Opus Growth

Optimization

Ready-made prompts for Claude, ChatGPT and other assistants connected to your ad accounts. Copy one, fill in the brackets and send.

Optimization

Find search terms that waste budget and turn them into negatives

Combines smart negative suggestions with the full search terms report to flag zero-conversion and overpriced terms, then previews a grouped batch of negatives that protects your brand and converting keywords.

Review the search terms in my Google Ads account [account name] for the last 30 days. Use the smart negative keyword suggestions and the full search terms report together. My brand name and its common spellings are: [brand terms].

Flag two groups:
1. Terms that spent more than [spend threshold] with zero conversions.
2. Terms with exactly one conversion where the cost of that term is above 2x my target CPA of [target CPA]. Calculate this yourself from the cost and conversions in the search terms report.

Skip any term the report lists under already_negative, and check my existing negatives (including shared lists) so you never propose a duplicate.

For the matched keyword, query search_term_view with the keyword segment. If that field does not come back for a term, write "not available" instead of guessing.

Give me one table sorted by cost, highest first. Columns: search term, campaign, ad group, matched keyword, clicks, cost, conversions, recommended negative, match type, level (campaign or ad group).

Decision rules:
- Use exact match for one-off variants.
- Use phrase match when the same unwanted word appears in several terms, for example "free", "jobs" or "DIY".
- Pull my converting keywords from the keyword report. Never suggest a negative that would block one of them or any of my brand terms.
- If a term is relevant but simply expensive, list it separately as a bid or ad copy issue, not as a negative.

End with the estimated monthly savings. Use the last 30 days of cost from the terms you would block as the estimate, and say that this is an upper bound. Then show a preview of the negatives you would add, grouped by campaign, as one batch per campaign. Wait for my approval before adding anything.

Optimization

Block searches for places outside your service area

A list of search terms mentioning locations you do not serve, the location setting that lets them in, and a previewed fix for both.

Optimization

Month-end budget pacing check across all client accounts

Runs pacing on every connected client account, compares projected month-end spend with each client's agreed monthly cap, and proposes CPA-ranked daily budget changes within guardrails, approved account by account.

List my connected Google Ads client accounts (include client accounts under my manager account) and run a budget pacing check on each one for the current month, passing each account's id directly. The monthly caps agreed with each client are: [account name: monthly budget list]. Skip any account that is not on this list and tell me which ones you skipped.

For each account, take month-to-date spend per campaign from a campaign-level performance report for this month, and the daily budget and pace from the pacing check. Calculate projected month-end spend as spend to date plus current daily budget times the days remaining. Add the campaigns up to get the account total and compare it with the client's cap.

Deliver one table sorted by risk. Include these columns:
- account
- campaign
- daily budget
- spend to date
- projected month-end spend
- account variance vs the monthly cap, as an amount and as a %
- flag

Use these flag rules, at account level:
- OVER: projected spend is more than 5% above the cap.
- UNDER: projected spend is more than 10% below the cap.
- ON TRACK: everything else.

For every campaign in a flagged account, add month-to-date conversions and cost per conversion. This shows whether the underspend is worth fixing.

Then, for each flagged account, propose new daily budgets based on the days remaining. The goal is to land total spend within ±3% of the cap by month-end. When adding budget, prioritise the campaigns with the lowest cost per conversion. When cutting, start with the highest. Respect the budget guardrails: no single increase above +50%, no more than +100% within 24 hours, no cut deeper than -80%. If the target cannot be reached within those limits, say so and suggest a second step later in the week or a pause instead.

Show the proposal as a before/after table per account, then a preview of the changes. Wait for my approval account by account before changing any budget. Do not touch bidding strategies or targets.

Optimization

Shift budget from rank-limited to budget-limited Search campaigns

A budget-neutral reallocation plan that moves money to Search campaigns losing impression share to budget at an acceptable CPA.

In [client account], analyse all Search campaigns for the last 30 days. Target cost per conversion for this client is [target CPA].

1. Run an impression share analysis. Put each campaign in one bucket: mainly losing to budget, mainly losing to rank, or not limited.
2. Add cost, conversions, cost per conversion and current daily budget for each campaign.

Decision rules:
- A campaign RECEIVES budget only if both are true:
  - search impression share lost to budget is above 10%
  - cost per conversion is at or below target
- A campaign DONATES budget if either is true:
  - it spends less than 85% of its daily budget
  - its cost per conversion is more than 30% above target
- Rank-limited campaigns get no extra budget. List them separately. Use keyword Quality Score data to say whether the main fix is bids or Quality Score.

Deliver:
- a table of all campaigns with bucket, metrics and verdict (receive / donate / hold)
- a reallocation plan that keeps total daily budget unchanged, with no single campaign changed by more than 30%

Show the before/after budgets. Wait for my approval before applying any change.

Optimization

Target CPA readiness check before switching a campaign's bidding

A go/no-go checklist with evidence and a safe starting target before moving a Google Ads campaign to Target CPA.

I want to move campaign [campaign name] in [client account] to Target CPA. Check whether it is ready before anything changes.

Run these five checks:

1. Volume. Count conversions in the last 30 days at campaign level.
   - 30 or more: ready for Target CPA.
   - 15 to 29: recommend Maximize Conversions without a target first.
   - Under 15: no change.
2. Tracking health. If any primary conversion action is paused, removed or recorded zero conversions in the last 30 days, verdict is NO-GO.
3. Goals. List which conversion categories are primary.
   - Flag any micro-conversion (page view, scroll, click) that is feeding bidding.
   - Show the counting type. For lead generation it should be "One".
4. Stability. Using change history, list budget, bid, keyword or ad changes made to this campaign in the last 14 days.
5. Starting target. Use the actual 30-day cost per conversion, rounded up. Never set it lower than actual.

Deliver a checklist with a pass/fail result and the evidence for each item. Finish with a one-line verdict: GO, GO with Maximize Conversions first, or NO-GO, plus the reason.

If the verdict is GO, preview the bidding strategy change. Wait for my approval before applying it.

Optimization

Meta ad set budget scaling and cutting with a 20% step rule

A rule-based budget plan for Meta ad sets that scales winners and trims losers without resetting learning.

In Meta ad account [ad account], review every active ad set in conversion-objective campaigns. Target cost per result is [target cost per result].

For each ad set, show:
- where the budget is set (ad set, or campaign with Advantage campaign budget)
- daily budget
- spend, results, cost per result and frequency for the last 7 days
- the same metrics for the last 30 days

Decision rules:
- SCALE (+20%, never more in one step): 7-day cost per result is at or below 80% of target, with at least 10 results in 7 days.
- CUT (-25%): both 7-day and 30-day cost per result are more than 130% of target.
- REFRESH CREATIVE (no budget change): 7-day frequency is above 3.5 and cost per result is rising.
- HOLD: everything else, plus any ad set with fewer than 5 results in 7 days.

Check the change history. Exclude anything whose budget or targeting was edited in the last 72 hours.

If the budget sits at campaign level, propose the change on the campaign instead of the ad sets.

Deliver a table with metrics, decision and proposed budget. Wait for my approval before updating anything.

Optimization

Triage every Google Ads recommendation into apply, dismiss or test

A decision table for every open Google recommendation, judged against your ROAS target, followed by a preview of the batch to apply or dismiss.

Account: [Google Ads account]. Our target ROAS is [target ROAS, e.g. 400%].

1. Take an account snapshot and pull campaign-level cost, conversion value and ROAS for the last 30 days.
2. List every open recommendation with its projected impact and the current optimization score.
3. Build one table with these columns: recommendation type, campaign, projected +conversions / +cost, campaign ROAS over the last 30 days, verdict (APPLY / DISMISS / TEST), and a one-line reason.

Decision rules:
- APPLY only if the campaign is at or above target ROAS and the change cannot raise spend without a matching conversion value gain.
- DISMISS anything that adds broad match, raises budget on campaigns below target ROAS, switches bidding to a goal that ignores conversion value, or is purely cosmetic for the optimization score.
- TEST covers anything plausible that is too risky to apply account-wide. For these, name the single campaign to try it on.

Do not apply or dismiss anything yet. Show the exact batch for each action as a preview and wait for my approval.

Optimization

Validate "raise your budget" recommendations against ROAS and lost share

A verdict on each budget recommendation based on marginal ROAS, impression share lost to budget and month-end pacing.

Account: [Google Ads account]. Our minimum acceptable ROAS is [minimum ROAS]. Monthly budget ceiling: [monthly budget].

List only the budget-related recommendations (raise budget, budget-limited campaigns) and check each one against the data before we spend more:
- Last 30 days per campaign: cost, conversion value, ROAS, and conversion value trend week over week.
- Search impression share lost to budget compared with share lost to rank. If rank loss is larger, more budget will not fix it.
- Budget pacing for the current month. Would the raise push total spend past our ceiling?

Output a table with these columns: campaign, current daily budget, Google's suggested budget, ROAS, impression share lost to budget, impression share lost to rank, projected month-end spend, verdict.

Decision rule: approve a raise only where ROAS is at least 20% above our minimum and budget loss is above 15%. In those cases, propose a capped increase of no more than 20% instead of Google's full number. Dismiss the rest.

Preview the budget changes and the dismissals. Change nothing until I confirm.

Optimization

Vet keyword and broad match recommendations with search term evidence

Tests Google's keyword and broad match recommendations against your own search term costs, conversions and ROAS, measures current wasted spend per campaign, and previews every apply, dismiss and add action.

Account: [Google Ads account]. Period: last 60 days. Target ROAS: [target ROAS]. Target CPA, used only if the account records no conversion value: [target CPA].

Google suggests new keywords and a switch to broad match. Judge these recommendations using our own data, not Google's projections:
1. List the recommendations and keep the keyword, broad match upgrade and "use broad match" types.
2. For each suggested keyword, find matching search terms in the search terms report and give cost and conversions. Pull conversion value for those terms with a query on search_term_view for the same 60 days, then calculate ROAS. If the account has no conversion value, judge on cost per conversion against my target CPA and say so.
3. Run keyword opportunities. Converting search terms that are not yet keywords are stronger candidates than Google's list.
4. Run smart negative keywords to see the account-wide leak to non-converting terms. For the per-campaign figure, sum the cost of zero-conversion search terms per campaign and divide by that campaign's total cost from a campaign-level performance report.

Output:
- Table A, keyword recommendations: keyword, search term evidence, ROAS (or cost per conversion), verdict.
- Table B, broad match recommendations: campaign, current wasted spend from non-converting terms as an amount and as a % of campaign cost, verdict. Dismiss broad match wherever that waste is above 10% of campaign cost.
- A list of keywords from our own converting terms (ROAS above my target) to add as exact or phrase match instead, each with the ad group it should go into, based on the ad group where the term already converts.

Preview every apply, dismiss and add action. Make no changes until I approve.

Optimization

Decide on ad and asset recommendations using real asset performance

A review of Google's ad, RSA and extension recommendations against actual headline, description and extension results, with your own rewrites where Google's fix is weak.

Account: [Google Ads account]. Campaign: [campaign name]. Period: last 30 days.

Google recommends improving ad strength, adding headlines and adding sitelinks, callouts or images. Before accepting any of them:
1. List the ad and asset recommendations for this campaign.
2. Read each RSA with its ad strength and Google's action items.
3. Pull headline and description performance: impressions, CTR, conversions, conversion value and conversion rate.
4. Pull extension performance per asset.

Then decide:
- If an "add headlines" recommendation would fill slots with generic text, dismiss it. Instead, write 3 replacement headlines based on our top converting headlines (price, shipping, returns, product benefit), each 30 characters or fewer.
- Flag headlines with 1,000+ impressions and zero conversions for replacement.
- Apply extension recommendations only for types we lack entirely. Skip new sitelinks if existing sitelinks have low CTR.

Deliver a table (recommendation, evidence, verdict) plus the proposed RSA edits. Preview every edit, apply and dismiss action and wait for my approval.

Optimization

Rank cross-channel opportunity recommendations by revenue impact

A prioritized top-10 list of Ads, Search Console and GA4 opportunities, each scored by revenue potential, with the ones to skip clearly marked.

Store: [website URL]. Google Ads account: [Google Ads account]. Period: last 28 days.

1. Generate the cross-channel opportunity recommendations and the unified marketing overview.
2. Enrich each opportunity with revenue context:
   - Ads items (wasteful keywords, keywords to add): cost, conversion value and ROAS.
   - Search Console page-2 quick wins: impressions, position, and whether the page is a product or category page. Check GA4 to see if that landing page generates purchases or revenue.
   - GA4 items: sessions, key events and revenue by landing page or source.
3. Score each item by estimated monthly revenue impact (high / medium / low) and effort (one change / several changes / content work).

Output: a ranked top-10 table with these columns: opportunity, channel, evidence, estimated revenue impact, effort, recommended action. Below it, list the opportunities to skip, with a reason. Typical reasons are informational queries with no purchase intent, pages that never convert, or spend too small to matter.

For any action involving a change in Ads, WordPress or Search Console, describe it as a proposed step only. I will ask for a preview separately before anything is changed.

Optimization

Match your Google Ads schedule to the hours you can answer calls

A day-and-hour performance breakdown with a proposed ad schedule and bid modifiers built around your real call-answering hours.

Review the ad schedule of my Google Ads campaign [campaign name] against the hours my team actually answers the phone: [business hours, e.g. Mon-Fri 08:00-19:00, Sat 09:00-14:00].

1. Read the current ad schedule and any bid modifiers.
2. Pull the last 90 days by day of week and hour (segments.day_of_week, segments.hour): cost, clicks, phone calls, conversions, cost per conversion and conversion rate.
3. Group the results into blocks: early morning, business hours, evening, night and weekend. Show one row per block with the metrics above and its share of total spend.

Decision rules:
- A block outside business hours with zero conversions after 20+ clicks, or with cost per conversion above 1.5x [target CPA]: propose -30% to -100%.
- A block inside business hours with cost per conversion at least 20% below target: propose +10% to +20%.
- A block with fewer than 20 clicks: leave it unchanged and flag it as low data.

4. Show the proposed schedule as a table (day, start, end, modifier) next to the current one, with the estimated spend that would shift.

Do not change anything yet. Show me the preview and wait for my approval before applying it.

Optimization

Stop paying for clicks from outside your service area

A district-level location report showing out-of-area spend, plus a previewed fix for the location option, exclusions and missing targets.

Audit where the leads for my Google Ads campaign [campaign name] really come from. My service area is: [cities or districts].

1. Read the current location targeting, exclusions and location option (presence vs presence or interest).
2. For the last 90 days, pull performance by the user's physical location from user_location_view and by area of interest from geographic_view: cost, clicks, phone calls, conversions and cost per conversion, down to city or district.
3. Report the following:
   - percentage of spend from outside the service area, with the top 10 locations by spend;
   - districts inside the area with 10+ clicks and cost per conversion above 1.5x [target CPA];
   - districts inside the area that convert well but are only covered by a broad city or province target.

Recommendations:
- If out-of-area spend is above 10%, switch the location option to presence only.
- Exclude out-of-area locations that spent money without converting.
- Add the strong districts as separate targets so they can be managed individually. Use geo_target_search for the correct IDs.

Show every change as a preview (setting, location name, ID, add or exclude) and wait for my approval before applying it.

Optimization

Prune weak RSA headlines and descriptions without losing ad history

A per-ad-group verdict table for every RSA asset, plus in-place replacement copy previewed for approval.

Review the responsive search ad assets in [campaign name] for the last 60 days, using asset-level data.

Only judge headlines and descriptions with at least 1,000 impressions. Flag an asset as weak if either of these is true:
- Its CTR is below 70% of the ad group's median asset CTR.
- It spent more than [target CPA] with zero conversions.

Deliver one table per ad group with these columns:
- asset text
- pinned position
- impressions
- CTR
- conversions
- cost per conversion
- the ad's current Ad Strength
- verdict (keep / replace / watch)

For each weak asset, write a replacement. It should:
- fit the same pinned position and character limit
- stay consistent with the ad's other messages
- avoid duplicating any existing headline

Limit changes to two assets per ad so the effect stays measurable. Do not propose any edit that would push Ad Strength below Good.

Use in-place RSA edits, not new ads, so each ad keeps its performance history. Show all proposed edits as a preview and wait for my approval before applying anything.

Optimization

Detect and pause fatigued Meta ads before they drain budget

A ranked list of fatigued Meta ads by wasted spend, creative patterns behind the fatigue and a safe pause list.

Find fatigued ads in the Meta ad account [account name].

Pull ad-level data for the last 28 days, split into two 14-day halves. For each half, include:
- spend
- impressions
- frequency
- link CTR
- results
- cost per result

Ignore ads with fewer than 5,000 impressions in the period.

Flag an ad as fatigued only when all three of these are true:
- Frequency is above 3.5.
- Link CTR fell by 25% or more from the first half to the second.
- Cost per result in the second half is above [target CPA].

For each flagged ad, read its creative: format, headline, primary text and opening hook.

Deliver:
1. A table of fatigued ads ranked by spend above target, showing campaign and ad set.
2. A short note on recurring patterns, such as which angles, formats or offers are wearing out fastest.
3. A pause list.

Never pause the last active ad in an ad set. If pausing would leave an ad set empty, mark that ad set as needing new creative instead.

Show the pause list as a preview and wait for my confirmation before pausing anything.

Optimization

Microsoft Ads ad-level pruning: edit or pause losing ads per ad group

Head-to-head ad comparisons inside each Bing ad group, with in-place RSA rewrites and a strict pause list.

Audit the ads in the Microsoft Advertising account [account name] at ad level for the last 90 days.

In every ad group with two or more active ads, compare the ads on:
- CTR
- conversion rate
- cost per conversion

Only judge ads with at least 1,000 impressions.

Flag an ad as the loser when both of these are true:
- Its CTR is 30% or more below the best ad in the same ad group.
- It has fewer conversions than that best ad.

Deliver one table per ad group showing, for each ad:
- the first headlines
- impressions
- CTR
- conversions
- cost per conversion
- verdict

Then propose actions using these rules:
- For losing RSAs, prefer editing the weakest headlines and descriptions in place over pausing, because rebuilding an ad resets its history. Write the new copy.
- Propose pausing only ads that lose on both CTR and conversions and have at least 30 clicks behind them.

Separately, list any ad groups that are running a single ad, since they have nothing to test against.

Show every edit and pause as a preview and wait for my approval before applying.

Optimization

Before/after impact report of ad and asset pruning for management

A change log, a before/after KPI table and a plain-language management summary of what the pruning achieved.

Over the past [number of days] days, I paused ads and replaced assets in Google Ads [Google Ads account] and Meta [Meta ad account]. I need to report the impact to management.

Step 1: Pull the change history from both accounts for that window. List every:
- ad pause or removal
- RSA, asset or extension edit
- creative swap

Include who made the change, when, and the old and new values where available.

Step 2: For each affected campaign, compare the 14 days before the first change with the 14 days after the last change, on:
- spend
- clicks
- CTR
- conversions
- cost per conversion
- ROAS, where conversion value is tracked

Flag any campaign where cost per conversion worsened by more than 15%, so I can consider reversing those changes.

Deliver:
1. A change log table.
2. A before/after table by campaign and platform.
3. A 5-sentence management summary in plain language. It should state the net effect on cost per conversion and the spend no longer going to weak ads. It should also name any caveats: seasonality, budget or bid changes made in the same period, and learning phases.

This is read-only. Do not change anything in either account.

Optimization

Pause Meta ad sets that overspend against your target CPA

A verdict table of Meta ad sets that are losing money, the daily spend you would free up, and pause actions prepared for your approval.

Optimization

Scale winning Meta ad sets with safe 20% budget steps

A shortlist of proven Meta winners with capped budget increases that avoid resetting the learning phase, ready for your approval.

Find the Meta ad sets in [ad account name] that deserve more budget, and propose a careful increase.

Use the last 14 days and compare them with the 14 days before. A winner must meet all three conditions:
- Cost per result at least 20% below [target CPA].
- At least 10 results in the period.
- Stable or improving cost per result versus the prior period.

For campaigns that use Advantage+ campaign budget, propose the change at campaign level instead of ad set level.

Increase rules:
- Raise each winner's daily budget by at most 20%.
- Keep the total added daily spend under [max extra daily budget]. If this cap forces a choice, prioritize the lowest cost per result.

Deliverable: a table with campaign or ad set, current daily budget, proposed budget, % change, cost per result now versus the prior period, and results. In two sentences, explain why the increases are capped at 20%.

Show me a preview of each budget change and wait for my approval before applying anything. Finally, tell me which date to re-check the results, 3 to 4 days after the change.

Optimization

Keep combined Meta and TikTok spend inside your monthly cap

A month-end spend projection across Meta and TikTok with budget cuts or headroom suggestions that land you on your monthly cap.

Optimization

Check whether recent Meta budget changes hurt your results

A timeline of Meta budget and status changes with before/after cost per result, a verdict for each, and reverts prepared for approval.

Optimization

RSA headline challenger test without losing ad history

Finds the weakest headlines inside your existing responsive search ads, writes challengers that test new angles, and edits the ads in place after your approval so their history stays intact.

For Google Ads campaign [campaign] in account [account], run a headline challenger test inside the existing responsive search ads instead of building new ones.

1. Pull RSA asset performance for this campaign for the last 60 days (use min_impressions so tiny assets drop out). Then read each ad's full copy, current pins and Ad Strength with get_ad_details. If the asset report does not return conversion rate for a headline, calculate it from conversions and clicks. If neither conversions nor clicks are usable for a row, judge that headline on CTR only and say so.

2. For each ad, identify up to 2 weak headlines. A headline counts as weak if it has at least 1,000 impressions and its CTR or conversion rate is below 70% of that ad's average across its headlines with 1,000+ impressions. If an ad has fewer than 3 headlines above 1,000 impressions, skip it and tell me there is not enough data yet.

3. For each weak headline, write 2 challenger headlines of 30 characters or fewer. Each one must test a different angle (price, proof, urgency, specific benefit). Do not just write synonyms of the old line. The first challenger replaces the weak headline. Add the second one only if the ad stays at 15 headlines or fewer; otherwise keep it as a backup for the next round.

4. Keep existing pins as they are. If a weak headline is itself pinned, give its replacement the same pin position and flag it in the table, because changing a pinned line changes every impression of that position.

Output a table with these columns: ad group, ad ID, current headline, pin, impressions, CTR, conversions, proposed replacement, angle tested, new character count, and Ad Strength now.

5. Edit the ads in place with update_rsa using the partial mode (remove_headlines and add_headlines), so each ad keeps its ID and performance history. Run it with confirm=false first, show me every preview, and apply only the ones I approve.

6. After applying, tell me the exact date and time of the change and confirm it with get_change_history (resource type AD_GROUP_AD). I will save that date myself so we can later compare equal 30-day windows before and after with rsa_asset_performance.

Optimization

Meta creative test readout: rank ads and pause the losers

A ranked ad-level table with hook and hold rates, a winner explanation and a pause list waiting for approval.

In Meta ad account [account], ad set [ad set] is running a creative test that started on [test start date]. My target cost per result is [target CPA].

Pull ad-level results since the start date: spend, impressions, link CTR, CPC, results and cost per result. For video ads, also calculate the hook rate (3-second plays divided by impressions) and the hold rate (ThruPlays divided by 3-second plays). Read each ad's creative so the table also shows format, headline and the first line of primary text.

Decision rules:
- Loser: the ad spent at least 2x the target cost per result with zero results, or its cost per result is 30%+ above the ad set average after 1,000+ link clicks.
- Winner: the lowest cost per result among ads with at least 20 results.
- Every other ad is still in test.

Give me the ranked table. Then add a short note on what the winning creative does differently (hook, offer, format, length) and what to test next. Propose which ads to pause, preview the list, and wait for my approval before changing anything.

Optimization

LinkedIn messaging test: which ad angle wins with B2B buyers

Reads the real copy of every active LinkedIn ad in a campaign, tags each by message angle, compares performance by angle and by ad, drafts new variants of the winner and proposes pauses for your approval.

In LinkedIn ad account [account], compare the messaging of all active ads in the Campaign Manager campaign [campaign] (an API campaign group) over the last 30 days.

1. Find the campaign group with list_linkedin_campaign_groups, then its ad sets with list_linkedin_campaigns, then the active creatives in each ad set with list_linkedin_creatives. If any list comes back truncated, tell me it is incomplete.

2. Read each ad's actual copy. For post-based ads (Sponsored Content, Document, Thought Leader) the copy lives in the post, so pull it with linkedin_ad_copy using the creative ids.

3. Pull per-creative metrics for the last 30 days with linkedin_campaign_analytics (pivot CREATIVE, narrowed to those ad set ids): impressions, clicks, CTR, spend, leads or conversions, and cost per lead. Check list_linkedin_conversions to see which conversions are attached to these ad sets. If no conversion is attached and the ads do not use a lead form, say so and judge on CTR and cost per click only.

4. Based on the copy, tag each ad with one message angle: pain point, ROI or numbers claim, social proof, product feature, or offer such as a demo or guide. Where an ad mixes angles, tag the one that dominates the headline and first line, and mark it "mixed" so I can review your call.

Report two tables: one by angle (summed) and one by creative, with the angle tag and a one-line quote of the copy.

Rules:
- Only draw conclusions from creatives with 3,000+ impressions.
- Treat CTR differences under 0.1 percentage points as noise.
- If an angle has only one qualifying creative, call its result directional, not proven.

Name the winning and losing angles and explain why in one or two sentences. Write 2 new ad variants of the winning angle for the next round. Keep intro text under 150 characters and headlines under 70. These are drafts for me to publish; if the ad set runs text ads, also give a version that fits 25 characters for the headline and 75 for the description.

Propose pausing creatives whose CTR is below 0.4% and that have no conversions after [spend threshold] in spend. Show the list, run update_linkedin_creative_status with confirm=false as a preview, and change nothing until I confirm. Skip creatives that are under review, since those cannot be paused, and tell me which ones they are.

Optimization

Test validity check: were the results contaminated by mid-test changes

A read-only audit of Google Ads and Meta change history, experiments and conversion tracking during a test window, ending with a Valid, Valid with caveats or Rerun needed verdict per test.

Before I report test results to client [client name], check whether the tests were contaminated. This is a read-only review. Make no changes.

Test window: [start date] to [end date].
Google Ads account: [Google Ads account], Meta ad account: [Meta ad account].
Tests and arms: [test name: arm A = campaign or ad set name/ID, arm B = campaign or ad set name/ID; repeat per test].

1. Confirm the accounts with list_accounts and list_meta_ad_accounts, and match each arm I listed to its real ID with list_meta_campaigns and list_meta_adsets (for Google, use the IDs that appear in the change history). If an arm name does not match anything, stop and ask me.

2. Google: pull get_change_history. It only reaches back 30 days, so if [start date] is older than that, tell me which part of the window you could not see. Filter the rows yourself to the test dates and the arm campaigns. Also run list_experiments (action list) and note any experiment that overlapped the window or locked an arm.

3. Meta: pull meta_change_history with since and until set to the test window, then keep only events on the arm campaigns, ad sets and ads.

4. List every change that could bias a result:
- budget changes above 20%
- bid strategy or target CPA/ROAS changes
- ads added, paused or edited
- keyword, negative keyword, audience or targeting edits
- conversion action or event changes

5. Tracking: run conversion_tracking_health and list_conversion_actions for the window on Google. On Meta, check list_meta_pixels (last fired) and list_meta_custom_conversions (never fired). A paused or silent conversion action must not be mistaken for a losing variant. Meta activity history does not show every pixel or event change, so if tracking looks off but no matching change is logged, say it could not be traced.

Output a timeline table with these columns: date, platform, user, entity, change, old value, new value, affected test arm, and impact (High, Medium or Low). High means the change hit only one arm, or it touched budget, bidding or conversions. Account-level changes affect all arms equally; mark them as such.

End with a verdict per test: Valid, Valid with caveats, or Rerun needed, with one line of reasoning each. List any blind spots (days outside the Google 30-day history, untraceable tracking changes) under the verdicts.

Optimization

Cut search terms that spend without selling or miss target ROAS

Flags Search and Shopping search terms that burn budget with no conversions or deliver less than half your target ROAS, proposes exact or phrase negatives at the right level, and applies them only after you approve the preview.

Review the search terms for [account] for the last 30 days. Include only Search and Shopping campaigns. My target ROAS is [target ROAS, e.g. 400%]. Our brand names: [brand names]. Products or product lines we sell: [product names or categories].

1. List the campaigns with list_entities and keep only Search and Shopping ones.

2. The search terms report returns clicks, cost and conversions but not conversion value, so pull the per-term data with google_ads_query on search_term_view for the last 30 days: search term, campaign, campaign type, ad group, clicks, cost, conversions and conversions_value, filtered to Search and Shopping channels. Use search_terms_report alongside it for the already_negative bucket. Note that Google hides very low-volume terms, so the list is never complete.

3. Flag every search term that meets either rule:
(a) cost above [minimum spend] with zero conversions, or
(b) at least 1 conversion, at least 20 clicks and ROAS (conversion value / cost) below half of my target.
If the account records no conversion value at all, skip rule (b) and tell me.

4. Exclude terms that contain any of our brand names or a product name from my list. If my product list is short, you can also pull product titles from shopping_performance_view with google_ads_query and treat them as protected.

5. Check list_negative_keywords and the already_negative bucket first so you do not duplicate an existing negative, including those in shared lists. Use smart_negative_keywords for its estimated savings and note where its suggestions agree with yours.

Deliver a table sorted by cost with these columns: search term, campaign, ad group, clicks, cost, conversions, conversion value, ROAS, rule hit, proposed match type and level.
- Match type: use exact by default. Use phrase only when a modifier is the problem, such as "free" or "used".
- Level: ad group when the term is fine for the account but wrong for that group; campaign when it is wrong for the whole campaign.

End with the total spend these negatives would have saved over the 30 days. Then run add_negative_keyword as one bulk call per campaign (and per ad group where the level is ad group) with confirm=false, show me the previews and apply nothing until I approve.

Optimization

Stop brand queries from inflating non-brand campaign ROAS

How much branded revenue sits inside non-brand campaigns, their true ROAS without it, and brand negatives to separate them.

Optimization

Seed TikTok search ad negatives with lessons from Google

Turns your proven Google Ads negatives and zero-conversion search terms into a prioritised negative keyword set for TikTok search ads, previewed per ad group before anything is added.

We run TikTok search ads in [TikTok account] alongside Google Ads in [Google account]. Before scaling TikTok, apply what we have already learned on Google.

1) Pull our Google negatives with list_negative_keywords (campaign and ad group level). Then use list_shared_negative_lists and get_shared_negative_list to read the keywords inside every shared negative list.

2) Run search_terms_report for the last 90 days with min_waste set to [minimum spend]. Take the terms with cost above that amount and zero conversions. Skip anything in the already_negative bucket, it is already covered by step 1.

3) Group everything from steps 1 and 2 into non-buyer intent themes that also apply on TikTok, such as free, tutorial, DIY, used, jobs, wholesale and price comparison sites. Use your judgement for the grouping. Drop anything specific to our Google setup, such as location negatives, competitor or brand negatives used to route traffic, and structural negatives between campaigns.

4) List our TikTok campaigns and ad groups with list_tiktok_campaigns and list_tiktok_adgroups. If the ad group data shows whether search ads are enabled (search_result_enabled or a search placement), keep only those ad groups. If it does not, show me the full ad group list and ask me which ones run search ads before going further: [ad groups with search ads, if known].

5) For each of those ad groups, read the current search negatives with manage_tiktok_search_negative_keywords (operation GET) so we do not propose duplicates.

Deliver a table grouped by theme with these columns: proposed TikTok negative, match type (BROAD, PHRASE or EXACT), the Google evidence behind it (cost and conversions over 90 days, or the existing list or level it comes from), and the ad groups it should be added to. Keep the set under 100 terms, prioritised by Google cost. Flag any term you are unsure about, for example words that can also signal a buyer on TikTok, such as "review" or "price".

Then show the manage_tiktok_search_negative_keywords ADD preview (confirm=false) per ad group and wait for my approval before applying anything.

Optimization

Negate low-ROAS paid queries where you already rank top 3 organically

Joins Google Ads search terms with Search Console rankings, finds low-ROAS paid queries you already win organically, previews exact match negatives and sets up a 4-week before/after check.

Compare paid search terms in [account] with organic rankings for [site] over the last 90 days. Our brand terms are: [brand terms].

1) Run google_ads_query on search_term_view for the last 90 days, returning campaign id and name, search term, clicks, cost, conversions and conversion value. Calculate ROAS per term (conversion value divided by cost). Keep the non-brand terms with ROAS below [target ROAS] and at least 30 clicks. If a term runs in several campaigns, keep one row per campaign.

2) Use gsc_list_sites to confirm the property for [site], then gsc_performance with the query dimension for the last 90 days, sorted by impressions with a limit high enough to cover the terms from step 1. Match each paid term to its organic query. Treat exact text matches as matches and list near matches separately for me to confirm, do not assume them.

3) Classify each query:
- A: average organic position 1 to 3 and organic CTR above 20%. Candidate for an exact match negative, because the paid click likely replaces an organic click.
- B: position 4 to 10. Keep paid, organic is not strong enough yet.
- C: not ranking. Keep paid, and review bids separately.

4) Check list_negative_keywords so nothing already blocked is proposed again.

Deliver a table with these columns: query, campaign, paid cost, paid conversions, paid ROAS, organic position, organic clicks, organic CTR and class.

For class A only, show one add_negative_keyword preview per campaign (confirm=false), passing all terms in a single batch as 'term|EXACT' at campaign level, and wait for my approval. Exact match negatives only block that exact query, so close variants will still show; say if any variant looks worth adding.

Then write a 4-week test plan: record today's baseline (paid clicks, paid cost, organic clicks for each class A query), and give me the exact calls to rerun in 4 weeks, google_ads_query for the same terms and gsc_performance with the query filter, both for the 28 days before and after the change. Success rule: total clicks (paid plus organic) on these queries drop by less than 10% while ad spend falls. If it fails, tell me which negatives to remove.

Optimization

Target CPA readiness check before switching bid strategy

A go or no-go verdict on moving a lead campaign to Target CPA, with the tracking fixes required first.

I want to move [campaign name] in Google Ads to Target CPA. Before changing anything, check whether the campaign is ready.

1. Conversion tracking health over the last 30 days:
   - List any primary conversion action that is paused, removed, or recording zero data.
   - Confirm that calls from ads and lead form submissions are both tracked and set as primary.
   - Flag page views, duplicate actions, or micro-conversions that are counted as primary.
2. Conversion settings: for lead actions, counting should be "one", not "every". Report the click-through lookback window and the attribution model for each action.
3. Volume: report the campaign's conversions and CPA for the last 30 days.

Decision rule:
- 30 or more conversions in 30 days with clean tracking: switch to Target CPA. Set the starting target at the actual 30-day CPA, not lower.
- 15 to 29 conversions: use Maximize Conversions without a target for now.
- Under 15 conversions, or broken tracking: stay on the current strategy and fix tracking first.

Give me the verdict, a prioritized fix list, and the proposed bid strategy change as a preview. Do not apply it until I approve.

Optimization

Meta lead ad set budget shifts based on cost per lead and lead quality

An ad set scorecard comparing two periods, plus budget increases, cuts and pauses that respect the learning phase.

Review the ad sets in my Meta lead campaign [campaign name]. Compare the last 14 days with the 14 days before.

1. For each ad set, report spend, leads, cost per lead, CPM, CTR, and frequency.
2. Download the instant form leads for the last 14 days. Count leads per ad set and flag the share with duplicate or invalid phone numbers. When ranking ad sets, use cost per valid lead.
3. Check whether the campaign uses Advantage campaign budget. If it does, propose a campaign budget change instead of ad set budgets.

Decision rule:
- Cost per valid lead at or below [target CPL] and frequency under 3: increase budget by 20%.
- Cost per valid lead above 1.5x target after spending at least 2x target: cut budget by 30%. Pause if there are zero leads.
- Frequency above 4: flag creative fatigue and recommend no budget increase.
- Make no more than one budget change per ad set every 3 to 4 days, so the learning phase does not reset.

Deliver the scorecard table and the list of proposed changes. Show the changes as a preview. Wait for my approval before updating anything.

Optimization

ChatGPT Ads test budget guardrails and ad group bid review

Spending limits that cap a ChatGPT Ads test, plus evidence-based bid changes for each ad group.

I am testing ChatGPT Ads for lead generation with a fixed test budget of [test budget] for this month. Put guardrails in place and review the bids.

1. Show the current account spending limits and spend so far. Include both the recurring daily limit and any date-range windows.
2. Confirm that the conversion pixel has received events recently and that the lead event is attached to the campaign. If it is not, stop and tell me before discussing bids.
3. For each campaign and ad group, list the budget and max bid. Then report the last 14 days by day: impressions, clicks, spend, CPC, and conversions.
4. For any ad group with zero or very low impressions, run the delivery diagnosis first. Raise the max bid by 15% only if the bid is named as the blocker.
5. If an ad group has spent 2x [target CPL] with no conversions, propose lowering its bid by 20% or pausing it.

Proposed guardrails:
- A date-range limit equal to the test budget for this month.
- A recurring daily limit of [daily cap].

Show every limit and bid change as a preview. Wait for my approval before applying anything.

Optimization

Triage every open Google Ads recommendation: apply, dismiss or test

A decision table that puts every open recommendation into apply, dismiss or test, with the real cost of each projected conversion.

In Google Ads account [account], list every open recommendation with its projected impact and the current optimization score. For context, pull the last 30 days of cost, conversions and cost per conversion for each campaign.

Sort every recommendation into one of three buckets:
- **Apply:** clear upside, consistent with our target CPA of [target CPA], and no loss of control.
- **Dismiss:** any of the following:
  - it raises spend without a conversion case,
  - it loosens match types or targeting we restricted on purpose,
  - it duplicates something already in place.
- **Test:** plausible but unproven. Say how you would test it.

Deliver one table with these columns:
- recommendation type
- campaign
- projected extra conversions
- projected extra cost
- implied CPA of those extra conversions (extra cost / extra conversions)
- bucket
- one-line reason

Flag every recommendation whose implied CPA is more than 30% above target. Rank the Apply bucket by projected extra conversions.

Do not apply or dismiss anything yet. Show me a preview of the Apply batch and the Dismiss batch as two separate lists, and wait for my explicit approval of each.

Optimization

Vet "raise your budget" recommendations before saying yes

A per-campaign verdict on each budget recommendation, based on budget-driven lost impression share, CPA and month-end pacing.

In Google Ads account [account], isolate every open recommendation that raises a campaign budget. For each affected campaign, check four things:
- search impression share lost to budget versus lost to rank, from the impression share analysis
- cost per conversion over the last 30 days
- month-end pacing from the budget pacing alerts
- the projected extra cost and extra conversions in the recommendation itself

Decision rule:
- **Approve** only if all three conditions hold:
  - impression share lost to budget is above 15%,
  - the last 30-day CPA is at or below [target CPA],
  - total projected monthly spend for the account stays under [monthly budget cap].
- **Dismiss** if most of the impression share loss comes from rank. Say what to fix instead: bids, Quality Score or ads.
- **Dismiss** if CPA is above target.

Output a table with these columns:
- campaign
- current daily budget
- recommended daily budget
- impression share lost to budget
- impression share lost to rank
- CPA
- pacing status
- verdict
- reason

Add a final line with the total extra monthly spend if every approved item goes live. Preview the approve and dismiss actions, and wait for my approval before applying anything.

Optimization

Validate bidding strategy recommendations before switching to Smart Bidding

A go or no-go on each bidding recommendation, based on conversion volume, tracking health and whether the right goal is primary.

In Google Ads account [account], list every open recommendation that changes a teklif stratejisi. That includes Maximize Conversions, Target CPA, Target ROAS and Maximize Conversion Value.

Before judging any of them, check the measurement behind each campaign:
- conversion tracking health, including any paused actions or actions with zero data
- which conversion goals are primary
- the counting type of each primary action (one versus every)
- the attribution model of each primary action

A recommendation passes only if all four conditions hold:
1. The campaign has at least 30 primary conversions in the last 30 days.
2. The primary goal is a real lead or sale, not a page view or a micro-event.
3. No primary action has recorded zero data in the period.
4. For Target CPA, the suggested target is within 20% of our real CPA of [target CPA].

Output a table with these columns:
- campaign
- current strategy
- recommended strategy
- conversions (30 days)
- primary goal
- tracking status
- verdict
- what must be fixed first, if the verdict is no

Keep tracking fixes as a separate list. Do not mix them into the recommendation decisions.

Preview the apply and dismiss actions and wait for my approval before changing anything.

Optimization

Judge ad and asset recommendations against actual RSA performance

A decision on each ad recommendation, plus our own replacement headlines where a change is justified, instead of generic suggested copy.

In Google Ads campaign [campaign], list the open recommendations that touch ads. That covers improving responsive search ads, adding headlines or descriptions, and raising ad strength.

For each ad they target, do two things:
- Read its full copy, ad strength and Google's action items.
- Pull asset-level performance for the last 60 days: impressions, tıklama oranı, conversions and cost per conversion per headline and description.

Decision rules:
- Never remove or replace an asset that produced conversions in the period, even if ad strength is "Poor".
- A recommendation is worth acting on only if both of these hold:
  - the ad has fewer than 10 headlines,
  - or it has assets with over 1,000 impressions and zero conversions.
- Dismiss pure "improve ad strength" items on ads that already beat the ad group's average CTR and CPA.

Where acting makes sense, do not use generic copy. Write 3 headlines (max 30 characters) and 1 description (max 90 characters) based on our offer: [key offer or USP].

Output:
- a table with these columns: ad, recommendation, asset evidence, verdict
- your proposed copy changes

Preview every RSA edit and every dismissal, and wait for my approval before making any change.

Optimization

One-page memo: platform recommendations we accept and decline this month

A management-ready memo that consolidates Google Ads, TikTok and cross-channel recommendations into a short accept and decline list with reasons.

I report to management monthly and need a one-page view of what the platforms are recommending and what we decided. This is read-only. Do not apply or dismiss anything.

1. Get the last 30 days funnel overview across Google Ads, Search Console and GA4 as context.
2. Pull the open Google Ads recommendations for [Google Ads account], with projected impact and optimization score.
3. Pull TikTok's suggestions for [TikTok advertiser id].
4. Pull the cross-channel opportunities.

Merge everything into one memo:
- **Top 5 to accept:** each item gets the platform, the action, the expected impact in conversions or cost, and the effort (low, medium, high).
- **Top 5 to decline:** each item gets a plain-language reason a non-specialist understands, for example "raises spend without lowering cost per lead".
- **Optimization score note:** one paragraph explaining why we do not chase 100%.
- **Risks:** anything that needs a tracking or budget decision from management.

Rank by expected business impact, not by the platform's own priority. Keep it under 300 words and use no jargon without a short explanation.

Optimization

Build one shared negative list from Ads and Search Console data

A themed shared negative keyword list built from non-converting paid terms and low-intent organic queries, ready to attach to all campaigns after approval.

I run a [business type] and want one shared negative keyword list that protects every campaign in my Google Ads account.

Start by reading my existing shared negative lists so you do not duplicate anything. Then collect candidates from two sources:

1. Google Ads search terms from the last 180 days, across all campaigns, with clicks and zero conversions.
2. Search Console queries for [site URL] from the last 3 months that show clear non-buyer intent: "how to", "what is", "free", "pdf", "jobs", "salary", "course", "DIY", "template".

Rules:
- Only include a word if it never appears in a converting Ads search term.
- Prefer phrase match for theme words.
- Use exact match for specific phrases.

Group the candidates by theme: jobs and careers, free and DIY, education and training, unrelated products, other.

Show me:
- the proposed list name
- each negative with its match type, theme and evidence (cost and clicks, or organic impressions)
- the campaigns you would attach the list to

This is a preview only. Do not create or attach the list until I approve.

Optimization

Clean up Microsoft Ads search terms and sync the gaps from Google

A Bing waste report plus a filtered list of Google negatives missing in Microsoft Ads, added only after you confirm.

My Microsoft Advertising account [Microsoft account] was imported from Google Ads a while ago. I have added negatives in Google since then, but not in Microsoft.

Step 1. Pull the Microsoft search terms report for the last 30 days. List terms with cost above [spend threshold] and zero conversions. Columns: search term, campaign, ad group, matched keyword and match type, clicks, cost.

Step 2. Read the current negatives on my Microsoft campaigns and ad groups. Compare them with all negatives in my Google Ads account, including shared lists. List the Google negatives that are missing in Microsoft.

Step 3. Filter that gap list. Microsoft traffic behaves differently, so only keep a Google negative if one of these is true:
- the word actually appears in Microsoft search terms
- it is universal junk, such as jobs, free or DIY

Final output: one table of proposed Microsoft negatives. Columns: negative, match type, campaign or ad group, source (Bing waste or Google sync), evidence.

Show a preview and wait for my approval before adding anything.

Optimization

Set up negative keywords for TikTok search ads

Reads the negatives already on your TikTok search campaigns and ad groups, checks whether query-level data exists, and otherwise builds a short starter list from your Google Ads negatives, previewed before anything is added.

I run TikTok ads in [TikTok ad account] with search placement on, and I have never set negatives there. My Google Ads account is [Google Ads account].

Step 1. List my TikTok campaigns and ad groups. Note which ad groups have search results enabled if the ad group data shows it. For each campaign and each search-enabled ad group, read the negative keywords already set (GET).

Step 2. Try to pull search query data for my ad groups for the last 30 days with the integrated report.
- If query-level rows come back: list queries with spend above [spend threshold] and zero conversions, plus queries with CTR below 0.5%.
- If TikTok does not return a search query dimension: say so clearly and do not invent queries. Instead, read my Google Ads negatives (campaign and ad group level plus shared negative lists) and build a starter list from them. Keep only universal low-intent words such as free, jobs, DIY, tutorial and review-only terms. Leave out anything product-specific to Google campaigns.

Step 3. Propose negatives per ad group (or per campaign if the same words apply to every ad group in it). Table columns: negative, campaign or ad group, match type (broad, phrase or exact), reason, evidence (spend and conversions, or "from Google list"). Skip anything already set from Step 1.

Keep the starter list short, around 15 to 30 words. TikTok search volume is small and over-blocking will hurt delivery.

Show a preview of each addition, one batch per ad group and match type, and wait for my approval before adding anything.

Optimization

Fix under-delivering TikTok ad groups: bid cap or budget?

Diagnose each active TikTok ad group as bid-limited, budget-limited, inefficient or too early from 7-day data against your target CPA, then preview one bid or budget change per ad group for approval.

In TikTok ad account [advertiser account], list the active campaigns and their ad groups. My target CPA is [target CPA]. TikTok's minimum daily budget for my account currency is [minimum daily budget].

For each ad group, show these values if the listing returns them:
- bidding method (lowest cost, cost cap or bid cap)
- current bid or cap
- daily budget, and whether the budget sits at ad group level or campaign level (check the campaign listing for campaign budgets)

If any of these fields are missing from the tool output, list which ad groups are missing which fields. Ask me to paste them from TikTok Ads Manager before you diagnose those ad groups.

Then pull the last 7 days at ad group level with tiktok_report_query (data_level AUCTION_ADGROUP, dimension adgroup_id):
- spend
- conversions
- cost per conversion (CPA)
- CTR
- CPM

Calculate average daily spend as 7-day spend divided by 7. Then show that figure as a % of the daily budget.

Diagnose each ad group with these rules, checked in this order:
- TOO EARLY: fewer than 10 conversions in 7 days. No change.
- BID-LIMITED: cost cap or bid cap, average daily spend under 50% of budget, and CPA at or below target. Propose a 10 to 15% increase to the bid or cap.
- BUDGET-LIMITED: average daily spend over 95% of budget, with CPA at or below 90% of target. Propose a budget increase of up to 20%.
- INEFFICIENT: CPA above 130% of target, with at least 2x target CPA spent. Propose either a 25% budget cut or a 10% lower cap, not both.
- Anything else: HEALTHY, no change.

Change rules:
- If the budget is set at campaign level, propose the budget change on the campaign instead. Flag every ad group it affects.
- Only propose bid or cap changes for ad groups already on cost cap or bid cap. Setting a bid switches an ad group to manual bidding, so never add one to a lowest cost ad group.
- No proposed budget may fall below [minimum daily budget].
- Never propose more than one change per ad group.

Deliver a table with these columns: ad group, bidding method, current bid/cap, budget (level), average daily spend %, conversions, CPA, CTR, CPM, diagnosis, proposed new value.

Then run each change as a preview with confirm=false:
- bids and caps with update_tiktok_adgroup
- ad group budgets with update_tiktok_adgroup_budget
- campaign budgets with update_tiktok_campaign_budget

Show me the previews and wait for my approval before applying anything with confirm=true.

Optimization

LinkedIn under-delivery check: bids vs LinkedIn's suggested range

A per-campaign diagnosis of why LinkedIn campaigns underspend, with new bids proposed inside LinkedIn's suggested range.

In LinkedIn ad account [ad account], list the active campaigns with their bidding type, current bid and daily budget.

Pull the last 14 days for each campaign:
- spend
- spend as a % of daily budget
- impressions
- clicks
- CTR
- conversions or leads
- cost per lead

For every campaign spending under 70% of its daily budget, run two checks:
1. Get LinkedIn's suggested bid, the bid floor and ceiling, and the minimum daily budget for its current targeting.
2. Get the audience forecast for the same targeting.

Diagnose each one as:
- BID TOO LOW: the current bid is below the suggested bid.
- AUDIENCE TOO SMALL: the forecast is under 50,000 members. Do not raise the bid. Suggest which targeting facet to widen.
- BUDGET BELOW MINIMUM.
- DELIVERING FINE.

For BID TOO LOW campaigns, propose a new bid at the suggested level, capped at [max bid]. If cost per lead is already above [target cost per lead], propose no bid increase and say so.

Deliver a table with the diagnosis, evidence and proposed bid. Preview the changes and wait for my approval before updating any campaign.

Optimization

Check conversion data before accepting any bidding recommendation

A readiness check of conversion values and goals that decides whether Google's bidding strategy recommendations are safe for a ROAS-driven store.

Account: [Google Ads account].

Google is recommending bidding changes, such as Maximize Conversions, Target CPA or Target ROAS. Before I accept any of them, check whether our conversion data can support value-based bidding:
1. List all bidding strategy recommendations with the campaign and the projected impact.
2. Audit conversion tracking health over the last 30 days: paused, removed or zero-data actions.
3. Check which conversion goals are primary. Confirm that Purchase is primary, and flag secondary-worthy actions such as add to cart or page views that are currently set as primary.
4. Read the conversion settings for purchase actions: counting type (it should be "one" is wrong for purchases, it should be "every"), default value, currency and lookback window.
5. Per campaign, report the number of purchase conversions and the conversion value over the last 30 days.

Verdict rule: recommend Target ROAS only where the campaign has 30+ purchases with real values in 30 days and tracking is clean. Dismiss Maximize Conversions or Target CPA recommendations on shopping and PMax campaigns, because they ignore order value. List any tracking fixes needed first.

Preview the dismissals and wait for my approval.

Optimization

Screen TikTok's optimization suggestions before acting on them

A ROAS-checked shortlist of TikTok's budget, bidding, targeting and creative suggestions, with previewed changes for the ones worth acting on.

TikTok ad account: [TikTok advertiser account]. Target ROAS: [target ROAS]. Period: last 14 days.

1. Pull TikTok's own recommendations for the account.
2. For every ad group a recommendation touches, report spend, purchases, purchase value, ROAS, CPA and frequency for the last 14 days, and compare them with the previous 14 days.
3. Classify each recommendation:
   - ACT: the ad group is above target ROAS and stable, and the suggestion is a budget increase, bid adjustment or audience expansion.
   - IGNORE: the ad group is below target ROAS, has fewer than 10 purchases in 14 days, or the suggestion would reset learning on a winning ad group.
   - CREATIVE: creative fatigue suggestions. Flag any ad group whose CTR dropped by more than 25% versus the prior period.
4. For ACT items, propose the exact change. Cap budget increases at 20% per step so learning is not reset.

Output a table: recommendation, ad group, ROAS current vs prior, purchases, verdict, proposed change.

Show a preview of every budget, bid or targeting change and wait for my confirmation before updating anything.

Optimization

Right-size your radius targeting using Business Profile and Ads data

Checks whether a campaign's radius center matches the Business Profile address, groups conversions into distance bands (exact where Google provides distance data, clearly labeled as approximate otherwise) and proposes the smallest radius that keeps most conversions.

Check whether the radius targeting on my Google Ads campaign [campaign name] fits where my customers actually are.

1. Read my business address from Google Business Profile, then pull the profile's insights for the last 90 days: calls, direction requests and website clicks. Use these totals as context for how much local demand comes through the listing itself.
2. Read the campaign's current proximity targeting: center latitude and longitude, radius and units. Tell me, as best you can, which area the center falls in and whether that looks consistent with the Business Profile address. There is no geocoding tool, so if a center correction might be needed, ask me for the business coordinates (copied from Google Maps) before proposing a new center.
3. Pull distance data for the last 90 days:
   - First try a GAQL query on distance_view for this campaign (segments.distance_bucket) with cost, clicks, conversions and cost per conversion. This works when location assets from Business Profile are linked to the account. If it returns data, group the buckets into 0-5 km, 5-10 km, 10-20 km and 20+ km.
   - If distance_view returns nothing, query user_location_view (or geographic_view) by city or district instead, resolve the location names with geo_target_search, and assign each one to an approximate band based on its distance from the business. Label these bands as approximate and list every location with its assigned band so I can correct any mistakes.
4. Show a table by distance band with spend, spend share, conversions, conversion share and cost per conversion.

Decision rule: recommend the smallest radius that keeps about 85% of conversions. Keep an outer band if its cost per conversion is below [target CPA], even if its volume is small. If the center is off and I have given you the correct coordinates, propose moving it there.

Show the current and proposed radius side by side, with the share of past spend and conversions that falls inside the proposed radius. Then prepare the change with manage_campaign_proximity (action update, confirm=false), show me the preview and wait for my confirmation before changing the proximity target.

Optimization

Google Ads experiments readout: declare winners and end losers

Compares the base and trial campaigns of every running and recent Google Ads experiment, gives a winner, loser or inconclusive verdict against your KPI and prepares the end actions for your approval.

Review the campaign experiments in Google Ads account [account name]. My primary KPI is [CPA or ROAS].

1. Find the customer_id for [account name] with list_accounts. Run list_experiments (action list) and keep the experiments that are running or ended in the last [30] days. Note each one's id, type, start and end dates and the campaigns it locks.

2. Map the arms. Each experiment has a base (control) campaign and a trial campaign. If list_experiments does not make clear which is which, run a GAQL query on experiment_arm (experiment_arm.name, experiment_arm.control, experiment_arm.campaigns, experiment_arm.traffic_split) to identify them.

3. Pull metrics per arm with google_ads_query on the campaign resource, filtered to the base and trial campaign ids and to segments.date between the experiment's start date and its end date (or yesterday if still running): metrics.cost_micros, metrics.clicks, metrics.conversions, metrics.conversions_value. Calculate cost per conversion, ROAS (conversion value / cost) and conversion rate (conversions / clicks) yourself. If a query fails or an arm returns no rows, say so for that experiment instead of guessing.

Return one table per experiment with both arms, the % difference for each metric, the days run, the traffic split and a verdict based on these rules:
- Winner: the trial beats the base on my primary KPI by at least 10%, the test ran 14+ days, and each arm has 30+ conversions.
- Loser: the trial is 10%+ worse under the same volume conditions.
- Inconclusive: any other result. Using the current daily conversion pace of the slower arm, estimate how many more days are needed to reach 30 conversions per arm.

Flag experiments that have run for more than 6 weeks without reaching the volume threshold. They are wasting the traffic split.

For winners, write a one-line recommendation I can act on in the interface (apply the change to the base campaign, or promote). Do not promote anything.

For losers and stale tests, propose ending them. Do not end anything yet. Run list_experiments with action end and confirm=false for each one as a preview, show me the exact list with experiment ids and the campaigns that will be unlocked, and wait for my approval before calling it with confirm=true.

Optimization

Diagnose TikTok creatives by hook, hold and offer, then prune

Every TikTok video ad classified by where it fails, with a specific fix per ad and a guarded pause list.

Rank the video ads in the TikTok ad account [advertiser name] for the last 21 days on creative quality, not cost alone.

Pull these metrics per ad:
- spend
- impressions
- CTR
- 2-second view rate
- 6-second view rate
- average watch time
- 100% completion rate
- conversions
- cost per conversion

Only judge ads with at least 3,000 impressions.

Classify each ad into one of four groups:
- Weak hook: 2-second view rate is under 25%.
- Weak hold: 2-second view rate is fine, but 6-second view rate is under 10%.
- Weak offer: watch metrics are healthy, but cost per conversion is above [target CPA].
- Winner: none of the above apply.

Deliver a table grouped by ad group showing the metrics and classification for each ad. For every non-winner, add one practical fix, for example:
- re-cut the first 2 seconds
- move the product earlier
- shorten to under 15 seconds
- change the CTA or offer

Propose pausing only ads that are both classified as weak hook and below their ad group's average conversions. Keep at least two active ads in every ad group.

Show the pause list as a preview and wait for my approval before changing any status.

Optimization

LinkedIn creative review: pause losers and learn which messages win

A creative scorecard by campaign, a messaging analysis of winners vs losers and new copy variants to test.

Review all active creatives in the LinkedIn ad account [account name] over the last 60 days.

Pull these metrics per creative:
- impressions
- clicks
- CTR
- spend
- leads or conversions
- cost per lead

Then read the actual ad copy of each creative.

Only judge creatives with at least 2,000 impressions. Flag a creative as underperforming if either of these is true:
- CTR is below 0.40% (for single-image Sponsored Content).
- Cost per lead is more than 1.5x [target CPL].

Deliver:
1. A table grouped by campaign, showing a copy excerpt, the metrics and a verdict (keep / pause / replace with new variant) for each creative.
2. A short analysis of what separates winners from losers. Look at headline style (stat-led, question-led, pain-led), the offer (demo, guide, webinar) and copy length.
3. Two new copy variants per campaign, built on the winning patterns and ready for me to review.

Propose pausing flagged creatives only where the campaign still keeps at least two active creatives.

Show the status changes as a preview and wait for my approval before applying them.

Optimization

Find and pause fatigued TikTok ads before they drain budget

A ranked list of TikTok ads whose CTR and cost per conversion are deteriorating, with pause actions previewed and new-creative needs flagged.

Optimization

Promote best-selling search terms to keywords and give each one owner

Finds converting search terms that are not keywords yet, adds them as exact match to the right ad group, and ends cross-campaign overlap with exact negatives, all behind one approved preview.

Find the search terms in [account] that sell but are not keywords yet. Look at the last 60 days.

1. Start with keyword_opportunities (its "converted but not added" list) and search_terms_report for the same 60 days.

2. The search terms report does not return conversion value or the triggering keyword, so pull them with google_ads_query on search_term_view: search term, campaign, ad group, segments.keyword.info.text and match_type, clicks, cost, conversions and conversions_value. If the account records no conversion value, judge on conversions and cost per conversion instead of ROAS, and tell me so before going further.

3. Get every existing positive keyword with list_entities (entity keywords) or keyword_report, and every negative with list_negative_keywords.

Keep only the terms that meet all three conditions:
- at least [minimum conversions, e.g. 3] conversions
- ROAS at or above [target ROAS]
- not already an exact or phrase match keyword anywhere in the account, and not blocked by an existing negative

4. For each term, suggest the ad group it belongs to and add it as exact match. Choose the ad group whose name, existing keywords and landing page theme are closest to the term; if two fit equally, take the one where the term already converts best. Mark any assignment you are unsure of as "check" so I can decide.

5. Flag every term that was triggered by two or more campaigns. The owner is the campaign whose ad group is the closest product match by the rule above; if that is a tie, the campaign with the higher ROAS on that term owns it. In the other campaigns, propose an exact match negative so each term has a single owner. Never propose a negative in the owner campaign.

Deliver a table with these columns: search term, triggering keyword and match type, campaign(s), conversions, conversion value, ROAS, destination ad group, owner reason, and action.

6. Run add_keywords (one call per destination ad group, match type EXACT) and add_negative_keyword (one bulk call per campaign, terms as 'text|EXACT') with confirm=false. Show all previews together and wait for my approval before applying any of them.

These prompts work best when your AI assistant is connected to your accounts with Opus Growth.

Connect your accounts free