply; a target return is a ratio rather than money, and Tempio treats it as one.

The target is a goal you hand the platform's own system, which it will miss in both directions while it works out how. A target is not a cap and reading it as one is the most expensive misunderstanding in this subject.

The cap is a refusal. Where a platform offers one, it holds the line at a number and takes the volume loss instead. It buys predictability with reach, and whether that is a good trade is a question about your business rather than about advertising.

Tempio puts limits around budget-bearing changes that have nothing to do with what a platform would accept, because the failure mode being guarded against is a loop rather than a typo.

A budget may not move more than one hundred per cent of what the entity carries today. Upward, that makes doubling the most a single change may do. Downward, a hundred per cent of the current figure is the whole of it, so one change can take a budget to zero — pulling spend back is not what the rule is for.

Every ad account carries its own daily ceiling, derived from what that account has actually done: three times the highest single-campaign daily spend in its last twenty-eight days, floored at a default stated in the account's own currency. The peak is taken over one campaign's day rather than the account's total, because that is the shape of the number a draft is asking for.

Applied increases on one connection are summed over a rolling twenty-four hours, and the sum plus the raise being approved has to stay under the account's cap. This is the only check that refuses a change every earlier check allowed, and it is the one that stops ten individually reasonable raises from being a doubled account by dinner — including ten driven by an agent doing exactly what it was asked.

None of that converts. Every comparison happens in the account's own currency, caps are stated in it, and no rate is applied anywhere on the writing side.

An approval rule in a Tempio workspace can clear a pause, a budget decrease or a bid decrease, under a cap it states in a currency it states. It cannot clear an increase, and that asymmetry is deliberate: the reversible, spend-reducing half of this subject is the half worth automating.

Reading a direction needs a pre-image — what the entity carried before. Where there is none, or where the budget mode changed, or where the amount moves a field the account's mirror never held, the direction cannot be read, and "cannot tell" is a refusal rather than a pass. That is worth knowing before you write a rule and wonder why it never fires.

The arrangement, stated once: your connections read your accounts, a change is written as a draft on the record with the state it would change from and the checks it passed, applying it is gated on approval — by you, or by a rule you wrote — and no platform write reaches a customer account until that platform has granted write access. Applying a cleared change to a platform is in private beta.

Every platform's delivery system needs a stretch of stable conditions to work out who to show an ad to, and a large change to a budget or a bid is a signal that the conditions changed. The cost of a change is therefore not only the money: it is the stretch of underperforming delivery that follows it.

That cost is real, and it is also the most over-claimed thing in this subject. Nobody outside a platform can tell you the exact threshold at which a change resets delivery, and these guides do not pretend to. What they do is give you the shape of the decision — fewer, larger, deliberate changes beat many small ones; changes at the start of a flight cost less than changes in the middle; and a change you can read the effect of is worth more than one you cannot.

Tempio helps with the last part in a plain way: every applied change is a row on the record with what it moved and from what, so the question "what did we change on the fourteenth" has an answer that is not somebody's memory.

They will not tell you a benchmark. There is no table here of what a good cost per acquisition looks like in your industry, because a figure stated as a fact about the world needs a source, and a median lifted from a vendor's blog post is not one. Every calculation in these guides runs on your own numbers instead — your margin, your account's own history — which is the better answer anyway.

They also will not tell you that one strategy beats another. The honest version of that answer is a decision procedure: what you know, what you are willing to lose while the system learns, and how much volatility the business can absorb. That is what each guide is.

Three of the other five subjects own a decision this one keeps handing over.

• Governance and approvals takes it the moment you want a change cleared without you: a bid ceiling is a budget decision until somebody writes a rule about it, and then it is a governance one.
• Measurement and metrics takes it whenever the argument is about whether a number moved at all, which is most arguments about bidding.
• Creative and testing takes it when the honest answer is that no bid setting will fix the asset.
• Media planning is upstream: the total these levers distribute was committed there, and a lever cannot fix a total that was wrong.
• Agent setup is how any of it gets drafted from a client rather than a screen.