Microsoft Advertising: The Paid Channel You Forgot

Most Australian paid search accounts I audit have never tested the second search network, and the reason is rarely strategy. It is habit. One platform works, the team knows it, and the question of what else is available never gets asked.

That is a fair position up to a point. It stops being fair when a business is paying more each year for the same clicks and has never spent a fortnight finding out whether there is extra demand next door.

This is written for the business already spending on paid search that wants a straight answer on whether Microsoft Advertising deserves a share of it. It covers what the network is today, who searches there and in what frame of mind, the operational differences that matter when you run it alongside Google Ads, the settings that quietly decide your results, the questions a report has to answer before anyone scales the channel, and the honest signs that it is the wrong fit. This is Microsoft Advertising only. It is not a Google Ads fundamentals guide, it does not cover performance campaigns, it avoids the channel comparison argument, and it does not relitigate quality measurement or keyword exclusion mechanics.

What Microsoft Advertising is today

The search network and its partners

Microsoft Advertising is the advertising platform for Microsoft's search and audience properties, and the largest part of its reach still comes from search. Its own product material puts the scale plainly. Per Microsoft Advertising's product site, "Microsoft Advertising reaches over 1 billion people monthly across Bing, Edge, Outlook, LinkedIn, and Xbox".

Search results on Bing and its partner sites are where most of the paid search activity sits. The other properties matter for reach and for audience targeting, which is a different job from capturing somebody who is actively looking for what you sell.

The network also serves outside Microsoft's own pages. Partner and syndicated search placements extend the reach, and those placements are the first setting I check, because a click from a partner site can behave very differently from a click on a Bing search result.

Display, video and retail placements explained

Beyond search, the platform sells display placements, video inventory and retail placements across its own properties and partner sites. Those behave like display anywhere. They are useful for reach and for remarketing, and weaker for capturing someone who is looking for something specific right now.

For a first test, search is where I would start. It maps most closely to the campaigns already running elsewhere, which makes the results comparable, and it is where the audience difference actually shows up in the numbers.

Who is actually searching there

The audience and the buying context

The audience skews differently from the market leader's, and that difference is the reason to test at all. Microsoft's own material makes one version of the point. Per Microsoft Advertising's product site, "Nearly 50 million users in the USA don't use Google for search".

Australia is a smaller market than the United States, and I would not carry a United States figure straight across to an Australian account. What travels is the reason those users exist. The network is strong with people who are at work, on a desktop, inside a Microsoft environment, and with an older and more professional skew than the market leader.

The buying context usually decides whether that suits you. Somebody searching on a work device during the working day behaves differently from somebody searching on a phone in the evening, and the enquiry that follows looks different too.

Which Australian businesses fit the network

In my experience it suits businesses selling to professional and older audiences, and businesses whose buyers are at work when they search. It is common in B2B, where the same person may use a work profile in one ecosystem and a personal one in the other.

It tends to suit professional services, software and technical products, healthcare practices with an older patient base, and trades serving commercial clients. It is a harder case for consumer products bought impulsively on a phone, where the audience overlap with the market leader is close to total.

How running it differs from running Google Ads

Importing campaigns, and what does not copy over

The fastest way to start is to import an existing campaign. The platform copies structure, keywords, ads and much of the targeting, which is why it is the usual first move.

What does not copy over is the part that catches people out. Network and partner selection, some audience settings, and anything specific to the platform's own ad formats all need checking after an import. I treat an imported campaign as a first draft that needs editing before it is judged.

Bidding, audiences and review differences

Bidding runs on familiar lines, with automated strategies taking more decisions over time. The difference that matters to a marketer is the audience layer. Because the platform sits inside Microsoft's own ecosystem, it can target profiles and settings that exist there, which is genuinely useful in B2B where a buying committee shares an environment.

Ad review is the other operational difference. The reviewer, the rules and the turnaround are not the same as on the other platform, so a campaign that goes live within the hour somewhere else can take longer here. Build that into the launch date rather than assuming it.

Judging whether the channel earns its place

Testing the extra reach instead of counting it

Reach is a claim until it is tested, and the honest test is narrow. The question is whether the network produces enquiries you would not have received otherwise, and that is answerable inside a fortnight of spend. Per Microsoft Advertising's product site, "With Microsoft Advertising, you can connect with more than 1B people where they spend over 4 hours a day working, browsing, and playing across Microsoft experiences."

A test that answers the question is small and bounded. One campaign, the offers that already work elsewhere, a distinct identifier so the enquiries can be attributed, and a date when the results get reviewed. Without the identifier, you will never know which channel earned the enquiry, and the review turns into an argument.

Measuring enquiries when clicks are few

Lower traffic means fewer clicks, and fewer clicks means movements that look like trends. A week where cost per click rises 30% on twelve clicks is not a signal. It is noise.

Judge the channel on enquiry quality over a longer window. Count the enquiries, check they are genuinely new rather than duplicated from another channel, and look at what they were worth rather than only what they cost.

The settings that decide your results

Network and partner selection

The most consequential setting is where your ads are allowed to appear. The default includes partner and syndicated placements, and those can carry a different quality of click from the search results you thought you were buying.

I would run search only for a first test, with partner placements excluded or reported separately, so the results mean something. They can always be switched on once you know what the search side delivers on its own.

Location, device and audience adjustments

Then check the adjustments, which quietly shape who sees the ads. Location targeting should match the area you actually service rather than the account default, device adjustments should reflect where the enquiries come from, and audience adjustments are worth layering on once there is data to support them.

One of these is worth checking on day one. If the location setting is wider than the area you service, the first month's spend will be spent teaching you that the phone enquiries from outside the area do not convert, which is an expensive way to learn something you already knew.

Reporting that shows the real picture

Separating network performance from noise

Report the search network separately from the partner placements, and report this channel separately from the other one. Blended figures tell you very little at this stage, because a small increment of spend disappears inside a larger account's average.

Watch the enquiry count alongside the cost. A channel can look efficient on cost per click while producing nothing that reaches a sales conversation, and that is the version of a good report that hides the truth.

Comparing enquiry quality, not just cost

The number that matters is the cost per quality enquiry, compared with what the business already pays elsewhere. That means whoever handles the enquiries has to be part of the review, because they are the ones who know which ones were real.

Microsoft's reporting of its own business is a separate measure, and worth reading carefully. Per Microsoft's FY2026 Q4 earnings press release, "Search advertising revenue excluding traffic acquisition costs increased 10% (up 9% in constant currency)". That is Microsoft's own audited revenue growth for the quarter ended 30 June 2026. It measures the company's revenue rather than advertiser performance, and it tells you the channel is being invested in rather than what you should expect to earn from it.

A realistic first ninety days

Structure, spend split and a review date

Ninety days is enough to answer one question: does this channel bring enquiries the business would not otherwise have received. Structure the test so it can answer that and nothing else.

One campaign, one ad group per service, the offers already proven elsewhere, a dedicated enquiry identifier, and a review date in the calendar from the day it launches. Fund enough clicks for a decision and no more. If you need to work out where a bounded test sits against everything else you fund, the marketing budget calculator gives you the arithmetic rather than an opinion.

What to check before scaling up

Before adding spend, confirm four things. The enquiries are new rather than duplicated from other channels. Somebody is following them up. The cost per quality enquiry is defensible against what the business already pays. And the accounts receiving the conversions are wired correctly, because a duplicate conversion tag will make a modest channel look excellent.

Signs the channel is not right for you yet

Thin demand in your category

The clearest sign is thin demand. If the searches your business depends on barely happen on this network, no amount of optimisation will manufacture them, and a test will spend real money proving something you could have checked in an afternoon.

Look up the terms you rely on and see whether they appear at all. Where demand is thin, the honest answer is to leave the channel alone and revisit it in a year. That is a cheaper recommendation than most agencies will give you, and it is the right one more often than people expect.

A team without capacity to follow up

The second sign is internal. A channel that produces twenty extra enquiries a month to a business that answers the phone twice a week is not a marketing win. It is a complaint waiting to happen, and those enquiries will be recorded as wasted spend when the review comes around.

Fix the follow-up first, then test the channel. If you want the question answered plainly before you commit anything, that is what my Google Ads management work covers, and where you are still deciding who should run it, consultant or agency sets out the difference between the two. How I work with businesses is the short version of that, and it is worth reading before you brief anybody, including me.

Frequently asked questions

Is Microsoft Advertising worth testing for an Australian business?

It depends on who searches for what you sell. The network is smaller than Google in Australia, so the honest test is whether your existing paid search campaigns find extra, genuinely new enquiries there, rather than whether the channel looks impressive in a report. A bounded test with a clear review date answers that question cheaply.

Can we just import our Google Ads campaigns?

You can, and it is the fastest way to start, but an import copies settings that do not always translate well. Treat the imported campaign as a first draft: check the network settings, the location targeting and the audiences, then prune before you judge the results.

How much should we spend on a first test?

That is a business decision rather than a formula, but the principle I use is to fund enough clicks for a decision and no more. If the channel needs a handful of enquiries before you can judge enquiry quality, size the test around that, and set a date when you will review it rather than letting it run untested for a year.

Does the smaller audience on the network make it harder to optimise?

It makes patience necessary. With fewer clicks, small fluctuations look like trends, so judge the channel on enquiry quality over a longer window rather than on week to week cost per click.

Which Australian businesses tend to do well on it?

In my experience it suits businesses selling to professional and older audiences, and businesses whose buyers are at work when they search. It is also common in B2B, where the same person may use a work profile in one ecosystem and a personal one in the other.

What should we check before scaling spend?

That the enquiries are new rather than duplicated from other channels, that they are followed up, and that the cost per quality enquiry is defensible against what the business already pays elsewhere. Reach is a claim. Followed up enquiries are evidence.