Google Ads Recommendations Report Explained for Clients

Google Ads recommendations report explained for clients

Sooner or later a client forwards you a screenshot from their Google Ads account with a number on it and one line of text: "Google says our account is only 62% optimised. Should I be worried?"

That number, and the list of suggestions behind it, is the google ads recommendations report. Google puts it front and centre in the interface and emails it to account owners, so clients find it. If you do not explain it first, Google explains it for you, and Google's framing is not always the one that serves the client.

This guide translates the recommendations tab into plain English: what the score measures, what each type of recommendation is really asking for, why applying all of them is a bad idea, and how to report on it.

Last updated: August 2026.

Key takeaways

• The Google Ads recommendations report is a list of automated suggestions plus an optimisation score from 0% to 100%, where 100% means Google considers the account fully set up to its own best-practice checklist.
• Optimisation score measures setup and adoption of Google's recommended settings, not results. An account can sit at 100% and still lose money, and an account at 65% can be highly profitable.
• Applying a recommendation is not free. Many bidding, budget and broad match suggestions increase spend or widen targeting, so they are proposals to review, not tasks to tick off.
• Dismissing a recommendation raises your optimisation score too, because dismissed items are removed from the calculation. That makes a deliberately curated score more honest than a chased one.
• Repairs and measurement recommendations, such as broken conversion tracking, are the ones worth actioning immediately. Everything else deserves a judgement call.

What is in this guide

• What is the Google Ads recommendations report?
• What does optimisation score actually measure?
• What are the main recommendation types?
• Should you apply every recommendation?
• How does auto-apply work, and how do you report it?
• How do you put this in a client report?
• FAQs

What is the Google Ads recommendations report?

The Google Ads recommendations report is an automated list of suggested account changes, shown in the Recommendations tab of Google Ads, together with an optimisation score that estimates how closely the account follows Google's recommended setup. It exists at campaign level, account level and manager (MCC) level, so an agency can see one score across a whole book of clients.

Google generates the list from the account's own performance data, settings, campaign types and industry trends, and it refreshes constantly, so the list your client screenshotted on Tuesday is often not the list you see on Thursday. Google documents the mechanics in Google Ads Help, and the same recommendation objects are exposed programmatically through the Google Ads API if you want to pull them into your own reporting.

Two bits of vocabulary worth teaching your client early:

• Optimisation score: the 0-100% figure. Australian and UK accounts show "optimisation"; US accounts show "optimization". Same metric.
• Recommendation: a single suggested change, each carrying an estimated uplift to the score (for example "+0.8%") if you apply it.

Google Ads is one of the most-connected paid channels we see: across the live client-to-platform connections on ReportsMate it accounts for roughly one in six, and the typical client account has about two platforms reporting into it. That matters here, because the recommendations tab only sees Google Ads. It does not know your client also runs Meta, or that half their revenue is organic.

What does optimisation score actually measure?

Optimisation score measures how much of Google's recommended configuration you have adopted. It does not measure profit, lead quality, or campaign performance. This is the single most useful sentence you can say to a client about it.

A 100% score means Google has nothing left to suggest. It does not mean the account is winning. Accounts can sit in the high 90s while pouring budget into broad match traffic that never converts, and accounts in the 60s can be deliberately narrow, tightly negatived and comfortably profitable.

The score is also gameable in a way clients rarely realise: dismissing a recommendation removes it from the calculation, so the score goes up. That is not a loophole, it is the design. Google treats "we considered this and it is not right for this account" as a valid answer.

One place the score does carry weight is the Google Partners programme, which uses a minimum optimisation score across managed accounts as one of its membership requirements. Check the current threshold in Google Ads Help before you quote it to anyone, because Google has adjusted the programme criteria more than once. If your agency needs to hold the badge, say that out loud to clients rather than pretending the score is a performance metric.

What are the main recommendation types in the Google Ads recommendations tab?

Google groups recommendations into categories. Here is what each one is really asking for, in the language you can put in front of a client.

The pattern is easy to teach: the top two rows are hygiene, the middle rows are craft, and the bottom two rows are money and strategy. Nobody should be auto-applying the bottom two.

If you want a fuller vocabulary lesson for clients on the underlying numbers, our guide to Google Ads metrics explained for clients covers impressions, CTR, CPC, conversions and cost per conversion in the same plain-English style.

Should you apply every Google Ads recommendation?

No. Recommendations are generated by an algorithm optimising for Google's view of account health, not for your client's margin, lead quality or capacity to service more work. Treat the list as a second opinion from a well-informed party with its own commercial interests, which is exactly what it is.

Three recurring examples of where blanket application hurts:

Budget increases. "Raise your daily budget to capture more traffic" is arithmetically true and commercially neutral. If the client's cost per acquisition is already at the ceiling their unit economics allow, more volume is more loss. Run the numbers first with a tool like our Google Ads budget calculator so the conversation is about maths, not opinion.

Broad match conversions. Switching exact and phrase keywords to broad match reliably increases reach. It also reliably increases irrelevant search terms unless the negative keyword list is disciplined. For a plumber, "how to fix a tap" is traffic, not a lead.

New campaign types. Adding a Performance Max campaign changes what you can see and control, and changes the shape of the report you send. That is a strategy conversation, not a one-click apply.

The honest version of your job is curation: apply the repairs, apply the sensible asset work, review the targeting, take the money questions to the client, then dismiss what you have consciously rejected.

How does auto-apply work, and how should you report on it?

Auto-apply lets Google implement selected recommendation types on your behalf, without you clicking anything. It is opt-in, controlled per recommendation type, and can be switched on at account level or across a manager account. Google keeps a history of what was auto-applied and, for a limited window, lets you undo changes.

Auto-apply recommendations reporting is where agencies get caught out. If a keyword match type quietly changed on the 3rd and cost per conversion moved on the 8th, and nobody in your team made that change, you will spend the client call guessing. Worse, if the client finds the change history before you mention it, you look like you were not watching.

Our position, from building a reporting product used by agencies across several ad platforms: whatever you auto-apply, report. A workable format:

1. Auto-apply settings in force this month - list the enabled types, or state "none enabled".
2. Changes Google applied automatically - date, what changed, why it matters.
3. Changes your team applied - the recommendations you actioned deliberately.
4. Changes your team dismissed - and the one-line reason.

That fourth item is the one clients remember. "We turned down Google's suggestion to broaden your keywords because it would have brought in DIY searches, not booking enquiries" is the sentence that proves you are managing the account rather than administering it.

How do you put recommendations into a Google Ads account health report?

A Google Ads account health report is a short, recurring summary of whether the account is set up correctly and running clean, separate from the performance numbers. Performance answers "did it work". Health answers "is it in good shape". Clients need both, and they conflate them constantly if you only send one.

A health section that works, in five lines:

• Optimisation score and direction - the number plus whether it went up or down, with one sentence of context.
• Repairs outstanding - anything broken, especially conversion tracking. Zero is the target.
• Recommendations actioned - what you applied and the intent behind it.
• Recommendations declined - what you rejected and why.
• Settings changes - auto-apply status, budget changes, new campaign types.

Keep it to a block near the top of the report, above the metrics. Then let the performance data do its own talking. If your client also has Search Console or Google Business Profile in the mix, the same health-then-performance structure works across all of them, which is how we structure the Google Ads section of an automated report.

Benchmarks belong in the health section too, because a click-through rate on its own means nothing without context. Our breakdown of what counts as a good CTR for Google Ads gives clients that context.

How ReportsMate handles this

We built ReportsMate email-first because, after years around agency reporting, the dashboards clients were handed almost never got logged into. The report that lands in the inbox is the one that gets read, and the account health story is exactly the part clients skip when they have to log in to find it.

Connect the client's Google Ads account through the Google Ads integration with one OAuth click, set a schedule, and the branded report goes out as your agency, from your sender domain, with AI-written commentary in plain English rather than platform jargon. GA4, Meta Ads, Search Console and Google Business Profile land in the same email, so the client sees one story instead of five logins. The full flow is on how it works.

We are obviously not neutral here, so the fair comparison: AgencyAnalytics, DashThis, Whatagraph, Swydo, Supermetrics and Looker Studio all do capable Google Ads reporting, and several offer deeper custom dashboarding than we do. The difference is delivery. Their model centres on a dashboard the client logs into. Ours centres on an email the client opens.

Google Ads recommendations FAQs

Q: What is a good optimisation score in Google Ads?

A: There is no universally correct number, because the score measures adoption of Google's suggested settings rather than performance. Well-managed accounts commonly sit between the high 70s and the low 90s once irrelevant recommendations have been dismissed. If your agency is in the Google Partners programme, check the current minimum score requirement in Google Ads Help, since that is a real threshold with real consequences. Otherwise, chasing 100% is a poor use of your time. A score of 80% built from deliberate decisions is worth more than 100% built from applying everything Google suggested.

Q: Does a higher optimisation score improve my Google Ads ranking?

A: No. Optimisation score does not feed into Ad Rank, Quality Score, or the auction. It is an account-level advisory metric, not a ranking input. Where it can indirectly affect results is when the underlying recommendation genuinely helps, for example fixing broken conversion tracking so automated bidding has accurate signals to work with. The fix helps; the score moving is just a side effect. Google Ads Help sets out the distinction between optimisation score and Quality Score clearly, and it is worth linking a curious client to the source rather than paraphrasing it.

Q: Is auto-apply safe to turn on?

A: For narrow categories, yes; for bidding, budgets and match types, we would not. The reasonable middle ground is enabling auto-apply for low-risk items such as ad asset suggestions, while keeping anything that changes spend or targeting under human control. Whatever you choose, two rules apply: check the auto-apply history regularly, and disclose the setting to the client. An unexplained account change is the fastest route to a client wondering who is actually running their account.

Q: Why did my optimisation score drop without me changing anything?

A: Because the score is recalculated against a moving target. New recommendation types appear, seasonal trends shift Google's estimates, competitor activity changes the auction, and campaigns that were paused or ended alter the mix. A score can fall while the account is performing better than last month. This is exactly why the score belongs in the health section of a report with a sentence of context beside it, never on its own as a headline number.

Q: Can clients see the recommendations tab themselves?

A: Yes, if they have access to the Google Ads account, and Google also emails suggestions to account owners. Assume the client will see it. That is a good argument for getting there first: when your monthly report already explains which recommendations you actioned and which you declined, a Google email becomes confirmation that you are on top of things rather than a prompt for an awkward question.

Q: How often should recommendations be reviewed?

A: Weekly for repairs, monthly for everything else. Repairs represent things that are broken, and a conversion tracking fault left running for three weeks corrupts a month of reporting. The rest can travel on your normal reporting cadence, which is the rhythm at which you review and communicate account changes. Most agencies land on monthly client reports with a weekly internal check, and automating the client-facing half is what frees the time to do the internal half properly.

The short version

Google's recommendations report is a useful input and a terrible boss. Fix what is broken, apply what genuinely helps, dismiss what does not fit the client's economics, and write down the reasoning. That last step turns an algorithmic to-do list into evidence that a human is thinking about the account. Automate the delivery, keep the health summary short, and spend the recovered hours on the decisions that actually move cost per acquisition.

Stop losing your Sundays to client reports. Start your free 14-day trial - no credit card, no setup, cancel anytime. Or compare plans first. Your clients get branded Google Ads reports in their inbox automatically, with the account health story already explained.