Microsoft Ads import from Google: keep the structure or rebuild the bidding too?

The Google import in Microsoft Ads is the obvious shortcut, and it saves real hours on structure. The mistake is treating it as finished. Bing volume is usually a fraction of Google's, so a Smart Bidding setup tuned on Google data can behave oddly once it's sitting on a much smaller pool. Audiences are the other gap: Customer Match lists and remarketing segments don't map one-to-one, so what came across may not be what you think it is. I'd keep the keywords and ad copy as imported, since the intent carries over fine, and rebuild the bidding and audience layer by hand. Has anyone run a big import and seen the bid strategy hold up on the smaller volume?

Depends on conversion volume, and I'd sort that out before touching the bid strategy. A Target CPA imported from Google was tuned on Google's data, so if the Bing campaign only gets a handful of conversions a month, the algorithm has almost nothing to learn from and can swing hard on a few clicks. In that case, run Maximize Clicks or manual CPC on a tight budget for a few weeks to collect real Bing conversions, then move to a target.

If volume is decent, keeping the imported target and watching the first couple of weeks is usually fine. Theo's point about audiences applies either way.

The imported target is often the bigger problem than the volume split. A Target CPA copied from Google is a Google number, and Bing usually has different CPCs and conversion rates for the same keywords, so it can be too tight even when volume looks fine. Rule of thumb: if you're under roughly 30 conversions in the last 30 days, skip the target entirely. Above that, run Maximize Conversions with no target for a couple of weeks, then set the target from the CPA you actually see on Bing. The algorithm chases whatever number you give it, so a Google-based target mostly buys you fewer conversions at a higher cost.

Agree on the target, but check the conversion goal before you count to 30. An imported goal can fire on every form submit, or on a different event than you expect, so the volume looks healthy while the CPA is off. Are you using the imported goal or a fresh UET one on Bing?

I'm not sure the 30-conversion cutoff holds up as a rule. It's a rough number, and what matters is how noisy your CPA is at that volume. Thirty conversions at a $40 CPA tells you a lot less than thirty at $4. On an imported goal, check whether it's primary or secondary and whether it's set to count in the conversions column, because that decides what the bidder actually optimizes against. The count can look healthy while the wrong event is driving it.

Jun's right that 30 is a rough line, but I don't think goal setup is a separate check from volume. On Bing, an imported goal that's set as secondary, or not counting toward conversions, still gets reported, so the count looks fine while the bidder ignores it. Look at the goal settings first, then count.

Agreed on goal settings coming first. One thing nobody's mentioned: Microsoft's UET tag can record conversions from the imported goal and from a new one at the same time, so make sure the imported goal isn't being counted twice against a fresh UET goal before you trust the volume.

The double count is easy to miss because both goals look fine on their own. Check whether the imported goal still fires off the old Google event on the same thank-you page the UET tag watches. If it does, one real lead can land twice, and the total will look better than what's in your CRM. Pull the leads for the same week from the CRM and compare them to the Bing count before you set any target.

Mira, I'd turn that into one rule: only one goal counts toward bidding for each real action, and everything else stays secondary or gets switched off. If the imported Google goal and the UET goal both fire on the same lead, your reported CPA looks better than it really is, and any target you set off that number will be too loose.

One counting goal per action also makes the CRM comparison simple. You're checking a single number against the leads instead of trying to work out which of two counts to trust. Once that matches for a couple of weeks, you can decide whether the imported goal is worth keeping as a secondary signal at all.