The Contrarian's 2027 PPC Planning Checklist

A 2027 Google Ads planning checklist for PPC should start with what Google has already decided for 2027, as those changes are already in your Google Ads account and a forecast is only a guess. A Budget limited campaign (one held back by its budget) that was delivering better results than its set Target CPA (target cost per acquisition) has been delivering closer to that typed target since August 17, 2026. Almost everything that ranks for a Google Ads planning checklist for 2027 is undated and is merely an ongoing list of things to optimize, and the paid search CTR statistic that gets used on planning slides came from informational searches only. Every point below is dated or sourced and culminates in a look at your account. The one from August is already spending money.

#	Check	When it lands	The account pull
1	DSA and call ads retire	February 2027	Manual AI Max upgrade in January, with a rollback date
2	Budget limited Target CPA	Since August 17, 2026	Target against trailing 30 day actual
3	Conversion delay and upload route	Since June 15, 2026	Days to value, and the API your CRM upload calls
4	Platform against CRM revenue	Monthly	Platform over CRM revenue, by campaign
5	Hidden search terms	Ongoing	N gram roll up of the export
6	Zero conversion terms	Ongoing	Sort by cost, work the top tenth
7	The 19.7 to 6.3 percent CTR figure	Before the deck ships	Cite Seer's April 2026 study with its scope
8	Incrementality test	Before budget moves	One test on the largest line
9	ChatGPT ads	2027 test budget	Capped line, criteria written first

Which Google changes already have a 2027 date?

The two retirements of functionality, Dynamic Search Ads and call ads, both end in February 2027. But the changes to bidding and upload functionality that sit behind these retirements are already here, so they can and should be incorporated into the 2027 plan before any growth targets are set.

1. Dynamic Search Ads and call ads both end in February

The automatic migration of Dynamic Search Ads (DSA) into AI Max for Search campaigns was previously set for September 2026 but has been pushed back to February 2027, a delay that Ginny Marvin, Google's Ads Liaison, announced in June 2026 citing advertiser feedback and the busy Q4 season (Search Engine Journal, June 12, 2026). For call ads, the option to create new ones was removed in February 2026, and existing call ads stop serving in February 2027, replaced by responsive search ads with a call asset. The upgrades of automatically created assets and campaign level broad match to AI Max for Search campaigns went ahead on schedule in September 2026, and AI Max is now on by default for new Search campaigns.

It's great that Google has moved the date out of Q4 and is encouraging voluntary upgrade of all campaigns before February 2027. Run these upgrade tools voluntarily during a booked week in January with a rollback date set beside it, then test each newly migrated campaign's search terms against December's and let those queries decide the rollback. It is also worth folding into that same review what the September upgrade did to match types and broad match.

2. A budget limited Target CPA now delivers closer to the target you typed

The way Google handles "Limited by budget" campaigns (labeled as such in the Status column) under target based bidding such as Target CPA has changed as of August 17, 2026. A concrete example (direct from Google): If your Target CPA is $10 and your recent actual is $5, Google will now deliver closer to that $10 CPA than before if you change nothing. A Bid Target Adjustment Tool was made available for Google Ads back on July 6, 2026 in order to decrease a bid target to match recent performance (for a Target CPA campaign).

Then open the campaigns limited by budget and list the target CPA against the last 30 days' actual CPA. For a large gap here, sort the search terms by cost since August 17 to see what you got spending extra money here; a higher CPA is a fair trade for queries you would have bid on by hand anyway, and the Bid Target Adjustment Tool lowers the target CPA on the rest. See the full update on this change for August 2026.

3. Does your conversion value reach the bidder in time?

Smart Bidding looks at the primary conversion action(s) listed in the Conversions column (plus any secondary actions listed in custom goals for Smart Bidding). That's what it will use to optimize to by default. Google's value based bidding guidance recommends that primary conversion actions have delays of under 7 days, notes that the upward ramp may take several months when all conversions have a 7+ day delay, and accepts a longer average delay as long as a share of conversions arrives within the week. A qualified lead score that arrives in days (vs. months) would thus be an earlier conversion action for a long sales cycle.

Since June 15, 2026, enhanced conversions for leads and new offline conversion imports have had to be uploaded via the Data Manager API, a separate Google interface for sending conversion data, instead of the Google Ads API. As announced in Google's developer notice and reported by PPC Land, from now on developer tokens that had not already adopted offline imports in the Google Ads API will get the error message CUSTOMER_NOT_ALLOWLISTED_FOR_THIS_FEATURE. However, allowlisted tokens will continue to work while they are being migrated to the Data Manager API. Check which API your CRM upload script or connector calls; we explain why it is worth sending the CRM conversion value to Smart Bidding in a separate article.

Which reports answer a narrower question than the one you asked?

The conversion columns on the right hand side of the report, and the search terms report below them, each count less than the label on the column or row suggests. An export of these reports pulled this month will therefore measure how large that gap is.

4. ROAS and CRM revenue will not match, so track the gap

Note that by default Google Ads records a conversion for the date of the click, while the Conversions (by conv. time) column records it on the day the conversion happened. However, your CRM will most likely record revenue on the date of the close of a deal (March in this example for a close from a click in January). So two systems that are tracking exactly the same activity will report in two different months without either one making a mistake, and once modeled conversions and two ad platforms claiming one sale are added, the totals diverge by design.

Simply put the platform reported revenue divided by the CRM reported closed revenue for each campaign in the monthly report. A number that stays constant and has a known "haircut" to it is a fine measurement instrument. A number that changes is an alarm and you need to determine if the change is due to a change in tracking or a change in mix. The ratio gets its full treatment in our separate ROAS vs revenue optimization piece.

5. How much of your spend can you still read by query?

Term level reporting within Google Ads changed in September 2020 in that the report now is limited to terms that a significant number of people searched for, and Seer Interactive saw the share of spend attached to a visible term across its own client set fall from 98.7 percent to 71.0 percent that week. Later in July 2025 Search Engine Land covered an analysis by Taikun Digital's Collin Slattery across about $20 million in spend (933 campaigns), which found that hidden searches (which are terms for which you cannot see the term in the report) ran a CPC 52% above that of visible terms and a CTR 44% below. Collin notes that the vast majority of this spread for hidden terms is due to hidden brand terms. Google has stated that the threshold exists to protect users' privacy alone, and the plan does not need to wait on that argument.

Export your data and see a list of individual words (ngrams) that make up all of the search terms that you can see in your data. A single word that appears in thousands of individual click rows would appear as a single row in this aggregated data, whereas before it would have appeared as thousands of individual rows, one for each click. More on long tail keywords here.

6. Nearly half of zero conversion spend sits in a tenth of the terms

The study of 32 accounts, including 5.2 million paid search terms that did not result in any conversions by Igor Ivitskiy of Doctor Ads revealed that the most expensive 10% of terms accounted for a median of 48.8% of total spend for terms that resulted in zero conversions. The top 5% of terms accounted for 34.6% of that same zero conversion spend, and a zero is only a proxy for waste, since some of those terms assisted a sale or converted after the window closed.

Filter the export of the zero conversion rows by cost descending. Divide the cost of the top tenth of terms by total zero conversion spend; a result near 48.8% puts the review hours at the head of that list. Reviews at the top of this list would then become a candidate for waste recovery prior to adding any additional channels.

Which headline statistics should leave your 2027 deck?

Take out the paid CTR collapse that a separate, larger 2026 study by Seer Interactive reads differently, and also the often repeated myth that you need to spend six figures in order to perform incrementality testing.

7. Seer's 19.7 to 6.3 percent paid CTR drop measured informational queries only

Seer Interactive's November 4, 2025 study found paid CTR on queries showing an AI Overview dropped from 19.70% in June 2024 to 6.34% in September 2025, but it covered just 3,119 informational queries and about 1.1 million paid impressions. A separate and larger Seer study, published April 24, 2026 across 53 accounts and 296.9 million paid impressions, tells a different story: paid CTR with an AI Overview went up from 14.64% in January 2025 to 16.21% in February 2026, and paid CTR without one went down from 25.98% to 21.85%. The Pew Research Center numbers that often get attached to it, 8% versus 15%, measured organic clicks on a traditional result with and without an AI summary among 900 US adults in March 2025.

As there is no AI Overview segment in Google Ads reports, splitting your own paid CTR by AI Overview presence requires SERP tracking data from outside the account; without that data, cite the April 2026 figures along with their scope, as our audit of AI Overviews and paid search CTR does.

8. An incrementality test can start at $5,000 now

An incrementality test (also referred to as an "experiment") is used to estimate how many conversions your ads caused by comparing users who are able to see your ads to a group held back from seeing them. An incrementality test recently came within reach of far smaller budgets: Google announced on November 11, 2025 that the minimum amount required to run one fell from upwards of $100,000 to $5,000, a figure that is Google's own, and for Search and Performance Max campaigns the Conversion Lift test is set up through your account representative. Run an incrementality test on your biggest budget item first before you send that money to another channel.

Where do ads in AI answers belong in a 2027 budget?

ChatGPT ads should be on a capped test line with specified success criteria before launch; Google AI answer ads should remain within existing campaigns that already serve up AI answers.

9. Cap the ChatGPT test and write its success criteria first

Importantly, Google's help documentation says ads appearing in Google's AI Overviews cannot be targeted directly, cannot be opted out of and get no segmented reporting, and eligible ads are sourced from existing Search, Shopping and Performance Max campaigns, thus no separate budget line is required for that spend. Furthermore, Google Marketing Live on May 20, 2026 provided an update on 2 new formats currently in testing, Conversational Discovery ads and Highlighted Answers. The only statistic provided to support the value of these features came from a Google commissioned Ipsos survey, conducted in December 2025 among 13,189 online shoppers in 25 countries who already use AI Mode or AI Overviews for shopping, and found that 75% reported faster, more confident decisions in AI Mode, which says nothing about advertiser outcomes. For a 2027 plan, those Google formats stay a watch item, while ChatGPT ads, which began in the US on February 9, 2026 on the Free and Go plans with a self serve Ads Manager opening to US businesses in May 2026, must be tested as a separate line.

Six months in, Search Engine Journal reported OpenAI saying it has no performance benchmarks across advertisers, industries or campaign types, with CPCs from $3 to $22, and one B2B test in that report used visitor deanonymization to trace 336 paid clicks on ChatGPT ads to 146 organizations, of which 5 were ideal customer profiles for that advertiser. It's worth noting that the Ads Manager for ChatGPT ads does not offer any way to find out this information for yourself within the platform, so you'd have to use a third party tool to do this kind of analysis.

Write down the fixed monthly ceiling amount for the ChatGPT test, and write down the acceptable average CPC and the number of accounts that meet your 'qualified' criteria that you'd need to see in order to extend for a second month. The Google side gets its own bar in our piece on Google AI Mode ads performance.

What order should the checks run in?

We can move now to the order of the checks: start with item 2, since the August change is already moving spend, then items 1 and 3 because they carry dates, and spread the report and test checks across the first quarter. Most of the checks above, and most of a sound PPC strategy for 2027, come down to this single, most important point at different scales: read the account at the query level before you trust an average from anywhere else, Google's included, because the search terms report export is still where a plan shows what the money bought, and that is the work behind query level paid search management. Finally, it takes no effort to 'retire' a statistic that has ceased to be useful (Item #7), so we can simply treat it as if it no longer appeared in the plan.