Your Microsoft Ads account may not have a Bing problem. It may have a Google-settings problem: someone copied a working campaign into a different auction and assumed the same settings would buy the same result.

Say you spend $20k a month on Google. Someone clicks Import, the campaigns appear in Microsoft Ads that afternoon, spend starts, and conversions do not. I used to tell clients that was a volume problem. I was wrong to start there. Import copies campaign decisions; it cannot copy the conditions that made those decisions work.

That is the fact everything else follows from. Keywords meet different searches. Network settings can expose ads to different inventory. Audiences and devices change the value of a bid, while tracking determines whether you can see what happened. If you manage the imported account as a smaller version of Google, you keep adjusting numbers inside a model that was never checked. I start by checking the model.

First, find out what Import carried over

Microsoft built Import to move campaigns quickly, not to decide whether their settings still make sense. Campaigns, ad groups, keywords, ads, bids, budgets, locations and schedules can transfer in one pass. That convenience is real. So is the catch: Microsoft’s import documentation lists settings and audiences that do not come across, including customer lists, affinity audiences, detailed demographics and YouTube visitor lists built in Google.

In our running example, a plumber bids $18 on phrase-match emergency plumber chicago. Google has a tight negative list and a history behind that bid. Import can carry the keyword and bid. It cannot carry the Google auction, and it may not carry every audience the campaign relied on. The number survives; the reasons for the number may not.

Google campaign settings being copied into a different search auction

Import is not necessarily a one-time event, either. Quick import applies defaults, while Customize lets you review settings before the pull; it can also pull from multiple Google accounts. A scheduled import can continue bringing selected Google changes across later. Before you fix a bid or pause a campaign in Microsoft, check what the next sync is allowed to change. Otherwise you can spend Tuesday correcting an account that Wednesday’s import puts back the way it was.

The first practical move is not a bid change. It is taking control of the copy.

Now follow the plumber’s $18 keyword into the auction. The phrase emergency plumber chicago looks familiar on the Microsoft keyword list. That does not mean the search term report beneath it will look like Google’s. Query inventory, searcher mix and close variants can differ. Lower volume also makes an unfamiliar query more noticeable: one expensive click can take a bigger bite out of a small campaign.

This is where the negative list matters. Microsoft offers phrase and exact negative match, but not broad negative match. Its negative matching also needs its own review for variants, including plurals and misspellings. A Google list that kept unwanted searches out cannot simply be assumed to provide identical coverage after import. The campaign may still look tightly built when you inspect its keywords. The search term report tells you what it actually bought.

I pull that report before touching bids. If the plumber is paying for searches outside the intended job, lowering $18 to $15 makes those searches cheaper; it does not make them relevant. Add the Microsoft negatives first, then judge the price of the traffic that remains. For a closer look at the distinction, see why the same keyword behaves differently on Bing.

A keyword is a bid for eligible searches, not a promise about which searches will appear. Check eligibility in the destination account.

Then check where the ads appeared

Search terms explain what someone typed. Distribution explains where the ad appeared. You need both before you can call a keyword a failure.

Our plumber had Search plus search partners selected on Google. According to Microsoft’s import documentation, that choice can map to Microsoft’s search networks, including Bing, AOL, Yahoo and syndicated partners. Search plus Display does not map to an equivalent Microsoft content network. The point is not that every partner click is bad. It is that a copied network choice can buy inventory you never evaluated in Microsoft.

Search budget leaking into placements that need a separate review

There is a second distinction to inspect: search ads may be eligible for native placements across properties such as MSN, Outlook and Edge through Search Ads Extended to Native. Native and search serve different moments. Someone typing an emergency plumbing query is declaring a need; someone encountering a native placement has not necessarily done that. A blended campaign total can conceal the difference, whether the native traffic performs well or badly.

Segment by Ad Distribution before judging the plumber’s keyword. Look at search, partners and any native delivery separately. If native deserves a test, give it its own budget and assessment rather than letting it inherit the search campaign’s success criteria. This is not a philosophical objection to another placement. It is basic accounting: know which inventory spent the money before deciding what to cut.

A transferred bid can lose its audience context

Suppose the plumber’s Google campaign used observation audiences to help interpret or adjust performance. Some Google-built audiences do not transfer through Import. The imported campaign can therefore keep a bid that was set with those layers in mind without keeping the same layers. That does not prove the bid is wrong. It means the evidence for it has changed.

Devices deserve the same treatment. Microsoft’s searchers and device mix need not mirror the plumber’s Google traffic. I used to copy device modifiers straight across because the campaign looked like the same campaign. That was the shortcut, not the analysis. If Microsoft produces more valuable desktop enquiries for this advertiser, a Google-derived mobile preference is a poor place to stop.

Rebuild relevant audience layers in Microsoft and set device bids from Microsoft performance. Start by seeing what transferred, what did not and where enquiries actually come from. Only then decide whether the $18 bid belongs in this auction. Otherwise you are pricing today’s traffic with another channel’s history.

Before calling it a failure, check what counts as a conversion

Our plumber has reviewed queries, distribution and bids. The dashboard still shows zero conversions. That could mean the traffic is not producing enquiries. It could also mean the account is not measuring them.

Microsoft’s UET tag needs to be present, use the right ID and fire for the action you intend to count. An inactive or unverified tag, or one blocked by consent, can leave a working enquiry path invisible to the account. No recorded conversion means automated bidding has less useful feedback, even if the phone rings or the form receives enquiries.

The opposite problem is counting one action twice. If an imported or separately configured goal sits alongside a UET goal for the same form submission, inspect whether both goals count that action. Do not assume Import always creates the duplicate; check the account in front of you. A duplicate can make bidding look better informed than it is and make reported CPA difficult to trust.

Dates can confuse the comparison, too. Microsoft and GA4 can attribute a conversion to different dates, so a lead that arrives after the click may appear in different reporting windows. Compare like with like before deciding one dashboard has found a sale the other missed. First establish that one real action produces one intended conversion. Then use that signal to judge the channel.

Trace the symptom back to the setting

At this point, the apparent mystery usually becomes a short chain of causes. The plumber’s $18 bid may be too high for the useful Microsoft searches it wins. Or the bid may be reasonable while irrelevant queries consume the budget. Or a distribution setting may put search creative in placements the advertiser never intended to fund. None of those problems is fixed by asking whether Bing is good or bad.

Connections between imported settings, traffic and reported conversions

An imported bid is a useful starting hypothesis, not a price floor. One suggested post-import adjustment is to lower bids 15 to 30%, then use Microsoft performance to set the ceiling. I would not make that cut before separating relevant search traffic from queries and placements I plan to exclude. A cheaper bad click is still a bad click. Once the traffic is cleaner, test the bid against the enquiries it produces.

The same cause-and-effect test works for measurement. If a tag is missing, the dashboard can say zero while enquiries arrive. If goals duplicate an action, the dashboard can suggest a healthier CPA than the real action count supports. Both problems can steer bidding in the wrong direction. The reported number is an output of the setup, not a verdict on the auction.

This is why I do not start with a blanket instruction to increase budget, rewrite every ad or abandon Microsoft. Identify which copied assumption produced the symptom, then change that assumption. A fuller account-level walkthrough appears in Microsoft Ads spending without converting? Audit the Google Import first.

Audit the import in the order that reveals the cause

Run this in order. Each check makes the next one easier to interpret: secure the account against unwanted sync changes, establish what traffic arrived, then price and measure the traffic you intend to keep. Otherwise you can tune a bid on inventory you were about to exclude.

Clipboard listing the checks for a Microsoft Ads import audit

  1. Import settings and sync schedule. Check the last import and what future syncs can overwrite. Remove status, bid or budget settings from scheduled imports when Microsoft needs independent control.
  2. Search terms and negatives. Review 30 days of Microsoft search terms, starting with spend that produced no conversions. Build phrase and exact negative coverage in Microsoft rather than assuming the Google list arrived intact.
  3. Network and placement distribution. Segment by Ad Distribution. Review search partners and any native delivery separately; exclude unwanted inventory and put a deliberate native test in its own campaign.
  4. Bids, devices and audiences. Evaluate imported bids against the traffic you kept. Rebuild useful In-market and remarketing layers, then set device adjustments from Microsoft results.
  5. UET tag and conversion goals. Confirm the correct tag fires for the intended action and that one action is counted once. Compare reporting dates before judging CPA. The duplicate-goal check is easy to skip; it can leave bidding optimising toward conversions that did not happen twice.

Use the model on the next account, not just this one

If you spend under about $3k a month on Microsoft and can give the account four honest hours, you can work through this yourself. Check the sync, run the audit, then give the cleaned-up campaign two full weeks before judging CPA. That will not make every campaign profitable. It will give you a better answer to the narrower question: was the import masking what this channel could do?

Hand it over when Microsoft spend is more than a rounding error, nobody owns the account weekly, or your agency treats it as Google with a different logo. Ask to see distribution, a Microsoft search term report and UET goals. If the report never gets past blended clicks and spend, they are managing the import’s appearance, not the account’s decisions.

Now take a new case: a dental implant clinic spends $8k on Google and $1.2k on Microsoft, with zero Microsoft consults reported. Its $42 implant keyword arrived at a Google-derived bid. Before calling that bid the problem, I would check the sync, search terms and distribution. Did the spend reach relevant searchers, or did partner or native placements take a share? Next I would check which audience layers made the trip and whether the clinic’s UET goal records one consult once. Only after those answers would I test a lower bid, perhaps starting with the proposed 20% cut.

That sequence does not promise the clinic a consult. It establishes whether its $1.2k bought the intended traffic and whether the account could recognise a consult if one arrived. That is what managing Microsoft as its own channel means: Google’s campaign can supply a starting structure, but Microsoft traffic has to earn its own bids, exclusions and measurement.

For an advertiser that does not want to own that work, groas manages Search Ads across Google, Bing and ChatGPT with ongoing work on bids, keywords and budgets, plus a dedicated manager, for a flat $1,499 a month. The important distinction is not who clicks Import. It is who checks the assumptions after the click.