Budget allocation and bid management

Every campaign built across this series shares a budget, and a real account rarely stays static — this part covers how to actually decide where budget should go, and the recurring, ongoing work of managing a live account rather than a one-time setup.

Impression Share: the metric that reveals budget constraints

Campaign: Search — Coffee Subscriptions
Search Impression Share: 62%
Search Lost IS (budget): 31%
Search Lost IS (rank): 7%

Impression Share shows what percentage of all eligible auctions a campaign's ads actually appeared in. "Lost IS (budget)" specifically means the campaign would have shown more often but ran out of daily budget first — a direct, quantified signal that this specific campaign could profitably spend more, if more were available. "Lost IS (rank)" instead means ads simply aren't winning the auction often enough (part 1's Ad Rank) — a different problem, solved by improving Quality Score or bid, not by adding budget.

A real reallocation decision

Campaign A (Coffee Subscriptions): Lost IS (budget) 31%, ROAS 380%
Campaign B (One-Time Purchases): Lost IS (budget) 4%, ROAS 210%

Campaign A is both performing well and budget-constrained — the clear signal to shift budget toward it. Campaign B has budget to spare already (low Lost IS from budget) and a weaker return — increasing its budget further wouldn't unlock much additional volume, and what it would unlock returns less per dollar than Campaign A already does. Reallocating a portion of Campaign B's budget to Campaign A, informed directly by this real data, is a higher-return decision than leaving both static.

A real weekly account-management routine

1. Review Search Terms report → add new negative keywords (part 12)
2. Check Impression Share by campaign → identify budget-constrained, high-performing campaigns
3. Review Quality Score changes on top keywords → investigate any that dropped
4. Check conversion tracking is still firing correctly → a broken tag silently breaks every automated bidding strategy relying on it
5. Review ad performance within each ad group → pause any RSA a full RSA rotation confirms is a genuine underperformer

This is a real, practical cadence — not exhaustive, but covering the highest-value recurring checks in roughly fifteen to thirty minutes a week for an account this series' size. Step 4 is easy to skip and genuinely costly to miss: a tracking tag broken by an unrelated site update can silently starve every automated bidding strategy of the real data it depends on, without any obvious symptom beyond a gradual, hard-to-diagnose performance decline.

Seasonal and day-of-week bid adjustments

Bright Leaf Coffee: increase bids 20% in the two weeks before major gift-giving holidays

Ad Schedule and seasonality adjustments let bids flex for predictable patterns — a gift-focused ad group genuinely converts at a higher rate in the weeks before a holiday, and bidding more aggressively during that specific window, informed by the previous year's real data where it exists, captures demand a flat, unchanging bid would leave on the table.

When to leave well enough alone

Not every week needs an active change — a campaign performing consistently within its target range doesn't need de novo intervention purely for the sake of activity. Excessive, frequent bid or budget changes can actually disrupt automated bidding strategies (part 10), which need a stable, sustained signal to optimize effectively; constant manual tinkering works against the very automation being relied on.

Next: the metrics that actually matter — reading CTR, CPC, CPA, ROAS, and Impression Share together, rather than any single one in isolation.