PPC Campaign Checklist

Pick one primary objective per campaign — leads, signups, ecommerce purchases, or app installs — and document the target CPA or ROAS the budget is sized against. Mixing objectives in one campaign confuses Smart Bidding and makes post-mortem attribution unreadable.

Match channel to intent: Google Search for high-intent queries, Performance Max for catalog-driven ecommerce, LinkedIn for B2B job-title targeting, Meta for broad lookalike prospecting. Document the channel choice so the creative team produces the right asset specs.

Group ad groups by tight thematic intent (SKAG-lite or theme-based) so Quality Score and ad relevance stay clean. Use a consistent naming convention — {Brand}_{Geo}_{Channel}_{Intent}_{Match} — that reporting and UTM templates can parse downstream.

Pull volume, CPC, and competition from Ahrefs or Semrush; cross-reference Google Keyword Planner for forecast. Tag each keyword by intent (informational / commercial / transactional) and seed the negative keyword list from low-intent variants you do not want to pay for.

New campaigns rarely have enough conversion history for tCPA or tROAS on day one — start on Maximize Clicks or manual CPC for the first 1-2 weeks, then transition once Smart Bidding has 30+ conversions in the window. Set daily pacing so the campaign cannot blow the monthly budget in week one.

Use Google Tag Manager Preview mode plus the Tag Assistant browser extension. Confirm GA4 page_view, Google Ads conversion linker, and any Meta or LinkedIn pixels fire on the landing page and conversion page. Pixel firing pre-consent is a GDPR violation — confirm the CMP gates ad pixels behind the consent state.

Mark the GA4 event as a key event and import to Google Ads as the primary conversion. The classic mis-map: form_submit firing on email-blur instead of actual submit, inflating conversions 3-5x. Trigger a real test conversion end-to-end before launch and verify it appears in both GA4 and Google Ads conversion reports.

Pull the team's canonical UTM document — source / medium / campaign / content / term — and use the shared link builder. Drift between campaigns (utm_source=facebook vs utm_source=Facebook vs utm_source=fb) breaks attribution reports and is the single most common reason monthly performance comparisons are uncomparable.

Build the dashboard before launch, not after. Pull CPC, CTR, CVR, CPA, ROAS, and impression share by ad group. Share the live link with the requesting stakeholder so the WBR cadence is set up before week one of spend.

Google Responsive Search Ads need 8-15 headlines and 4 descriptions to feed the asset rotator. Pin headline 1 only when legal or brand demands it — pinning everything kills RSA performance. Mirror the landing page H1 in at least one headline to protect Quality Score.

Designer delivers in the platform's required ratios: 1:1, 4:5, and 9:16 for Meta and TikTok; 1.91:1 for LinkedIn single-image; landscape and square plus portrait for Google Performance Max. Confirm font licenses cover paid-media use; confirm any stock or AI-generated imagery has documented commercial license.

Sitelinks, callouts, structured snippets, lead form, location, and call extensions on Google Ads. Each extension extends the ad's real estate on SERP and lifts CTR; missing extensions is leaving free clicks on the table. Confirm sitelink destinations match advertised content (policy violation otherwise).

Headline-to-headline match drives Quality Score. Test load time on PageSpeed Insights — Core Web Vitals (LCP, INP, CLS) influence ad rank. Form has visible privacy notice, the form length matches ad's offer (a webinar registration shouldn't ask for 12 fields), and the thank-you page fires the conversion event.

Walk Google Ads and Meta policy hits: trademark in copy, superlative claims without substantiation, restricted categories (financial, health, alcohol), personalization-restricted categories. Disapprovals after launch waste budget and delay learning-phase exit.

Pharma and medical-device need FDA ISI; financial advisors need FINRA Rule 2210 review; substantiated comparative claims need a Lanham Act-defensible substantiation file. Skip this step when the campaign is genuinely outside a regulated category, but document that decision.

Capture timestamped approval from legal reviewer and brand owner. Slack DMs are not an audit trail — record the approver name, date, and any conditions. Required when the regulated-review flag is Yes; recommended for all paid sends regardless.

Launch at 25-50% of intended daily budget for the first 3-5 days while you confirm conversions track end-to-end and impression share looks healthy. Scale up only after the learning phase shows stable CPA — not on day one.

Confirm impressions, clicks, and conversions are all flowing. Common 48-hour gotchas: zero impressions (low bid or disapproved ads), clicks but zero conversions (broken pixel or thank-you page), runaway spend (bid strategy mis-set or budget cap missing).

Re-test the GTM trigger, the GA4 event, and the Google Ads conversion import. Check the consent-mode state — pixels gated behind a CMP fail to fire when consent is denied, which is correct behavior but reads as broken to a non-privacy-aware reviewer. Pause spend until the fix lands and a fresh test conversion appears in both platforms.

Pull the search terms report; add irrelevant and low-intent queries to the negatives list at campaign or shared-list level. This is the single highest-ROI optimization in the first 30 days — broad-match terms drift quickly and burn budget on queries the campaign was never meant to compete on.

Use the platform's experiment feature (Google Ads Experiments, Meta A/B Test) — not two parallel ad sets, which contaminate audiences. Test one variable at a time: headline, creative, or landing page. Plan for 1-2 weeks at sufficient volume to reach 95% confidence on the primary metric.

Apply the team's standing thresholds — for example, pause keywords with 100+ clicks and zero conversions, or ads with CTR below 0.5x ad-group average after 2,000 impressions. Document the decision in the optimization log so the next reviewer doesn't re-enable them by accident.

Report against the target CPA or ROAS captured at kickoff — not against a moving goalpost. Include impression share lost to budget vs. rank, learning-phase status, and a recommendation: scale, hold, or sunset.

Capture what worked, what didn't, and which hypotheses were tested. Future campaigns in this account inherit the lessons; without the log they re-run the same failed tests.