My short answer: in the first 30 days I diagnose more than I optimise. When taking over an account it's tempting to start changing bids right away, but every decision made without knowing what you measure rests on luck.
Week 1: Is measurement right?
I verify the Google Tag Manager container, GA4 events, Google Ads conversion actions and the Meta Pixel / Conversions API setup one by one. For e-commerce, the Merchant Center feed and currency/tax settings too. Far more often than people think, "poor performance" is actually missing or double-counting conversion tracking.
Week 2: Reading the account and the market
Account structure, search-term report, audience definitions, negative lists, creative variety. What are competitors doing, what does seasonality look like, which country/market takes which share? By the end of this week I have a "where does the budget go" map.
Week 3: Quick wins
Low-risk, clear-impact fixes: wasteful search terms, wrong location/language targeting, broken landing pages, forgotten campaigns. These build trust and free up budget.
Week 4: The plan
On day thirty I present the client or management a three-column list: stop, scale, test. Every item has a hypothesis, a success metric and a date. The real optimisation starts that day.
"Optimisation on wrong data is not optimisation."