Optimize Google Ads settings for freelancers: make a defined change for a clear reason

Define what optimisation should accomplish

A setting is worth reviewing when it affects the client's intended advertising behaviour. A freelance consultant should begin with that purpose rather than changing options simply because an audit or interface makes them easy to find.

The objective may concern geographic relevance, spending control, measurement meaning or the structure of an offer. Each one suggests a different observation and supported action. “Optimise everything” does not define a reviewable request.

ppctor provides settings discovery, current account reads and approval-based edit preparation. These capabilities help the freelancer move from a broad concern to a defined target and operation without treating every finding as an automatic instruction.

Keep the client's business context visible. An option that suits a local appointment service can serve a different purpose for a travel business. A technically available edit is not evidence that the edit is appropriate.

State the expected account behaviour in ordinary language. The client should understand what the proposed change controls and what it cannot guarantee. A location option can alter targeting logic without proving every future user's exact physical location.

The useful outcome is a clear decision supported by evidence. It can be to retain the current setting, investigate another fact or prepare a specific change. Responsible optimisation does not require a mutation at every review.

Discover the setting before reading its value

The assistant's settings catalogue describes available fields and entity levels. Discovering that catalogue is a local capability check, not a provider read. It tells the assistant how a setting can be requested.

A current-value question then needs the relevant account evidence. The fact that a field exists in the catalogue does not establish what the client's campaign currently uses or whether a particular write is supported.

Keep these stages distinct in the conversation. “This campaign setting can be read” and “this campaign currently uses this value” are different claims. The second requires returned account data.

Use the relevant field and compatible resource instead of guessing from a performance report. Observed traffic does not by itself reveal a campaign's configured network option or its positive location-targeting setting.

For a follow-up about another setting on the same target, retain the target context deliberately. Do not broaden the read to every campaign merely because several settings can be discovered at once.

This approach avoids unnecessary provider calls while preserving evidence for factual claims. Local discovery answers a capability question, and a scoped account read answers the user's current-state question.

Use audit findings as review evidence

An audit finding identifies a condition the app considers worth reviewing. It can provide a useful starting point for the freelancer, but it does not replace the client's objective or prove the commercial effect of the condition.

Read the named target and the check's scope. Some checks concern account settings, others campaign options or deeper assets. A finding on one entity should not become an account-wide conclusion without supporting evidence.

Keep unavailable checks visible. The audit retains could-not-check information rather than counting an unread area as a pass or issue. A clean summary cannot certify settings that the completed run did not inspect.

A diagnostic condition is not a quantified waste estimate. A low Quality Score or an empty-looking structure can deserve investigation, but the finding alone does not establish lost revenue or future savings from a change.

Confirm current state before a proposal. An audit result can precede another person's manual edit. The execution flow will recheck relevant settings, but the freelancer should also make the recommendation fit the latest available context.

Use the audit to organise work into clear decisions. Retain what was observed, why it matters to the client's goal and which supported operation is being considered. Avoid presenting a list of findings as a universal prescription.

Choose the correct entity level

Settings belong to specific account entities. An account setting, campaign option and ad-group control require different targets. The action should request only the context needed for its valid scope.

For example, an account-wide app-category exclusion does not need the assistant to discover campaigns. The scope is the connected account. Inventing a campaign target would confuse the user's intention and add unnecessary work.

A campaign option needs the named campaign. An ad-group setting needs that group and its parent campaign. A keyword or ad operation needs its deeper context so similarly named entities can be distinguished.

Performance Max uses asset groups rather than Search ad groups. Standard Shopping uses ad groups with listing structures. Keep the hierarchy accurate when describing settings and proposed product-targeting work.

Shared resources require relationship context. A shared negative list can affect several attached campaigns. A content edit and a change to one attachment are different operations even when they concern the same named list.

The client should see readable names and meaningful parents rather than internal identifiers. The application resolves resources behind the scenes, while the proposal gives the owner enough context to recognise the exact live target.

Separate location targets from presence options

Selected locations describe the places configured for targeting. The positive location option describes how presence and interest are treated. A request to target a country does not necessarily specify both choices.

Google's advanced-location guidance distinguishes presence from presence-or-interest targeting and notes that location signals do not guarantee perfect accuracy. Review the option in the client's business context.

A local business that serves people physically near its premises may have a different requirement from a business serving people planning travel. Do not change the option merely because one audit interpretation sounds universally safer.

The focused presence action changes the campaign's positive targeting option. It is separate from adding or removing the campaign's actual geographic criteria. The proposal should make that distinction clear.

When the user requests both a country and Presence Only, account for both logical settings. A successful country addition should not be presented as fulfilment of the presence instruction unless that operation was also prepared and executed.

After the change, read the relevant configured result. A later geographic performance report can help investigate observed activity, but it does not independently prove every targeting signal or explain all out-of-area traffic.

Review location and radius edits

The focused location workflow can add supported geographic targets, remove explicit positive location or address-radius criteria and add specified radii. Its proposal should identify the places and radius values in readable language.

Removing a criterion requires that it belong to the selected campaign and the supported positive-target population. The workflow should not treat any resource name mentioned in a conversation as an authorised removal target.

Review radius units carefully. Miles and kilometres produce different areas, and a bare number does not establish the intended unit. Confirm the address and the requested distance together.

The current location edit preserves location exclusions and the Presence setting. If the user wants to change that option too, prepare it separately. Adding a place does not implicitly reset every geographic setting.

A radius around a linked store has additional prerequisites. The relevant populated location group must already exist. Creating such a group is a separate account prerequisite, and provider population can be asynchronous.

Keep these requirements visible before approval. A dependency description helps explain why a requested radius cannot simply be attached immediately. It should not cause the assistant to create unrequested infrastructure during an ordinary targeting change.

Understand a complete language selection

The focused campaign-language action supplies a desired complete set. It is not merely an append operation for one language. Review the full intended selection before confirming it.

An empty set represents all languages in this action. That has a different meaning from an unavailable value or an accidental omission. The proposal should make the requested account behaviour understandable.

Keep campaign targeting separate from keyword-research language preferences. A planner request does not itself alter the live campaign's language criteria. Likewise, editing campaign languages is not a request for research output.

Use the client's actual offer, creative and audience context to assess the selection. The language of one keyword does not establish the correct campaign configuration across every group or ad.

When the client says “add English”, clarify whether they want to retain existing languages if the action requires the complete desired set. A correct replacement should reflect the final selection rather than an incomplete fragment.

Record the verified selection after execution. A later performance comparison should retain the timing of the change and avoid attributing every outcome movement to language targeting without further evidence.

Review network and automation choices

Network and automation options can change where or how advertising operates. Read the current campaign setting before recommending a supported disable action. Do not infer it solely from one example of traffic.

The focused catalogue includes actions such as disabling Search Partners, disabling Display Network in the relevant campaign context and disabling AI Max. Each has its own field and target requirements.

Availability does not prove that disabling an option improves every account. The freelancer should connect the recommendation to the client's goal and relevant evidence rather than describe every expansion mechanism as waste.

An ad-group optimized-targeting disable action is a different scope from a campaign-level network change. The proposal should retain the group and parent campaign so the owner recognises what will change.

Read the operation's actual direction. A focused disable tool is not necessarily a general toggle that supports every possible new value. Do not promise the inverse operation without checking the available action.

After an approved change, review the actual configured result and then observe a suitable performance period. Verification of the setting confirms the mutation; it does not establish the commercial effect of the option.

Check a daily-budget adjustment

A campaign budget adjustment needs the supported budget configuration, current amount and intended operation. The focused workflow supports non-shared daily budgets rather than every budget resource the provider exposes.

Setting a monetary amount differs from raising or lowering by a percentage. The first specifies the new value in the account currency. The percentage operations derive a result from the applicable current value.

Review any floor or ceiling included in the request. A bounded adjustment should preserve the client's constraint. Do not silently ignore it because an unbounded percentage operation would be easier to prepare.

The campaign budget is separate from the pacer's cycle reference. Updating a live setting does not automatically revise the client's planning agreement, and changing that reference does not enforce a new platform cap.

Use performance and business context before recommending an increase. A low recent cost can indicate several conditions, and a larger budget does not guarantee additional useful outcomes.

For a shared or unsupported budget configuration, explain the actual boundary. The freelancer can still assess the business question, but the confirmation should not submit an improvised operation that bypasses the action's target rules.

Review an existing CPA or ROAS target

The focused target actions adjust existing CPA or ROAS targets. They do not create every possible bidding strategy merely because the client asks for a different acquisition goal.

Read the campaign's current strategy context and applicable target. A visible desired business CPA does not prove that the campaign has a supported editable target in its current setup.

CPA uses the account currency for an absolute set operation. ROAS uses display percentage, so 400 means 400% in the focused action. Make the units explicit when the client discusses return as a multiple.

A proportional raise or lower also needs careful interpretation. Increasing a target and increasing a budget are different controls. The proposal should state which one changes and preserve any other settings not requested.

Compare the target with a consistently defined outcome. Purchase-only return and all-primary value return can differ. The client should understand which observation supports the recommendation.

Do not promise that the new target will immediately produce the stated cost or return. It is a live control, while recorded outcomes require later evidence. Review both the verified setting and subsequent activity separately.

Treat conversion configuration carefully

Conversion roles and counting settings affect how account outcomes are interpreted. The focused catalogue includes supported role and counting operations, but the applicable action type, origin and provider restrictions can constrain them.

Do not assume that every visible conversion action is freely editable. A locked or unsupported action needs an accurate explanation. The tool should respect the target rules instead of forcing a write through a familiar field.

A bulk role operation concerns explicit targets. The proposed primary selections and remaining secondary selections should be clear. It is not permission to reclassify every unrelated action in the account.

Keep measurement changes separate from claims that tracking is broken. A role recommendation can depend on the client's objective without proving duplicate events, missing transactions or implementation errors.

The app's business reports use primary actions by default and add the selected actions without a deduplication repair. Changing a role can therefore change what the reported population includes even if the underlying event collection is unchanged.

Document the timing and intended definition. A later conversion comparison should explain any population change rather than present the resulting count movement as a pure performance improvement.

Keep tracking flags separate from measurement health

Auto-tagging and tracking templates are account configuration questions. A current flag can be read and a supported change prepared, but the flag alone does not establish that measurement is working throughout the client's site.

Enabling auto-tagging changes the relevant account setting. It does not verify redirects, consent behaviour, destination processing or every analytics link. Those require their own observations and responsible owners.

Clearing an account tracking template can affect a broader context than one campaign. Review why it exists and what other tracking setup depends on it before recommending its removal.

A missing or unfamiliar template is not automatically a defect. The freelancer should establish the intended tracking design rather than invent a replacement from generic advice.

Likewise, a Revenue figure appearing in a report is not proof that all tracking is accurate. Confirm which conversion actions and values the measurement includes before interpreting it as sales data.

Record the exact configuration work separately from implementation testing. The client can then see which part the freelancer completed and which website or analytics investigation remains due.

Review exclusions and shared lists

An exclusion should represent a deliberate business decision about what advertising should avoid. Start with the actual search, placement or criterion evidence and the client's offering. A low outcome alone does not establish irrelevance.

The focused negative-keyword additions support exact match at the specified campaign, ad group or shared-list level. Review that match behaviour rather than assuming the action creates a broad exclusion for every related search.

A shared negative list must be attached to campaigns to have the intended effect there. Creating the list, adding text and attaching it are separate operations. The proposal should account for each requested step.

Detaching a list from one campaign does not remove the list or its other associations. Editing its contents can affect attached campaigns more broadly. Choose the operation that reflects the client's intended scope.

Placement exclusions also have valid target levels and kinds. A website, app, YouTube video and YouTube channel are not interchangeable input forms. Review the selected kind and values before confirmation.

Keep the business reason in the work note. A verified exclusion confirms the relevant configuration change, not a quantified future saving. A later performance review can assess observed activity without claiming that every excluded opportunity was waste.

Preserve relevance decisions that came from the client

The client may confirm that a search concerns an unavailable service or an offer they deliberately do not provide. Record that fact with the selected exclusion. It is stronger decision context than a vague judgement that a term looks poor.

Read Performance Max group context

Performance Max settings and signals need their actual campaign and asset-group context. A request about one group should not silently apply to every group in the campaign, especially when their product selections or creative differ.

The focused search-theme action adds specified themes or removes explicit existing resources. Read which operation is being prepared. A suggestion of theme wording is creative assistance until the user requests a live addition.

Attaching an audience signal adds the selected signal rather than replacing all existing signals. The proposal should preserve that distinction so the client understands the resulting configuration.

Creating an audience is a separate account-level operation from attaching it to a group. A combined request can have dependencies, but those should be explicit rather than hidden extra work.

Final URL expansion has a focused action with a provider-preflight requirement and an unestablished historical live-acceptance note. Do not describe the catalogue entry as an unconditional guarantee of a currently verified live write.

A signal or theme does not itself prove a narrowly constrained delivery population. The freelancer should explain the actual supported operation and avoid describing an added signal as a complete replacement of the campaign's audience behaviour.

After the edit, review the intended group and then a relevant outcome period. A correctly attached signal is an account result; its isolated performance effect needs a separate, cautious interpretation.

Distinguish selected product edits from replacement

Product-targeting work can concern an exact selected occurrence or a complete listing selection. These actions have different scope and should not be described with the same generic phrase, “update products”.

The selected inclusion action uses resolved listing leaves and group context. It preserves unrelated partitions and can expand a broader leaf only for the selected products. The relevant selection evidence is part of the request.

The same product can appear in several groups. Including it in one group does not automatically include it in another. Keep campaign and asset-group or ad-group context visible when the client asks for a targeted change.

The focused Performance Max retargeting action replaces the group's product selection. It is broader than toggling an exact selected leaf. Review the complete requested item set before approving that operation.

The focused Standard Shopping replacement supports flat item trees and preserves bids where the inheritance is unambiguous. An ambiguous inherited bid should lead to a refusal or clarification rather than an invented value.

Performance filters need their own period and measurement. Asking to include excluded products with ROAS above a threshold combines relationship evidence with historical performance. A missing performance value should not become an assumed zero.

Read the proposal as a hierarchy decision. The client should understand where each selected product occurrence belongs and whether unrelated selections remain. A unique-product list alone does not describe the complete target scope.

Review the full creative operation

Creative work can be part of settings optimisation, but drafting and uploading are separate tasks. The assistant can suggest headlines or descriptions from supplied context without provider research. A live edit requires a defined target and proposal.

The focused RSA replacement supplies the complete headline and description sets, final URL and display paths. It removes pins, clears omitted paths and retains the existing status. Review all these effects before confirmation.

A new RSA is created paused in a standard Search ad group. This does not enable its parent group or campaign. If the user also requests enablement, the relevant action must appear separately in the proposed steps.

Confirm product facts and commercial promises with the client. Character-limit validation can make a request technically acceptable, but it cannot establish whether a discount, guarantee or delivery statement is true.

URL normalisation does not test the destination. A supplied public address can be prepared for the account request without proving page availability or checkout behaviour. Keep the website owner's investigation distinct.

A verified creative edit does not guarantee policy approval or delivery. The client should understand the remaining review or launch decision rather than treat a successful mutation receipt as proof that the new ad is already generating results.

For an existing attachment or extension, check whether the action changes shared content or only the relationship. A supported existing image-asset reference is not an upload tool for a new image file.

Follow an invented local-service review

The following business and sequence are invented. Birch Repairs provides appointments within an agreed local service area. Its freelance consultant reviews the settings of a Search campaign called Household Repairs after the client raises a concern about geographic relevance.

The consultant first reads the selected geographic criteria and positive location option. The client confirms that the service is intended for people in or regularly present within the served area. That business fact guides the recommendation.

The current geographic set already matches the agreed places, so the consultant does not propose an unnecessary country or radius replacement. The identified decision concerns the presence option rather than the selected locations.

The assistant prepares the supported presence proposal for the exact campaign. The authorised connected owner reviews the action and confirms it. A verified result establishes that the relevant setting changed.

The consultant then plans a later review using a defined period and consistent lead measurement. The new option does not guarantee perfect geographic accuracy, and a lead movement alone would not prove the option's isolated effect.

The client note distinguishes the retrieved setting, confirmed service requirement, verified edit and subsequent review. The example illustrates a focused settings decision rather than a universal instruction to change every campaign to Presence.

A useful decision can leave neighbouring settings unchanged

In this example, the consultant does not clear tracking, replace languages or alter budgets simply because those actions are available. The review concerns one defined mismatch. Other settings require their own reason and evidence before becoming another proposal.

Respect unavailable and conditional actions

A comprehensive catalogue still needs accurate availability information. Some actions are unavailable because the existing implementation lacks a supported endpoint or required verification. Others require a provider preflight or a particular current configuration.

The focused auto-apply subscription action is unavailable; its recorded audit implementation uses an unsupported endpoint. Manage that subscription through Google Ads rather than claiming the Slack proposal can safely change it.

The automatic HTTPS final-URL action is also unavailable because destination verification is missing. A public-URL syntax check does not supply the necessary reachability evidence. Explain the actual boundary.

Existing CPA or ROAS target adjustments require applicable existing targets. A budget action requires the supported non-shared daily configuration. These restrictions should be checked before approval, not discovered through an improvised live write.

Provider support and app verification are separate facts. A field or API method existing in documentation does not prove that the current application has a complete safe workflow for the user's requested operation.

Keep the client response specific. State which selection or prerequisite is needed, or which work must occur in Google Ads. An unavailable action does not justify executing a different partial request without the user's informed decision.

Honest boundaries make the supported tools more useful. The freelancer can plan the actual work without presenting catalogue breadth as proof that every possible setting is editable in every account context.

Sequence related work deliberately

Related changes can have a dependency order. A new group needs its campaign, and a keyword addition needs the group. The local dependency framework describes such relationships for supported requests.

A simple existing-target setting does not require an unrelated creation chain. Local relationship knowledge helps choose the order without making provider reads that the request does not need.

Existing resources still require appropriate resolution. A dependency table cannot prove that a named campaign exists or that its current type supports the intended child. Use account evidence for those factual questions.

Review each logical step in ordinary language. Creating keywords and enabling the group are separate intentions. A proposal should not omit one merely because the keyword list occupies most of the visible detail.

Keep initial statuses explicit for new structures. A paused creation can be intentional staging, while enabling is another decision. Dependency order does not automatically choose the client's launch state.

Do not promise that every multi-write request is transactionally reversible. If the complete result is uncertain, some work may already have applied. The subsequent investigation must establish current state before another live request.

The freelancer should keep related actions coherent rather than combine every available edit into one decision. Clear scope and steps make the confirmation review meaningful and the later work record understandable.

Use Slack for a focused settings conversation

A useful Slack exchange starts in the channel connected to the intended account. Ask for the current setting on the relevant target, then explain the client's requirement or the observed concern.

The assistant can discover available settings locally and retrieve the current account values through the appropriate tools. Those are different stages. A capability explanation should not be presented as an observed configuration.

Follow-up questions should change only the intended part. If the client asks about another option on the same campaign, preserve the campaign context. If the request changes to an account-wide setting, state that scope explicitly.

Creative and explanatory replies can remain conversational. The assistant does not need to call a provider merely to rewrite the recommendation politely or explain a known concept. Current account claims still require evidence.

Ask for a table or chart when it helps the decision, such as comparing relevant campaign performance. A current boolean setting may be clearer in a short sentence than an unnecessary visual.

Before a live instruction, make the target unambiguous. The proposal should reflect the resolved campaign or deeper asset and the exact requested operation. Do not rely on an unclear “it” after a long list of entities.

Keep the thread's outcome readable for the client. The account observation, recommendation, proposal and execution result should remain distinct, so a later reader can see what was actually changed.

Review and confirm the exact proposal

A supported edit is prepared for review before the live request is made. The authorised connected owner must click Confirm and Execute. A conversational agreement or a report recommendation does not substitute for that control.

Review the named target, parent context, exact operation and values. For a replacement, read the complete desired set and any clearing behaviour. For a shared resource, understand the intended associations.

The confirmation authority checks the channel's account owner, Slack installer user and proposal actor. Participation in the channel does not create unrestricted approval rights for every colleague.

The proposal expires after fifteen minutes. If the client returns later, prepare a fresh version from current context. An old recommendation remains useful discussion without becoming permanent permission to mutate the account.

Settings are checked again before execution. If relevant state changed, the proposal can become stale. The appropriate response is a new reviewable request rather than ignoring the changed before-and-after context.

During execution, the visible progress state shows that the confirmed work is being checked and applied. It does not establish completion. Wait for the recorded outcome before adding the task to verified completed work.

For the freelancer, a clear proposal protects the business discussion from ambiguity. The client can approve the exact account behaviour that was recommended, rather than a broad instruction whose implementation remains hidden.

Interpret the execution outcome

A verified result establishes the relevant account outcome of the approved edit. It confirms what was changed or created according to the applicable checks. It does not establish the client's commercial return.

Some operations use provider acknowledgement without independently verified final completion. Keep that narrower status in the client report. A positive acceptance message should not be rewritten as a fully completed asynchronous outcome.

An uncertain result means that a request may have reached Google while the complete outcome could not be confirmed. Some or all work may already have applied. Do not treat uncertainty as proof that nothing changed.

The workflow avoids blind retries of uncertain mutations. Inspect the relevant current target before another request, especially for creations that could be duplicated. A follow-up should concern only the work still needed.

Slack status delivery is separate from execution. The original proposal update and thread receipt communicate the recorded result; their appearance does not imply multiple independent live writes.

Record the exact stage for the client. A pending proposal, verified setting and unresolved execution have different follow-up requirements. This gives a later consultant a reliable account history rather than a list of intentions.

If the edit was rejected before a live write, explain the relevant current-state or validation issue. Correct the request or prerequisite instead of presenting the rejection as a general inability to optimise the account.

Review subsequent evidence consistently

After a verified settings change, choose a defined follow-up question. It may concern current configuration, policy state, delivery or performance. Each requires appropriate evidence and should not be inferred from the execution receipt alone.

Use consistent metrics and populations when comparing periods. A conversion-role change can alter the report's included actions, while a status change can alter the enabled population. Explain those differences before describing a performance movement.

Keep the timeline of other work and confirmed business changes. Stock, offers, website behaviour and additional edits can affect the subsequent result. Avoid crediting every difference to the last proposal.

A short operational check can establish that a setting remains as intended. A commercial evaluation may need another observation window. The freelancer should not present one as a substitute for the other.

If a further edit is recommended, prepare a new proposal using current evidence. The earlier approval does not authorise a recurring series of automatic changes. Each supported live decision retains its target and control.

During handover, preserve the original reason, completed operation and next review point. The next responsible person should know whether a result is verified, accepted or uncertain before they continue the work.

A useful optimisation record closes the loop honestly: what was observed, what the client chose, what changed and what the later evidence shows. It does not need a guaranteed-success narrative to demonstrate responsible account work.

Freelancer settings questions

Should every audit finding be changed?

No. A finding is a review prompt within a defined check. Assess current evidence and the client's objective. Unavailable checks are not passes, and diagnostic conditions do not quantify financial waste by themselves.

Does settings discovery read the account?

The catalogue discovery is local. A current-value answer requires the relevant provider read. Capability information and observed configuration should remain separate in the response.

Does adding a country also change Presence?

The selected geographic criteria and positive location option are distinct. The focused location edit retains Presence settings. Account for both operations when the user explicitly requests both.

Can every daily budget be edited by the focused action?

It supports non-shared daily campaign budgets with defined set or percentage operations. Other configurations need an accurate capability explanation rather than an improvised write.

Does a verified change prove improvement?

No. It establishes the relevant account result. Later performance needs a defined period, consistent measurement and appropriate business context. The execution receipt does not quantify profit or isolate causation.

What makes a settings review useful?

A clear purpose, current evidence and correct target make the decision understandable. Explicit replacement effects, authorised confirmation and an honest follow-up keep the work accountable to the client's objective.