{"id":29148,"date":"2026-08-05T11:11:53","date_gmt":"2026-08-05T11:11:53","guid":{"rendered":"https:\/\/mazhu.localshop.lk\/?page_id=29148"},"modified":"2026-08-05T11:17:10","modified_gmt":"2026-08-05T11:17:10","slug":"portfolio-bid-strategies-stopped-hitting-targets-in-december-2025-here-is-what-i-think-happened","status":"publish","type":"page","link":"https:\/\/mazhu.localshop.lk\/?page_id=29148","title":{"rendered":"Portfolio bid strategies stopped hitting targets in December 2025. Here is what I think happened.\u00a0"},"content":{"rendered":"\t\t
\n\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t\n\t\t\t

\u00a0In November and December of 2025, every portfolio bid strategy I manage started missing its monthly targets. Not one account. Every account. Simultaneously.\u00a0<\/p>

Prior to that point, portfolio strategies had been reliably hitting their thirty-day average targets for years. CPL targets, ROAS targets, it did not matter. The portfolio approach worked because it pooled conversion data across multiple campaigns and optimised toward a shared goal. It was one of the most stable and effective tools in the account management toolkit.\u00a0<\/p>

Then, within the space of about six weeks, that stopped being true. The strategies started missing by fifteen percent in some accounts and thirty percent in others. No single campaign had obviously broken. No tracking had dropped off. No conversion events had changed. The accounts looked fine. The targets were just not being hit.\u00a0<\/p>

I am still watching this across my accounts. My conclusion is that Google shifted its primary optimisation mechanism away from portfolio-level bidding toward campaign-level bidding. I cannot prove this. Google did not announce it. But the pattern across accounts is too consistent to be coincidence, and the timing aligns with everything else Google has been building toward.\u00a0<\/p>

Why simultaneous across all accounts matters\u00a0<\/span><\/p>

When something goes wrong in one account, the cause is usually in that account. A broken tag. A change to the website. A budget exhaustion problem. Something specific to that client or that category.\u00a0<\/p>

When the same thing goes wrong across every account simultaneously, across different industries, different budget levels, and different conversion types, the cause is not in the accounts. It is in the platform.\u00a0<\/p>

I have been running daily reports across client accounts for long enough to recognise the difference between account-level variance and platform-level shifts. The December 2025 pattern was a platform shift. The only question was what had changed and why.\u00a0<\/p>

What portfolio bidding was\u00a0actually doing\u00a0<\/span><\/p>

To understand why this matters, it helps to understand what portfolio bidding was designed to do and why it worked.\u00a0<\/p>

A portfolio bid strategy pools multiple campaigns into a single optimisation unit. You set one target, whether that is a CPL, a ROAS, or a CPA, and the algorithm distributes budget and adjusts bids across all the campaigns in the portfolio to hit that single target. A campaign that is having a strong day gets more budget. A campaign that is underperforming gets pulled back. The portfolio self-balances.\u00a0<\/p>

The practical benefit was that you could have campaigns with\u00a0relatively low\u00a0individual conversion volume and still give the algorithm enough aggregate data to learn from. A campaign with fifteen conversions a month sitting in a portfolio with four other campaigns might collectively have eighty or ninety conversions a month. That aggregate volume was what the algorithm used to optimise. Individual campaign data volume mattered less.\u00a0<\/p>

This approach worked extremely well for a long time. It was stable, predictable, and genuinely effective for accounts that could not generate the conversion volume required for campaign-level bidding to work independently.<\/p>

What I think changed\u00a0<\/span><\/p>

My conclusion is that Google moved the primary optimisation mechanism from portfolio level to campaign level. Portfolio strategies still exist in the interface. They still function in a technical sense. But the weight the algorithm places on portfolio-level signals appears to have reduced significantly\u00a0relative\u00a0to campaign-level signals.\u00a0<\/p>

The practical effect is that campaigns sitting inside a portfolio but generating low individual conversion volume are no longer being carried by the aggregate pool the way they once were. Each campaign is now being evaluated more on its own data. Campaigns without sufficient individual conversion volume start underperforming, and the portfolio\u00a0as a whole misses\u00a0its target because several of its components are individually starved of signal.\u00a0<\/p>

This aligns exactly with what Google has been building toward for the past two years. Google Performance Max is a campaign-level tool. Google AI Max is a campaign-level tool. The entire direction of the platform is toward fewer, larger, better-fed campaigns rather than many small, tightly themed ones pooled together. Portfolio bidding was a workaround for a data volume problem. Google\u00a0appears to be\u00a0solving that problem differently now, by pushing consolidation rather than pooling.\u00a0<\/p>

Why Google did not announce this\u00a0<\/span><\/p>

I do not know with certainty why this change was not communicated. What I can say is that it is consistent with how Google has handled similar transitions in the past.\u00a0<\/p>

Platform changes that require advertisers to restructure their accounts are rarely announced as explicit changes. They tend to appear as gradual shifts in performance that create pressure to adopt newer tools and structures. The pressure to\u00a0consolidate, to move to Google Performance Max, to adopt Google AI Max, all of it creates a natural migration away from older approaches without Google needing to formally deprecate them.\u00a0<\/p>

Portfolio bidding was not turned off. It was quietly made less effective. The accounts that noticed were the ones with daily monitoring. The accounts that did not notice were the ones reviewing performance monthly, by which point the signal was buried in averaging and the cause was much harder to\u00a0identify.\u00a0<\/p>

What this means for how accounts need to be structured now\u00a0<\/span><\/p>

The account architecture that worked well under portfolio bidding was built around tightly themed, granular ad groups and campaigns. Small, focused, precise. The portfolio absorbed the data volume problem and the algorithm optimised across the whole.\u00a0<\/p>

That structure now works against you if each campaign does not have sufficient individual conversion volume. The same precision that made the account clean and logical becomes a liability when the algorithm needs thirty or more conversion events per campaign per month to function effectively.\u00a0<\/p>

The direction to move in is consolidation. Fewer campaigns. More conversion data per campaign. Ad groups with broader thematic coverage rather than highly granular sub-themes. This is a meaningful structural change for any account built on the old model, and it\u00a0has to\u00a0be managed carefully.\u00a0<\/p>

How to manage the transition without breaking what is working\u00a0<\/span><\/p>

This is where I have had to be most deliberate. The temptation when you understand what needs to change is to make all the changes at once. That is\u00a0almost always\u00a0the wrong move, and it is particularly wrong here.\u00a0<\/p>

Consolidating\u00a0campaigns means the algorithm needs to relearn. Every campaign that gets merged into another one goes through a learning phase as the new combined structure settles. If you\u00a0consolidate\u00a0everything simultaneously, you have an entire account in learning at the same time. For a large account with high daily conversion volume, that learning period is short. For a small account, it can run for\u00a0weeks\u00a0and the performance impact is material.\u00a0<\/p>

What I do instead:\u00a0<\/span><\/p>

I start by\u00a0identifying\u00a0which campaigns are generating the most conversions. These are the campaigns I\u00a0consolidate\u00a0into, not the campaigns I merge away.\u00a0<\/p>

I look at which campaigns have fewer than thirty conversion events per month against their target. These are the candidates for consolidation. I do not start with the biggest or most complex ones. I start with the smallest and least active, where the risk of disruption is lowest.\u00a0<\/p>

I\u00a0consolidate\u00a0one campaign at a time and wait for the receiving campaign to restabilise before touching the next one. This takes longer. It is the right approach. An account that\u00a0consolidates\u00a0slowly over six weeks without any performance disruption is significantly better than an account that\u00a0consolidates\u00a0in a week and then spends a month recovering.\u00a0<\/p>

Once the structure is\u00a0consolidated\u00a0and stable, I reassess the bid strategy. At that\u00a0point\u00a0the question is whether the remaining campaigns have enough individual conversion volume to perform well on campaign-level bidding, or whether a portfolio is still doing useful work. In some accounts the portfolio\u00a0remains\u00a0useful. In others, campaign-level targets perform better with the\u00a0consolidated\u00a0structure. The only way to know is to test it.\u00a0<\/p>

The broader shift this is part of\u00a0<\/span><\/p>

The December 2025 portfolio strategy change did not happen in isolation. Google Smart Bidding started the move toward giving the algorithm more data per unit of optimisation. Google Performance Max accelerated it. Google AI Max is the current expression of it. The portfolio shift is another step in the same direction.\u00a0<\/p>

Whether this is better for advertisers is a separate question from whether it is happening. It is happening. The accounts that adapt will perform. The accounts that do not will find their portfolio strategies quietly underdelivering with no obvious explanation.\u00a0<\/p>

Consolidating\u00a0campaigns and getting the structure right is necessary work. But structure alone does not fix the underlying problem if the account is still\u00a0optimising toward the wrong thing. The signal feeding the algorithm matters as much as the shape of the account.\u00a0<\/p>\n\t\t\t\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"

\u00a0In November and December of 2025, every portfolio bid strategy I manage started missing its monthly targets. Not one account.<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":29026,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"elementor_theme","meta":{"footnotes":""},"class_list":["post-29148","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=\/wp\/v2\/pages\/29148","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=29148"}],"version-history":[{"count":4,"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=\/wp\/v2\/pages\/29148\/revisions"}],"predecessor-version":[{"id":29152,"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=\/wp\/v2\/pages\/29148\/revisions\/29152"}],"up":[{"embeddable":true,"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=\/wp\/v2\/pages\/29026"}],"wp:attachment":[{"href":"https:\/\/mazhu.localshop.lk\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=29148"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}