Google Ads management: a weekly optimisation routine

Google Ads management drifts to one of two extremes in most teams: either the account is left alone for weeks, or bids are poked at daily and the system's learning is undermined. Between the two sits a routine that takes roughly an hour a week and has a defined order. In this article we set out that routine in six steps, explaining which order to work in and why. We do not give target figures; every account has a different baseline. What we give is which data each decision should be made from.

Why weekly, and not daily?

Automated bidding strategies learn from data, and every significant change made during the learning period restarts that process. Daily intervention never leaves the system a stable baseline. Monthly maintenance, on the other hand, is too late: if budget is flowing to the wrong place, it has flowed there for four weeks. A weekly rhythm is long enough for most accounts to accumulate statistically meaningful data, and short enough to contain the error.

On low-volume accounts, stretching this rhythm to two weeks may be more appropriate. The criterion is not the calendar but the data: if not enough conversions have accumulated to support a decision, postponing it that week beats deciding on a guess. We summarised how we run the routine on the enterprise side on our paid media service page.

The six-step routine

1. Search terms report. The first and highest-return step of the routine. Open last week's search terms and read where the budget actually went in real queries. Add the irrelevant ones to the negative list and note the ones performing well that are not yet keywords. This step alone closes the largest source of waste in most accounts.

2. Conversion tracking health. Before making an optimisation decision, make sure the data is correct. If the conversion count has unexpectedly dropped to zero or doubled, that is usually a tagging change rather than performance. There is no such thing as a right decision made on wrong data; if this step is skipped, the remaining five are wasted.

3. Budget distribution. Check where the budget is constrained between campaigns and where it is going unspent. Keeping an underspending campaign alive while another campaign consistently exhausts its budget and hits its target is a common mistake. Keep changes small; large one-off budget jumps restart the learning period.

4. Ad copy and assets. Change the underperforming headlines, but not all of them at once. If you change everything simultaneously, you cannot learn what worked. One or two asset changes a week compound into something measurable within a few months.

5. Quality and landing page signals. Google defines ad quality in search campaigns through three components: expected clickthrough rate, ad relevance and landing page experience. This trio is a diagnostic tool, not a number to optimise directly. The component flagged below average tells you where to look that week. If the landing page side is weak, the fix is on the page rather than in the account; our conversion and user experience work addresses that side.

6. Recommendations and optimisation score. Review the recommendations list in the account interface, but do not apply all of them. Recommendations are generated automatically and some do not suit your scope; match type expansion and auto-apply settings in particular call for a deliberate decision. Note which ones you applied and why; that note can be the only record explaining what happened in the account three months later.

What to leave out of the routine

There is also work that should not be done weekly. Changing the bidding strategy, rebuilding the campaign structure, moving the target ROAS or CPA, and opening a new campaign type all belong on that list. These are monthly or quarterly decisions; when they enter the weekly routine the account stays permanently in learning and no period can be interpreted on its own.

Another common mistake is trying to fix a bad week. A single week's fluctuation is noise in most accounts. Look for at least two consecutive periods and sufficient conversion volume before deciding. Hasty intervention is one of the most frequent causes of cutting off a trend that was correcting itself.

Tying the routine to a record

What makes the routine durable is writing down what was done. Three lines a week is enough: what changed, why it changed, which metric to watch. Without this record, nobody remembers the reason for a performance change three months later and the account history turns into a guessing game.

In monthly reporting, these weekly records supply the reasons underneath the charts. We gathered our structure for which report feeds which decision on our reporting page. If you want to discuss scope, fee models and KPIs, the Google Ads consulting article is a better starting point.

In short, Google Ads management is a matter of discipline rather than brilliance. Reading the search terms, verifying the data, directing the budget to the right place and writing down what you did: when those four are done consistently, the account is already ahead of most competitors. Save the remaining time for the big structural decisions.

Monthly and quarterly decisions

The work we left out of the weekly routine does not disappear, it simply moves to a longer cycle. In the monthly cycle you look at these: whether the campaign structure still fits the business objective, whether the target metrics are realistic, how the new keyword opportunities accumulated from search terms should be placed into the structure, and changes in competitor visibility.

The quarterly cycle holds bigger questions: is the channel mix right, does the budget split between search and other channels reflect business results, is the measurement infrastructure still adequate? Keeping these decisions out of the weekly routine matters; when they mix in, the big decisions get rushed and the account stays in permanent flux.

Separating the cycles has another practical benefit: expectation management within the team. When it is clear what the weekly report covers and what the monthly report covers, strategic questions are not asked every week and operational detail is not debated every month. Meeting length shortens accordingly.

Three principles that hold whoever runs the account

Whether you run it in-house or bring in outside support, three principles stay the same. First, no change is made without a rationale; if the rationale cannot be written down, it is too early for the change. Second, no decision is taken before the data is verified. Third, only one major variable is moved at a time.

These three principles look dull, and they are. But what determines the outcome in ad accounts is usually not a brilliant idea; it is catching the error early and making the correction consistently. The routine exists to do exactly that.

Tie the routine to a document rather than to a person. When the person managing the account changes, a one-page document setting out the six steps lets the successor follow the same order from the first week. The most frequent cause of lost continuity in company accounts is the routine remaining a habit that nobody wrote down.

Sources

• Google Ads Help — About optimization score: what the optimisation score is, how recommendations are generated, and how the score differs from Quality Score.
• Google Ads Help — About Quality Score for Search campaigns: the expected clickthrough rate, ad relevance and landing page experience components and their diagnostic use.