By December 31, 2027, no white-label Google Ads proposal worth signing will charge separately for weekly bid management. If I am wrong, the winning proposals in early 2028 will still list it as a paid deliverable. I do not expect to see many.

I used to sell that work. Weekly bid changes, match type rebuilds, late-night negative keyword mining. It was real work. Now search term matching can choose the query, Smart Bidding sets the auction bid, and final URL expansion can choose the landing page. None of that makes a good account automatic. It changes what a fulfillment partner is worth paying to do. My four calls for 2027 are about what agencies will stop buying and what they will demand instead: pages, conversion data, and fewer hours billed for watching a machine work.

Prediction 1: Separate bid-management fees disappear by December 31, 2027

The shift did not happen in one release. Two 2026 turning points show how the job changed, and why the old line item is the first one I expect agencies to cut.

September 2026: AI Max takes on more of the matching work

Google told advertisers on April 15, 2026 that DSA, automatically created assets, and campaign-level broad match would upgrade to AI Max from September 2026, then moved DSA to February 2027 while the other two kept the September schedule. By September, AI Max had left beta after hundreds of thousands of advertisers tried it; new DSA creation was blocked, and eligible campaigns were auto-migrated by month’s end.

The date matters less than the work those settings move. Dynamic ad groups become standard ad groups with search term matching, text customization, and final URL expansion enabled. Google can choose a matching query, customize the text, and select an eligible page on the domain. A practitioner still has to decide what belongs in the account and what must stay out. But weekly bid checks are a thin thing to sell when the auction bid is already automated. The valuable review moves upstream, to the inputs that determine which traffic and outcomes the system pursues.

December 2026: PMax expands the page and image decision

With final URL expansion on, Google can replace an asset-group URL with another relevant page on the same domain based on intent; with it off, PMax sends traffic only to page-feed and asset-group URLs. Leave discovery open without exclusions, and a paid click can reach a blog post, careers page, support document, or thin location page. Then, from December 4, 2026, new PMax campaigns default image extraction on across the UI, Editor, and API, even when account-level dynamic images is off. Google can scan landing pages and serve extracted images across Search, YouTube, Display, Discover, Gmail, and Maps.

That made page selection a governance job, not an afterthought. If final URL expansion is on, PMax can reach pages beyond the supplied URL, widening the set of images eligible for extraction. I used to call exclusions hygiene. I was wrong. They can also be creative direction. The practical test for a partner is whether they know which pages and images the campaign may use, not whether they can produce a tidy account of last Tuesday’s bid review.

Now look at the invoice. White-label partners may charge flat fees of $400 to $1,200 per account per month. The agency’s all-in cost can reach 1.5 to 2 times the partner invoice after internal QA and communication, with typical resale markup at 40 to 60%. Say you resell an account at $800 and pay $500. If $300 of that partner fee buys bid reviews that change nothing, a client asking what they bought makes for an uncomfortable meeting.

By the end of 2027, I expect winning scopes to stop pricing those reviews separately and put the fee against work a client can inspect. What proves me wrong: show me ten winning white-label scopes in Q1 2028 that still charge separately for weekly bid management. If agencies keep buying that deliverable, this call fails.

Prediction 2: Clients check the landing pages, not the bid log

Google’s pitch for full AI Max is an average 7% more conversions or conversion value at similar CPA or ROAS versus search term matching alone. The catch is familiar to anyone who has cleaned up a lead account: the result depends on what the account calls a conversion, so junk leads can inflate the apparent lift. Pair a weak page with loose counting and the system can get better at buying an outcome nobody wanted. Faster learning is not much comfort when it learns the wrong lesson.

The page matters before the click becomes a lead, too. When search term matching widens the queries an ad can answer and final URL expansion widens the pages it can send people to, every eligible page deserves a hard look. Does it answer that intent? Does its headline match the offer? Is its call to action the one you would pay to put in front of this visitor? Those are decisions a partner can make and a client can check. “We monitored the bids” does not tell the client whether a careers page was an acceptable destination.

By December 31, 2027, I expect white-label scopes that retain clients to name their page inventory: one focused page per core ad group, monthly URL-exclusion reviews, and approved image sources where PMax draws creative from the site. That is a contract standard, not a claim that every account must use the same page layout. It puts ownership where the partner can exercise it. It is also the approach groas sells through Search Ads for one flat fee, with a landing page for every ad group built in, rather than billing hours to watch bids.

I learned how visible this gap can be when a client paid me for a year of weekly bid notes while location pages bled budget to irrelevant placements. The notes looked diligent. The CPA moved only after we fixed the pages and the counting. I would rather put the pages in the scope and make someone answer for them than produce another handsome record of adjustments.

What proves me wrong: if broad, one-page-per-account builds still beat one-page-per-ad-group builds on qualified cost per acquisition in head-to-head tests at the end of 2027, keep the single page. The client should buy the structure that wins, not the one with the prettier scope.

Prediction 3: Conversion-data hygiene becomes a paid deliverable by December 31, 2027

Smart Bidding acts on the conversions it receives, not the sale the business meant to measure. Tracking can still fail at the tag, the attribution window, the imported GA4 event, enhanced conversions, or internal traffic. Enhanced conversions uses SHA256-hashed first-party data to recover conversions that blocked tags would otherwise miss. That work is less glamorous than an optimization deck. It is also closer to the decision the bidding system makes every auction.

The upload window puts a deadline on the job: offline uploads more than 90 days post-click are not imported, and the limit for enhanced conversions for leads is 63 days. Miss the window, count every form start as a win through an imported GA4 event, or let office IPs pollute a primary conversion, and the account can train toward the wrong outcome. The problem is not that automation refuses to work. It is that it works on the signal you supplied.

That is why I expect serious white-label scopes to price the data work explicitly by the end of 2027. Name who checks the GA4 import. Name who keeps Data Manager uploads on schedule. Name who confirms that the primary conversion represents a qualified lead or a real sale and that internal traffic stays out. Clean signal is part of the bid strategy now. A partner cannot promise qualified CPA while treating the definition of “qualified” as somebody else’s housekeeping.

This also gives an agency something better to ask for than a monthly explanation of bid movements. Ask to see the primary conversion definition and the upload schedule. If the partner cannot show either, a lower reported CPA may only mean the system found an easier event to count. That is not the same as making customer acquisition cheaper.

What proves me wrong: if accounts with sloppy imports keep matching clean accounts on qualified CPA through 2027, data hygiene has not earned its own paid line. Until then, I would rather pay for a checked conversion than another bid note.

Prediction 4: Custom scripts shrink to a niche skill by December 31, 2027

The maintenance math is ugly. Ads API versions sunset on a conveyor: v20 and v20.1 on June 10, 2026; v21 in August 2026; v22 in October 2026; v23 in February 2027; v24 in May 2027; and v25 in August 2027. Miss a sunset and a bid script, Looker dashboard, CRM conversion upload, or feed sync can break without a warning email. The client does not get a refund for the hours spent patching a tool that was sold as efficiency.

I kept a 40-script library alive for two years. Much of it adjusted bids that Smart Bidding now handles, while every API notice brought another round of maintenance with no obvious client-facing gain. That does not make every script worthless. It makes a script pack a poor substitute for a measurable service. If custom logic cannot show what it improves after its upkeep is counted, calling it proprietary does not rescue it.

By the end of 2027, I expect scripts to be a monitoring and maintenance skill, not a headline product. Keep someone who can maintain an upload and read a change log. Be much more skeptical of a team selling anomaly detectors and budget pacers as the reason to sign. The question is what the script changes in qualified CPA, measured against an account with clean inputs and AI Max already doing its job.

What proves me wrong: if a script vendor shows incrementality on qualified CPA against that clean AI Max baseline through 2027, I will renew the whole library myself. A working control group beats my opinion every time.

What to change in your white-label contract now

These calls have separate tripwires, but they point to one buying decision. I covered the hours version in the 2028 white-label predictions piece. This one is about what replaces those hours. If Google pauses auto-migration or hands query and URL control back to advertisers, the wider thesis fails. Until that happens, I would stop paying for bid watching as though it were the center of the job.

Before you renew a white-label partner, ask for three paid deliverables you can audit:

  • Named page inventory. Give each core ad group a focused page with its own headline, proof, and call to action. Do not make one contact page answer six different intents without testing whether it can.
  • Monthly URL and image-source review. Record which pages final URL expansion may reach, which need exclusions, and which site images PMax may use. Keep blog, careers, and support pages out unless you chose them for the campaign.
  • Conversion-data ownership. Define the primary conversion as a qualified lead or real sale. Put Data Manager uploads on a schedule within the applicable window, filter internal traffic, and check the GA4 import before it trains Smart Bidding.

Then compare those deliverables with the bid-management line on your next invoice. Ask the partner to name the pages they built and the primary conversion the system uses. If they cannot, move the budget to someone who can. If I am right about December 31, 2027, the partner earns the fee by owning the inputs. Google does the bidding.