One of the most common mistakes on new Google Ads campaigns is applying aggressive targets and constraints before the algorithm has any data to work with. It feels like responsible management: setting a maximum CPC, capping the Target CPA at what you need to be profitable, restricting the daily budget tightly. But applied too early, these constraints don't protect the campaign. They prevent it from getting off the ground.
I took on a new client in the HR consultancy space last year who had launched their own campaign, set a Target CPA of £30 on day one, and concluded after three weeks that Google Ads didn't work for them. When I looked at the account, the campaign had spent £47 in three weeks. It had generated four clicks. The Target CPA was so far below what the market could deliver that the algorithm had essentially stopped bidding. I relaunched on Maximise Conversions, and within six weeks they had 22 leads at an average CPA of £58. Not perfect, but real data to work with.
The counterintuitive truth here is that setting tighter constraints on a new campaign usually makes results worse, not better. The algorithm needs room to explore. Restricting it from the start is the equivalent of hiring someone and then refusing to let them do the job.
How Google's Smart Bidding actually works
Smart Bidding strategies such as Maximise Conversions, Target CPA, and Target ROAS are machine learning systems. They work by observing patterns: which searches, times of day, devices, audience segments, and bidding environments produce conversions, and which don't. Over time, they use those patterns to bid more aggressively where conversions are likely and less aggressively where they're not.
The critical point: they need conversion data to do this. A campaign with 5 conversions in its history cannot identify patterns. A campaign with 50 to 100 conversions has enough signal to make meaningful predictions. The threshold Google recommends for Target CPA is approximately 30 to 50 conversions per month per campaign before the strategy can function reliably.
Before that threshold is met, Smart Bidding is essentially guessing. And if you've set an aggressive Target CPA target alongside it, you've created a situation where the algorithm is trying to hit a target it doesn't have enough data to pursue, which typically results in either severe underspending (it won't bid where it can't achieve the target) or volatile spending while it searches for a pattern that isn't yet visible in the data.
What happens when you set a Target CPA too low too early
Here's a concrete example. A new campaign targeting solicitor services launches with a Target CPA of £20, the number that would make the economics work perfectly. But in that sector, average CPCs are £8 to £15, and a typical conversion rate on a new campaign without optimised landing pages is 2 to 4%. That implies a real CPA of £200 to £750 per lead, not £20.
The algorithm knows its conversion history. It sees that it has never achieved anything close to a £20 CPA. Its response: it stops bidding on most impressions because almost no auction looks like it can deliver the target. The campaign either goes dormant, showing ads rarely, spending almost nothing, or it runs erratically, chasing unlikely conversions.
The advertiser sees poor performance, concludes Google Ads doesn't work for their business, and pauses the campaign. In reality, the campaign was never allowed to function.
The right approach for new campaigns
Start on Maximise Conversions with a realistic daily budget cap. This tells Google to spend the budget generating as many conversions as it can, without a specific cost target. It's the most permissive Smart Bidding strategy, which is exactly what a new campaign needs: permission to explore and accumulate data.
During this phase, the job of the account manager is active monitoring rather than constraint: checking the search terms report weekly, adding negative keywords, reviewing geographic performance, and ensuring conversion tracking is accurate. The goal is to build a clean, well-structured dataset that gives the algorithm something to work with.
Once the campaign has reached 30 to 50 conversions, look at the average CPA you've achieved. That real number is the starting point for setting a Target CPA. Set it slightly above the current average initially, perhaps 10 to 20% higher, to give the algorithm room to find volume. Then tighten it gradually, in 10 to 15% increments, over subsequent months as efficiency improves.
When budget constraints are legitimate
Constraining budget with a daily cap is always appropriate. It's how you control total spend. The daily cap doesn't break the algorithm; it just sets the ceiling on how much it can spend. The problem is when the daily cap is set so low that the campaign can't accumulate data fast enough (for example, a £10/day cap in a sector with £8 CPCs means fewer than two clicks per day and near-zero conversion data).
The other legitimate constraint is pausing or reducing bids on specific keywords or ad groups that are clearly performing poorly after sufficient data has accumulated. Pausing a keyword that has 100 clicks and zero conversions over 90 days is a reasonable, data-backed decision. Pausing it after 10 clicks because it hasn't converted yet is not.
The pattern I see
Most under-performing accounts I audit either launched with constraints they imposed before having data, or had someone apply an aggressive Target CPA target at the campaign level after a week of poor performance in a panic response. Both errors prevent the algorithm from doing what it's designed to do.
The discipline required in the early stages of a campaign isn't to constrain. It's to monitor carefully without overreacting, build good negative keyword lists, and give the system the space and data it needs to start functioning properly. I see this play out regularly with new businesses in North Wales. The instinct to protect budget by setting tight targets from day one ends up preventing the campaign from ever getting off the ground.
Google Ads management in North Wales. If you'd like help structuring a new campaign launch or auditing an existing campaign that's struggling to gain traction, get in touch.
