Importing Google Ads campaigns into Microsoft Ads: copy as-is or rebuild?

Microsoft's import tool can bring over a whole Google account in a few clicks, and the temptation is to take everything. The problem is that some settings don't mean the same thing on Bing. A tCPA learned on Google's conversion volume gets imported as-is, but Bing's audience pool is smaller and skews older and more desktop-heavy, so a target that worked on Google can starve delivery there.

Keywords, match types and negative lists usually transfer fine. Bid strategies and audience layers are where I'd rebuild by hand. A good rule of thumb is to import keywords and ad copy, then set bidding fresh with Maximize Clicks or manual CPC and let Microsoft's own data build up before moving to a target. Has anyone seen an imported tCPA hold up without a reset?

The tCPA could be failing earlier than the bid target itself, on the conversion side. Imported goals can keep Google's count settings while UET fires on different events, so the target ends up training on a different signal than the one you think it's optimizing for. The sanity check is to compare Bing's conversion count against your own backend numbers.

I'd push back on resetting to Maximize Clicks. It feels safe because it drops the target, but it also drops the only signal telling Microsoft which clicks are worth paying for. Bing's volume is already thinner, so you'd be spending on clicks with no conversion feedback, and what you learn in that phase is mostly about traffic, not what converts.

If Jun's check shows the conversions are firing correctly, I'd keep a conversion-based strategy. Maximize Conversions with a daily budget cap gives the algorithm a signal and limits the damage. Or keep the tCPA but loosen it to around 1.3 times your historical Google CPA for the first few weeks, then tighten it once the conversion count is stable.

I'd skip the 1.3x rule. It ignores margin, and margin is what should set the target for e-commerce. Start from break-even ROAS per product group, then use Bing's own data to see whether that target is too tight.

Check how many campaigns you're bringing over before you pick a bid strategy. Google accounts often split one offer into lots of small campaigns by device, location or ad group theme, and each one runs its own bidder. Bing volume per campaign is thinner, so a tCPA campaign with five conversions a month has almost nothing to learn from. The usual rule of thumb for target-based bidding is around 30 conversions a month per campaign, and small splits rarely get there on Bing.

So I'd merge campaigns that share a goal and a similar CPA range into fewer, larger ones. Keep the keyword and negative lists as they are, and only split out where the economics really differ, like a product group with a much different margin. Sasha's point about keeping a conversion signal still holds, but it works better when the conversions are pooled.

Rowan's right that five conversions a month can't train a bidder, but merging by goal and CPA range alone is a mistake. Google accounts often split by theme because the ads say different things. Put a brand campaign and a generic "best X" campaign under one tCPA and the bidder learns an average across two very different intents, while the copy can only speak to one of them.

So I'd check whether the groups share a message and a landing page before merging. If they do, pool them. If they don't, pool by message first and accept the thinner conversion count.

Mira's right about message. Also check the Google-side conversion actions before you merge. Google's import can carry a secondary "view" or micro-conversion over as a primary goal, so the pooled campaigns end up bidding on different events.

I wouldn't treat 30 conversions a month as a hard line. I think Microsoft's own guidance for tCPA is closer to 15 in the last 30 days, though check the current docs because I might be misremembering. What matters more is whether the count is steady. A campaign with 18 conversions spread evenly across the month can train a bidder fine, while 40 conversions that land in one burst after a promo teach it very little. Week-to-week variance tells you more than the monthly total, and it keeps you from merging unrelated campaigns just to hit a round number.

I wouldn't split campaigns by message just for bidding. Message is a copy and landing page decision, and separate ad groups can carry their own ads and URLs inside one campaign. The bidder only sees the campaign goal and the conversions coming in, so it doesn't care what the ads say.

The one split that does make sense is brand versus generic, and it's about cost. Brand clicks convert far cheaper, so if they sit in the same campaign they pull the average down and make the generic traffic look artificially expensive. I'd keep brand on its own bidder, pool the generic themes together, and handle message at the ad group level.

Agreed on brand, and that's where the import is most likely to hurt on Bing. Brand searches are a much smaller slice of Bing volume, so a brand campaign can sit at five or ten conversions a month even when it does well on Google. A tCPA has almost nothing to learn from at that level. I'd keep brand on Manual CPC or Maximize Clicks with a sensible cap, and let the generic campaigns carry the target-based bidding, since that's where the volume actually is.