Microsoft Ads removes Max CPC on 1 October. Portfolio bid strategies keep the cap

Microsoft Advertising emailed advertisers on 20 August 2026 to say that from 1 October, the Max CPC ceiling disappears from new campaigns built on automated bidding. It is a small setting and an easy email to skim past. It is also the last hard cost ceiling most Microsoft accounts have, and there are three weeks left to decide whether you need it.

Exactly what changes on 1 October

The change is narrower than the headlines suggest, and the boundaries are where the decision lives.

Microsoft's stated reasoning is that advertisers using conversion-based bidding with targets hit their goals more reliably than those relying on legacy controls, because a Max CPC overrides the stated goal and can cause spend pacing irregularities. That is a fair description of what a bid cap does. A cap that binds is, by definition, refusing auctions the bidder's model wanted to enter, and the model then has less data and a distorted view of the auctions it did win.

One open question: at the time of writing, Microsoft has not published how the Max CPC field will behave for campaign creation through the API, Microsoft Advertising Editor or bulk upload after 1 October. The interface change is confirmed; the other creation paths are not documented. Do not build a workaround on the assumption that bulk upload will keep accepting the field.

First, work out what your Max CPC was actually doing

Most accounts have a Max CPC set because somebody set one years ago, not because anyone has recently checked whether it binds. Before deciding whether to preserve the capability, pull the last 90 days and answer one question per campaign: is the cap binding? If your average CPC sits well below the cap and your impression share lost to rank is not moving, the cap is decorative and losing it costs you nothing.

Where the cap does bind, it is usually doing one of four jobs, and only three of them are worth preserving.

That fourth row is the common one. A cap on an untargeted maximise strategy is a budget-shaped constraint pretending to be an efficiency constraint: it controls the price per click, which is not the number you are judged on. If that is your setup, the honest fix is a Target CPA or Target ROAS derived from your actual unit economics, not a workaround to keep the ceiling.

The portfolio bid strategy route, and what it costs you

If a cap is genuinely load-bearing, portfolio bid strategies are the supported way to keep one, for new campaigns as well as existing. A portfolio strategy is a shared, account-level strategy that several campaigns can join, optimising across them collectively rather than one campaign at a time.

That collective optimisation is the trade. Weigh it honestly:

• You gain a bid ceiling that survives 1 October, plus genuine cross-campaign optimisation, which usually helps when campaigns compete for overlapping queries.
• You lose some per-campaign independence. Campaigns in one portfolio share a target and are optimised as a set, so a campaign with materially different economics does not belong in the same portfolio as its neighbours.
• You add a reporting layer. Diagnosing a single campaign's performance is harder when its bids are set by a strategy shared with five others, and that cost is paid every month, not once.

A workable rule: group into a portfolio only campaigns that share a target and a purpose — all non-brand prospecting at the same CPA target, say, or one product category with one ROAS target. Do not create a single account-wide portfolio just to keep a cap. You would be trading a real diagnostic capability for a setting you may not need.

Should you pre-create campaigns before 1 October?

Campaigns created before 1 October keep Max CPC, so there is an obvious hedge: spin up the campaigns you expect to need in Q4 now, while the setting is still available. That is legitimate for work you were going to do anyway. It is a bad idea as a general strategy, for three reasons.

1. Grandfathering has no stated end date, which is not the same as having none. Microsoft has said existing campaigns keep the setting. It has not said for how long, and a control removed from creation is usually removed from editing later.
2. Empty shell campaigns are not free. A campaign created now and launched in November starts learning in November regardless, and a paused shell accumulates no history that helps it.
3. You would be optimising to keep a control that Microsoft's own data says hurts goal attainment. If the cap is not binding, preserving it is effort spent on nothing.

The defensible version: if your Q4 structure is already agreed and you were going to build those campaigns in the next three weeks, build them now. If you are inventing campaigns to bank a setting, stop.

A pre-1-October checklist

1. Export the last 90 days by campaign with average CPC, the current Max CPC value, impression share lost to rank, and conversions. One sheet, all campaigns.
2. Flag every campaign where average CPC is within roughly 15% of the cap. Those are the binding ones. Everything else is decorative and needs no action.
3. For each binding campaign, name the job from the table above. If it is the fourth job, set a real target instead.
4. Move genuine brand-defence campaigns to Target Impression Share where holding a position is the actual goal. That strategy is unaffected by this change.
5. Build portfolio strategies only where campaigns share a target and a purpose, and document which campaigns sit in which portfolio before you migrate.
6. Set Q4 targets from unit economics, not from last quarter's average CPA. If you do not know your contribution margin per order, that is the more urgent problem.
7. Write down the current Max CPC values before you change anything. If a migration goes badly you will want the previous state, and it is not recoverable from the interface once the setting is gone.

This is a pattern, not an incident

Microsoft is following the same path Google has been on all year. Google is auto-upgrading Search campaigns using automatically created assets and campaign-level broad match to AI Max in September 2026, with Dynamic Search Ads on a later timeline, and Microsoft rolled out its own AI Max for Search worldwide in August. In every case the direction is the same: manual constraints come out, targets and signals go in, and the advertiser's leverage moves from settings to inputs.

That shift is not automatically bad, but it changes where the work is. When you cannot cap the price, the things you can still control are the conversion data you feed the model, the targets you set, the negatives and exclusions that define the auctions you enter, and the landing page that decides what a click is worth. Those are the levers that survive the next removal too, which is a reason to invest in them rather than in the next workaround. If your CPCs are rising faster than your revenue, the diagnosis usually sits in that list rather than in the bidding settings — something we have written about in more detail in scaling ad spend without killing CAC and in our performance marketing work generally.

Frequently asked questions

When does Microsoft Advertising remove Max CPC?

From 1 October 2026. Microsoft emailed advertisers on 20 August 2026 to say that Max CPC will no longer be available when creating new non-portfolio campaigns using Maximize Clicks, Maximize Conversions or Maximize Conversion Value, including those with a Target CPA or Target ROAS.

Do existing Microsoft Ads campaigns lose their Max CPC?

No. Campaigns created before 1 October 2026 that already use Max CPC keep the setting. Microsoft has not stated how long that will remain true, so treat it as grandfathering rather than a permanent guarantee.

How can I still set a bid cap in Microsoft Ads after 1 October 2026?

Use a portfolio bid strategy. Max CPC remains available for both new and existing campaigns that use one. The trade-off is that a portfolio strategy optimises across all campaigns that join it, so only group campaigns that share a target and a purpose, and expect single-campaign diagnosis to get harder.

Which Microsoft Ads bid strategies are not affected?

Target Impression Share and Enhanced CPC are unaffected by this change. Target Impression Share is usually the better tool anyway when the real goal is defending a position on brand or competitor terms.

Why is Microsoft removing Max CPC?

Microsoft's stated reasoning is that advertisers using conversion-based bidding with targets meet their goals more easily than those relying on legacy controls, because a Max CPC overrides the stated goal and can cause spend pacing irregularities. A binding cap refuses auctions the bidding model wanted to enter, which both limits delivery and distorts the data the model learns from.

Should I create campaigns before 1 October to keep Max CPC?

Only if you were going to build those campaigns anyway in the next few weeks. Pre-creating shell campaigns to bank the setting is weak: grandfathering has no stated end date, a paused campaign accumulates no useful learning history, and in most accounts the cap is not binding in the first place.

How do I know whether my Max CPC is actually doing anything?

Pull the last 90 days by campaign and compare average CPC against the Max CPC value alongside impression share lost to rank. If average CPC sits well below the cap and impression share lost to rank is stable, the cap is decorative and losing it costs nothing. Flag campaigns where average CPC is within roughly 15% of the cap; those are the ones that need a decision.