Verify what bidding counts as success
Start with the conversion actions used by the campaign. Check account-default goals, campaign-specific goals, and any custom goals. A page view or form start can be useful for diagnosis while remaining unsuitable as the outcome that drives acquisition bidding. Confirm the intended goal set instead of assuming every action marked primary belongs in every campaign.
Complete a test submission, purchase, or call through the real customer path. Check that it records once with the correct value and currency. Verify that the same purchase is not counted twice through separate tracking methods. For lead generation, distinguish unique leads from repeated form submissions and inspect the counting setting.
- Confirm the primary goal represents an action the business accepts.
- Keep diagnostic actions separate from the main outcome report.
- Review attribution and conversion windows before judging recent traffic.
- Check imported offline outcomes for delays or missing records.
- Record any measurement change in the change log.
An obvious configuration error can justify immediate correction, but changing the goal set can affect Smart Bidding learning. Document the decision and avoid changing budgets, targets, and attribution at the same time without a clear reason. If tracking fails, repair it before using a low reported conversion rate to exclude otherwise relevant traffic. The weekly audit workflow supports recurring measurement checks after the initial review.
Remove verified irrelevant intent before grading performance
Export search terms with campaign, ad group, keyword, match type, clicks, cost, and conversions. Read the actual query rather than judging the triggering keyword alone. Sort first for obvious business mismatch, then review expensive terms whose outcomes have had time to mature.
A service provider can exclude a query for equipment rental when it offers only completed installations. That decision comes from the offer, not a minimum click count. A relevant installation query with no sale is different. Check conversion delay, destination, pricing, and lead quality before excluding it.
Use cost above twice target CPA with zero mature conversions as an investigation trigger. With a $75 target CPA, the review threshold is $150. This does not mean every term crossing $150 becomes a negative. Treat the trigger as a way to prioritize scarce review time.
| Finding | Initial response | Check before applying |
|---|---|---|
| Specific irrelevant query | Consider a negative exact | Does the exclusion preserve valuable related intent? |
| Repeated irrelevant phrase | Consider a negative phrase | Can that phrase appear in useful searches? |
| Relevant term with high mature CPA | Inspect bid, offer, and destination | Is tracking reliable and evidence sufficient? |
| Relevant term with recent clicks only | Keep within a spending boundary | Has the conversion window matured? |
Review n-grams to find recurring patterns, then inspect the complete queries behind them. Do not add overlapping pattern costs together when reporting waste. Google does not expose every query, so keep total campaign spend separate from the reviewed search term subset. The negative keyword workflow supports exact and phrase exclusions and shared lists. Use the negative keyword examples to check scope before sharing a list across campaigns.
Match location eligibility to where you can serve
Review both the selected locations and the advanced location option. A setting based on presence or interest can reach people who show interest in a location without being physically there. Presence focuses on people in or regularly in the selected area. Neither setting provides perfect real-time physical location certainty.
For a local business that serves only nearby customers, presence is often the better starting point. Check the location report and actual lead serviceability before changing it. A hotel, destination service, or relocation business may intentionally value people researching from elsewhere. In that case, interest can match the buying journey.
- Confirm the service area with the business owner.
- Review matched location and user location information where available.
- Inspect exclusions and overlapping geographic targets.
- Compare qualified inquiries, not just raw conversion counts.
- Document why presence or interest fits the offer.
Do not exclude a state solely because a short reporting window shows expensive clicks. Inspect mature outcomes and whether the campaign intended to reach that area. Correcting a clearly wrong location setting is a useful first-week action. Geographic performance pruning needs a stronger evidence base, especially when individual regions have little traffic.
Separate eligibility decisions from bid adjustments
Check the ad schedule against account time zone, business hours, call handling, and the landing page experience. A business that cannot answer calls overnight may need a narrower call-focused schedule. A website that accepts qualified inquiries at any hour may still benefit from evening traffic. Conversion time is not necessarily the time of the ad click.
Segment performance by hour and day using a mature observation window. Compare lead quality and response coverage before removing hours. Sparse overnight data does not justify a permanent shutdown, and closing all evenings can remove research visits that convert later.
Review device results alongside the customer path. Test mobile forms, payment steps, click-to-call behavior, and page usability. An apparently weak device can reveal a broken interaction rather than bad demand. Fixing the destination can be more useful than removing the traffic.
Manual bid adjustments and Smart Bidding behave differently. Many traditional device and schedule adjustments do not apply to Smart Bidding strategies. Check the current strategy's supported controls before making an edit. Ad scheduling can still control eligibility where supported, which is a different decision from changing bids inside the eligible hours.
- Check account time zone before editing hours.
- Use click-time cohorts when comparing acquisition efficiency.
- Verify lead handling during the hours you retain.
- Repair device-specific failures before excluding devices.
- Record any eligibility restriction and its business reason.
Avoid stacking several bid adjustments without calculating the combined exposure when manual bidding applies. Keep the first-week change focused on a verified problem and give the remaining traffic time to generate mature outcomes.
Inspect Search Partners and Display expansion separately
Open the campaign's network settings and confirm where the Search budget is eligible to run. Search Partners and Display Network expansion are different sources of delivery. Evaluate each separately where the campaign type offers those options. These Search controls do not describe the full inventory or settings of Performance Max.
Segment Search performance by network. Compare spend, mature conversions, qualified leads, CPA, and conversion value. A lower CPC on partner inventory is not a business result if lead quality is weak. Equally, a different network should not be disabled simply because its click behavior differs from Google Search.
For a tightly scoped Search test, turning off optional Display expansion can make the initial acquisition review easier to interpret. If the expansion is intentional, give it a defined objective and inspect outcomes rather than treating its traffic as interchangeable with Search queries. Use separate campaigns when a distinct campaign type and funding boundary better fit the strategy.
- Confirm whether Search Partners is enabled.
- Check the Display option available in the campaign settings.
- Compare mature business outcomes by network.
- Investigate tracking and lead quality before blaming the source.
- Record the budget and outcome impact after a network change.
If partner traffic has no qualified outcomes after meaningful mature spend, a bounded exclusion test can be reasonable. Set the review window in advance. Do not claim the change saved its entire historical cost because the remaining inventory can absorb the released budget.
Expand reach only with reliable goals and boundaries
Review exact, phrase, and broad match by search term relevance and campaign role. Exact is tighter than phrase or broad, but it can still match close variants and the same intent. Phrase is not a literal substring restriction on every positive match. Inspect actual queries instead of assuming the label guarantees relevance.
Broad match with Smart Bidding can be appropriate when conversion measurement is reliable and the campaign has a meaningful business goal. It can also expand into unsuitable intent when the goal set rewards weak actions. Treat measurement, targets, exclusions, and spending limits as prerequisites for a controlled test.
Create a test whose spend is affordable if results disappoint. For an initial operating boundary, reserve no more than 10% of the approved acquisition allocation for an uncertain expansion. This is a budget choice, not a platform requirement. Use an eligible campaign experiment where suitable, or document the comparison limits of a separate campaign.
- Verify primary goals and accurate conversion values.
- Use the CPA or ROAS strategy that fits the business outcome.
- Keep proven irrelevant intent excluded.
- Define a total test allowance and review dates.
- Inspect search terms and qualified outcomes as evidence matures.
- Stop expansion when the authorized spending boundary is reached.
Do not combine a broad match test with a large budget increase, new landing page, and target change unless the situation requires it. Those changes make interpretation harder. Review recommendations as proposals rather than accepting them without checking commercial fit. The Google Ads optimization workflow connects routine reviews with ongoing execution.
Schedule the work in an order you can evaluate
Begin with measurement and business eligibility. Then remove clear irrelevance and inspect allocation. Save uncertain performance judgments for a window that includes enough mature evidence. The goal of week one is a defensible set of fixes and a monitoring plan, not a promised percentage reduction in spend.
| Timing | Priority | Completion evidence |
|---|---|---|
| Start of week | Validate conversion goals and tracking | A verified action records correctly |
| Early review | Check queries and location eligibility | Each exclusion has a business reason |
| Midweek review | Inspect networks, schedule, and devices | Settings match the intended acquisition path |
| End of week | Review pacing and expansion tests | Spend stays inside authorized boundaries |
For each edit, record the starting setting, reason, owner, and earliest useful evaluation date. Keep an immediate check for delivery errors and a later check for mature efficiency. Roll back when the setting is wrong, rather than waiting for a formal reporting date. Undoing an edit changes future eligibility or settings; it does not recover past spend.
Use the 30-day optimization checklist to extend the review beyond immediate fixes. Keep bid and budget rules aligned with the bidding strategy so automation does not issue conflicting instructions while you assess the changes.
Questions about first-week optimization
You can verify many settings and technical fixes within seven days, but performance conclusions depend on conversion delay and available evidence. Check eligibility immediately, then assess mature outcomes later. A week of data can be insufficient for a low-volume campaign.
No. Review relevance, mature spend, conversion delay, and measurement first. A clearly irrelevant query can justify an exclusion immediately. A relevant keyword with a small number of recent clicks usually needs a spending boundary and more observation rather than an automatic pause.
No. Evaluate Search Partners using mature qualified outcomes and the campaign's business goal. Disable or test removal when the evidence and spending boundary justify it. Lower click cost alone does not prove value, and a network label alone does not prove waste.
Keep verified rules working after the first week
PpcAutomate runs Google Ads and Microsoft Ads accounts on autopilot with n-gram query mining, exact or phrase negatives, shared lists, and bid and budget rules toward target CPA or ROAS. Daily change caps, anomaly alerts, monthly reporting, and a change log with one-click undo keep actions visible. Growth includes Approve or Autopilot mode per rule, weekly audits, Slack alerts, and an ad copy test monitor. Define the business boundaries and use recurring rules for the decisions you can evaluate consistently.
Start with the one-time free homepage demo. Upload a Google Ads or Microsoft Ads search terms CSV to inspect query spending against your target. The file is processed in your browser. Use the findings to choose the first exclusions to review, then apply the broader checklist to settings and measurement that the CSV cannot verify.