Sometimes the most useful thing a paid search review can say is that the next increase will not pay. Here is how our PPC team decides that: the structure that shows it, the checks we run, and when cutting is wrong. It draws on the Google Ads account we run for Bruc Bond in Singapore and the United Kingdom.
Why an agency would ever recommend less
Much of the industry charges management as a percentage of ad spend. Under that model the short-term incentive runs the wrong way: a bigger budget means a bigger fee, whatever the budget returns. An agency paid that way has a standing reason to keep recommending increases past the point where they help.
We price management differently, and our pricing page says so: the monthly fee tracks the platforms and hours a clean account needs, not a percentage of your budget. Media spend is billed to you by each platform.
That removes the incentive to inflate spend. It does not remove the work of knowing where more spend stops paying, which is what the rest of this post is about.
Where the ceiling hides: one campaign averaging two markets
Bruc Bond offers business banking and international payments, with a SaaS platform built on the same capabilities. Its two priority markets, Singapore and the United Kingdom, each have their own search behaviour. Its buyers must be businesses, not consumers. And two products, banking and the platform, compete for the same budget.
We ran its Google Ads with separate campaigns for each market, and the banking services and the SaaS platform as distinct campaigns with their own messages. Keywords and negatives were built around what businesses search when they need corporate banking or cross-border payments, keeping consumer queries out. The full account map is in the Bruc Bond case study.
That structure is the precondition for any honest budget call. When one campaign averages two markets, a market that has run out of good auctions hides behind one that has not. The blended cost looks fine, so budget keeps rising into the weaker market’s worst queries. Split them and each shows its own ceiling.
The case record carries no performance figures, so we use its structure here, not its results. What it shows is the order of work: separate what needs judging separately, then judge each part on its own numbers.
The checks we run before recommending a budget change
This is the sequence our team works through, per campaign rather than per account. You can run it on your own account with read access to Google Ads.
- List each campaign by market and product. If one campaign serves two markets or two products, note it; its numbers are an average and cannot support a budget decision on their own.
- For each campaign, find the last three or four budget changes. Record spend, conversions and cost per conversion for a settled period before and after each one.
- Read the two impression share columns: search impression share lost to budget, and lost to rank. Budget loss says more money could buy more of the same auctions. Rank loss says it could not.
- Open the search terms report for the weeks after the last increase. Count the new queries that a business buyer would never type. That is where extra budget usually goes.
- Check the landing page conversion rate over the same period. If it fell while clicks rose, the ceiling is the page, not the auction.
The table below is how we read the result and what we recommend in place of a straight increase.
| What the checks show | What it means | What we recommend instead |
|---|---|---|
| Cost per conversion rises in step with each budget increase, across several changes | The campaign has bought most of the good auctions; extra spend buys worse ones | Hold spend and move the difference to a market, product or channel that has not hit its ceiling |
| Impression share lost to rank is high, lost to budget is low | More budget cannot win these auctions | Fix relevance, ads and landing pages before any increase |
| New irrelevant queries appear after each increase | Matching has widened to fill the budget | Add negatives and tighten match types, then re-test the increase |
| Clicks rise, landing page conversion falls | The page is the limit | Spend on conversion work rather than more clicks to the same page |
| Impression share lost to budget is high and cost per conversion holds | The campaign is not at its ceiling | Increase, in steps, and repeat the checks |
Telling a bad week from a real ceiling
Auctions are noisy for reasons outside your account: a competitor’s short promotion, a seasonal event, a platform change. Cutting budget after one dramatic week produces a budget that swings and never settles long enough to prove anything.
A real ceiling repeats. It shows as the same relationship between spend and cost per conversion across several independent changes, upwards and downwards. A temporary spike breaks that relationship for a window you can explain, then recovers. Before we recommend a lasting cut, we want to see the pattern more than once.
- What it is
- A new campaign, or one that just changed bidding
- What to do
- Wait: early numbers are not a ceiling
- What it is
- A competitor promotion, seasonal event or platform change
- What to do
- Do not cut: the budget would swing, never settle
- What it is
- One spend to cost pattern, across several independent changes
- What to do
- Recommend a lasting cut once the pattern repeats
The same thinking applies on the organic side, where a page can win traffic that never turns into revenue; we cover that in the keyword that costs you money.
When spending less is the wrong answer
We do not recommend cutting by default, and some accounts should spend more. Holding or cutting is the wrong call when:
- A converting campaign is losing a large share of impressions to budget while its cost per conversion holds. The budget cap is turning away buyers who would convert.
- A campaign is new or has just changed bidding strategy. Its early numbers are a learning period, not a ceiling.
- A known seasonal peak is coming, and the account has history showing that the peak converts.
- Competitors bid on your brand terms. Brand defence can be worth paying for even when the organic listing sits beneath it.
The recommendation is only as good as the account structure under it. If your markets and products share campaigns, the first job is to separate them; the budget call comes after.
Written by
Shweta Agarwal
Chief Operating Officer, Vinove
Shweta runs the operating side of the group, holding delivery, process and accountability consistent across the group. Her focus is on the systems that let teams grow without losing the standard they started with.
LinkedIn →
