Execution
The technical work behind the commercial thinking.
Commercial judgement decides what should happen. This is the layer that makes it happen inside the account. Google and Microsoft Ads are managed in-house by the people who built the plan.
Google Shopping
Catalogue structure first: what is eligible, what is profitable, what is actually in stock. Campaign structure follows the economics of the products rather than the shape of the website.
Performance Max
Asset groups and listing groups built so budget can be steered by product, with exclusions, brand handling and feed labels doing the work. Read through native channel performance, asset group detail and search-term reporting, subject to Google's disclosure thresholds.
Search
Brand, category, competitor and problem-led demand treated as different purchases, with match types and negatives used to keep them apart while there is enough signal to run them apart.
Product feeds
Titles, attributes, identifiers and imagery treated as a ranking and eligibility surface. Custom labels carry margin, stock and lifecycle context into the account. Labels organise the catalogue; they do not set bids on their own.
First-party data
Customer lists, new versus returning treatment, and consented enhanced conversions where the data quality supports it.
Measurement
Tracking validated before anything is optimised: deduplication, conversion definitions, currency and value accuracy, and a check that the account and the finance view can be reconciled.
Tracking
Server and browser events, consent handling, and identifiers that only ever carry what the visitor has agreed to. No event fires on an enquiry that was not actually recorded.
Incrementality
Where budget and account size allow, geo or campaign-level tests separate demand we created from demand that would have arrived anyway.
Campaign architecture
Split by economics and signal volume. Splitting for its own sake fragments learning; leaving everything blended hides the decisions worth making.
Query control
Search terms reviewed as a demand map, not just a negative-keyword chore. What the account is buying should be explainable in a sentence.
Product segmentation
Products grouped by the commercial job they are doing, so budget and targets can differ where the economics differ.
Budget allocation
Allocated against commercial priority and constrained by stock and margin, then reviewed against live signals rather than a monthly habit.
Bidding
Targets set per job. Where the data supports it, conversion values reflect contribution rather than gross revenue, so value-based bidding optimises towards what the business keeps.
Experimentation
Structured tests with a stated hypothesis and a decision rule agreed in advance, so a result changes something.
Microsoft Ads
Microsoft Ads is also available; scope is agreed with you.