Microsoft Ads as an efficiency layer: importing from Google without importing its mistakes
Microsoft Advertising occupies a strange position in most media plans: everyone agrees it is "probably worth doing", almost nobody does it well, and the gap between those two facts is where the efficiency lives. The channel's pitch is genuinely simple — meaningful search volume at CPCs that are typically lower than Google's in English-language markets, reachable by adapting structures you have already paid to prove. The execution failure is equally simple: teams treat the Google import button as the strategy rather than the starting point.
Why the one-click import disappoints
Google Import copies campaigns, keywords, ads, and — dangerously — assumptions. Three of those assumptions break on contact with Microsoft's auction. Bid targets: Google tCPA values were learned in a deeper, more competitive auction; imported verbatim, they either overpay wildly or never win volume. Audience layers: Google's in-market and affinity segments map loosely or not at all, and imported observation layers silently distort delivery. Query behaviour: Microsoft's matching is looser in places, and its audience skews older, more desktop, more B2B-hours — the same keyword buys different intent.
Worse is the standing sync. Auto-reimport every week means every Google experiment, mistake, and Google-specific optimisation flows into an auction it wasn't made for. The account can never learn its own lessons because it is overwritten weekly by someone else's.
The import hygiene checklist
Import once, then cut the cord. Use import for the initial build only. From day one, the Microsoft account is managed on its own performance data.
Re-map conversion goals deliberately. Stand up UET properly — ideally through the same server-side container as the rest of your stack — and map goals one-to-one against the Google conversion actions you trust, so cross-platform CPA comparisons mean something. Import does not do this for you.
Rebuild bid strategy targets from scratch. Start conservative, let the thinner auction reveal its own equilibrium, and expect longer learning at lower conversion volumes. Patience here is cheap; imported targets are not.
Separate the negative lists. Microsoft's query mix differs enough that Google negatives both over-block and under-block. Build Microsoft negatives from Microsoft search term reports.
Strip what doesn't translate. Google audience layers, Google-specific ad customisers, and campaign types without true equivalents come out at import. What stays earns its place on Microsoft data.
Add what Google can't do. LinkedIn profile targeting — bid adjustments by industry, company, and job function — is Microsoft's unique lever, and for B2B and prosumer products it changes the math. Use it as a bid layer on your proven keywords before building anything exotic.
Managing the divergence
A healthy Microsoft account drifts away from its Google parent over time — different top keywords, different device splits, different dayparting. That drift is the efficiency you came for. We review it monthly with one governing question: is the marginal pound here still cheaper than the marginal pound on Google? While the answer is yes, budget flows to the divergence. When it stops being yes, the account has found its natural size — typically a meaningful single-digit to low-double-digit share of the search programme, varying by market and category.
Set expectations accordingly: Microsoft Ads is not a second Google. It is an efficiency layer — smaller, quieter, and, run with discipline, some of the cheapest incremental volume in paid search.