Campaign Structure & Budgeting

The agency that restructured one account and doubled revenue

In 2023, a digital marketing agency took over a Google Ads account that had been running for two years. The previous agency had delivered mediocre results: £18,000/month spend, £31,000/month revenue (1.7× ROAS).

The account had 340 active keywords across 22 campaigns, all sharing budget from the same campaign types with no clear structure. Branded keywords (high-converting, cheap) competed with generic keywords (expensive, low-converting) for the same daily budget. Remarketing campaigns ran alongside cold prospecting in the same campaign structures.

The new agency rebuilt from scratch in 60 days:

• 3 campaigns: Branded (keywords with the company name), Category (non-branded but high-intent), Competitor (conquesting competitor brand searches)
• Remarketing in a separate campaign with separate budget
• Branded campaign separated so its high-Quality-Score clicks didn't subsidise expensive generic keywords

Result after 90 days: same £18,000 spend. Revenue: £68,000 (3.8× ROAS — more than double).

The ads didn't change significantly. The structure did.

(Illustrative scenario based on patterns common in paid search account management. Specific figures are representative of real-world outcomes — not a verified account of a specific named company.)

The account structure hierarchy

Google Ads account structure:

Why branded campaigns must be separated: Your brand name searches have extremely high Quality Scores (people searching your exact name), low CPCs (little competition), and the highest conversion rates (they already know you). If these compete for budget against expensive generic keywords, one of two things happens: either brand keywords use budget that should go to generic (wasted — brand searches would convert anyway), or generic keywords steal from brand (losing cheap, high-converting clicks to fund expensive, lower-converting ones).

Separate branded campaigns get their own budget and can be bid down — protecting the budget for harder-to-win generic clicks.

Meta Ads account structure

Meta's best-practice structure has evolved with AI-driven optimisation:

Campaign Budget Optimisation (CBO) — the modern approach: Set budget at the campaign level. Meta's algorithm automatically distributes budget across ad sets within the campaign, concentrating spend on the ad sets that are converting. This is generally better than setting budgets per ad set (Ad Set Budget Optimisation) for established campaigns with conversion data.

Campaign structure:

Why prospecting and retargeting must be separate campaigns: Audience economics differ dramatically. Cold audience CPMs are higher; conversion rates are lower; creative needs are different. Mixing them in the same campaign causes Meta's algorithm to disproportionately fund retargeting (which has better short-term signals) at the expense of cold audience growth — starving the top of the funnel.

Budget allocation: the art and science

Starting budget considerations:

A useful framework: before you can profitably scale, you need data. Data costs money. Treat your initial ad budget as a learning investment, not an expected profit centre.

Minimum meaningful budget per platform (to gather data that's worth optimising):

• Google Search: £15–30/day per campaign (£500–1,000/month per campaign)
• Meta Ads: £20–40/day per campaign (£600–1,200/month)
• LinkedIn Ads: £50+/day recommended (platform minimum is lower, but CPCs are significantly higher; £1,500+/month needed to gather useful data)

Budget allocation by funnel stage:

The funnel dependency: Hot retargeting audiences are created by the activity above them. If you cut cold prospecting budget, within 30–60 days your hot retargeting audience shrinks (fewer website visitors → fewer cart abandoners). Optimising only for short-term ROAS by cutting cold prospecting is a common mistake that destroys long-term performance.

Budget pacing and spend monitoring

Daily budget vs. monthly budget: Most platforms set daily budgets. The actual daily spend can vary — Google Ads may spend up to 2× the daily budget on high-opportunity days, balanced by underspending on low-opportunity days, staying within the monthly equivalent.

Pacing issues to watch:

• Overspending: Spending significantly over daily budget early in the month leaves insufficient budget for later (when learnings are better and performance is typically stronger)
• Underspending: Budget isn't being spent = targeting is too narrow, bids are too low, or the auction is too competitive. Not a sign of health — unused budget generates zero data and zero return.

There Are No Dumb Questions

"Should I spread budget equally across campaigns or concentrate it?"

Concentrate it on what's converting. A common mistake is spreading budget evenly across 10 campaigns, giving each £50/day — insufficient data to optimise any of them. Better: 3 campaigns with £150/day each, built around the highest-conviction opportunities. Once those are proven and profitable, add campaigns with the incremental budget that expansion generates.

"How often should I adjust campaign budgets?"

Weekly for significant adjustments (scaling successful campaigns); monthly for strategic budget reallocation across channels. Changing budgets daily in response to day-to-day fluctuations is a mistake — daily performance varies naturally, and constant adjustments destabilise algorithm learning. Make budget changes with purpose, not in reaction to noise.

Back to the agency

Same spend. Same keywords. Same market. The only thing that changed was the structure — and revenue went from £31,000 to £68,000 per month. When branded and generic keywords competed for the same budget, the algorithm couldn't allocate spend intelligently. Separating them gave each campaign clean signals and its own budget logic. Structure determines which data the algorithm learns from, and bad structure produces bad learning regardless of how good the ads are. The previous agency wasn't failing because of bad creative — it was failing because the account's architecture made good performance structurally impossible.

Key takeaways

• Branded campaigns must be separated. Brand searches are cheap, high-converting, and should be protected from competing with expensive generic keywords for the same budget.
• Cold prospecting and retargeting must be in separate campaigns. Mixed objectives cause algorithm confusion and budget misallocation; retargeting consistently out-signals cold audiences, starving the top of the funnel.
• Treat initial budget as a learning investment. The first £1,000–2,000 buys data that makes subsequent campaigns profitable. Expect marginal or negative ROAS in month 1.
• Budget allocation follows the funnel: 50–60% cold, 25–30% warm retargeting, 10–20% hot retargeting. Cutting cold prospecting destroys future retargeting pool.
• Concentrate before diversifying. 3 campaigns with £150/day generates more optimisable data than 10 campaigns at £45/day.