Diversifying beyond Google: a 90-day plan for performance teams

Search diversification fails for a predictable reason: it is run as a side project with no end state. Someone imports the Google account into Microsoft, spend trickles, nobody owns the measurement, and three months later the verdict is “Microsoft didn’t really work for us.” That is not a platform result. It is a planning failure. Here is a 90-day plan that treats the second channel as a deliberate build with a decision at the end, not an experiment left to drift.

Days 1-30 baseline and build

The first month is about two things only: knowing exactly where you stand, and standing up a channel that is built rather than borrowed.

Start with the baseline, because you cannot prove incrementality later if you did not measure carefully now. Record your current Google performance by campaign type: CPA, ROAS, conversion rate, and the share of total demand you are capturing. Note your concentration risk honestly, meaning how much of your acquisition depends on a single auction you do not control. This is the number diversification is meant to reduce, and you want it written down before you start so the comparison at day 90 is real.

Then build the Microsoft channel. Use the import tool to seed structure, but treat it as the starting line and nothing more. The work in month one is rebuilding the parts that should be platform-native: bidding tuned to Microsoft’s lighter competition rather than copied from Google’s auction pressure, the audience layers that have no Google equivalent, and feed and shopping structure set up properly rather than carried over half-configured. Microsoft Ads typically runs at materially lower CPCs than Google, commonly cited at around 33% lower on average, but you only capture that gap if the account is built to spend efficiently, not just to exist.

The most important deliverable of month one is the measurement scaffolding. Conversion tracking verified on both platforms, separate targets set for Microsoft rather than Google’s inherited benchmarks, and a clear definition of what success at day 90 looks like. If you skip this, everything that follows is anecdote. The import is a starting line, not a strategy, and month one is where you prove you understand the difference.

Days 31-60 native optimisation

Month two is where the channel earns its keep, and it is the part most diversification attempts skip entirely. Having built something platform-native, you now optimise it as such, which means leaning into the things Microsoft does that Google cannot.

The headline capability is LinkedIn Profile Targeting, which lets you filter on job title, company and industry and is exclusive to Microsoft Advertising. If your buyer is defined by their role rather than their demographics, this is the single biggest reason the channel exists, and month two is when you layer it onto your search intent properly. There is no equivalent move on Google, so this is genuinely incremental capability rather than a cheaper version of what you already do.

Beyond targeting, this is the month to work the audience profile to your advantage. A meaningful share of Microsoft’s audience is higher-income and desktop-dominant; in the US, around 41% of Bing users earn over $100k. For considered, desktop-led purchases, that skew is a feature, and your messaging, landing pages and bidding should reflect it rather than ignore it. Tune bids where the lighter competition gives you headroom, prune what the first month’s data shows is not converting, and push the campaign types where Microsoft’s audience composition plays in your favour.

The discipline this month is to optimise against Microsoft’s own targets, not Google’s. A third lower CPC against a differently skewed audience is not “Google but worse.” It is a different job, and judging it by the wrong yardstick is how good channels get killed prematurely. Our method is built around exactly this distinction between importing and building.

Days 61-90 incrementality and decision

The final month answers the only question that matters to a board: is this channel adding demand you would not otherwise have captured, or just moving it around?

That is an incrementality question, and you measure it deliberately rather than assuming it. The cleanest read comes from holding other variables steady and watching whether total acquisition rose as Microsoft spend ramped, or whether Microsoft simply cannibalised conversions Google would have won anyway. Geo-based holdouts, controlled spend changes, and careful before-and-after comparison against the baseline you recorded in month one all help. The point is to attribute the lift, or the absence of it, with enough confidence to make a budget call.

Set the comparison up honestly. The metrics that should travel across platforms, conversion rate by intent, CPA against the same target, ROAS against the same margin, are the ones you compare. The absolute inputs, CPC chief among them, you do not, because they were never going to match and matching them was never the goal. The question is incremental return per pound, and by day 90 you should be able to answer it with data rather than instinct.

Then decide. The honest outcomes are three: scale the channel because it is incremental and efficient, hold it at a deliberate size because it adds diversification value even where it does not scale, or wind it down because the evidence genuinely says so. All three are legitimate. What is not legitimate is the non-decision, where Microsoft drifts on at a token budget because nobody ever made the call. The plan exists to force the call.

Scaling from here

If the day-90 decision is to scale, the work changes shape from setup to portfolio management, and a few principles keep it honest.

Run Microsoft and Google as one portfolio with distinct jobs, not as a contest. The question at every review is where the next pound earns the most incremental return, which sometimes points at Google and sometimes at Microsoft, and the right answer changes as each channel matures. Set and revisit Microsoft’s targets on their own terms, because as you scale you will eventually meet the volume ceiling that the lighter competition implies, and you want to recognise that ceiling as a natural limit rather than a failure.

Keep benchmarking on a quarterly cadence so drift does not erode what you built. The channel that gets the smaller share of attention is the one where small problems compound unwatched, and a scaled Microsoft line deserves the same scrutiny as the Google one it now sits beside. The deeper reason any of this is worth doing is concentration risk: a programme that depends entirely on one auction you do not control is fragile in a way that no amount of Google optimisation can fix. Diversification, properly built and properly measured, is how you address that. If you want a partner to run the 90 days with you, get in touch.