Running PPC Campaigns in the AI Era

PPC is no longer a controllable acquisition channel, it's moving to an AI-influenced distribution system. Best practice is now a whole other reality, because the game is no longer just bidding on intent.

Increasingly, it’s about earning the right to relevance, by feeding algorithms more real opinions and easily discernible facts (profit and quality, not just volume), and building brand demand early in the journey so there’s something worth capturing at that critical commercial conversion stage.

Visibility Before Intent, Profit Before ROAS, Signals Before Spend.

 

Here are seven key areas of consideration any responsible PPC manager needs to be all over in 2026 to get the best outcomes for their clients.

1) Search is becoming “answer-first,” and that breaks the old PPC methodology

The basic concept of search ads used to be simple:

1) user expresses intent

2) advertiser matches keywords

3) landing page converts

But multiple speakers at April's Brighton SEO hero Conference pointed at the same disruptor to this logic: AI compresses and reroutes journeys, and in many cases removes the click entirely. “No-click” behaviour is rising, queries are getting longer and more conversational, and the SERP is shrinking into an answer layer where ads don’t neatly “sit next to the answer” anymore.

The deeper implication

If the platform increasingly predicts intent (instead of matching it), the advertiser’s job changes. You can’t only optimise for the moment of explicit intent; you need to influence the moments that create intent - and you need to do it in ways AI systems can read as credible and relevant across channels.

2) PPC is splitting into two jobs, demand creation and demand capture (and yes, you need both)

A recurring theme at the Hero Conference was that performance is not growth. When you optimise purely for ROAS, you often end up building a finite pool of existing demand that can include existing customers while not capturing users in that crucial middle and upper funnel.

Several talks discussed a 3 stage model:

• Demand generation: create and seed intent among non-aware audiences (with strong exclusions to avoid wasting budget on people already in-market / already brand aware).
  
  
• Demand nurture: give the algorithm more creative and narrative layers so people build preference while they research (often across social, video, native, CTV).
  
  
• Demand capture: be present when buyers finally express intent, often after “silent research” with fewer reliable keyword signals than before.
  
  

The dangers of not adopting the 3 stage approach

Brands who keep treating PPC purely as “the conversion channel” will fight a losing battle: higher CPAs, more volatility, and shrinking margins.

Brands who use PPC as a full-funnel system will see compounding effects: branded search lift, assisted conversions, and an easier time converting the demand they helped create.

3) AI doesn’t remove strategy, it makes strategy the only differentiator

Across talks, there was a consistent warning: platforms will happily optimise, but they’ll optimise toward the easiest outcomes unless you teach them otherwise.

If you don’t tell them what good value, margin, offline quality, and first-party relevance look like, the system will just go after what it can see, and that’s often low-quality conversions.

That’s why moving from the role of operator to the architect matters. The new PPC campaign manager is a:

• Signal architect (determining what data the algorithm receives)
  
  

• Test designer (validate inputs, not just outputs, through experimentation)
  
  

• Commercial integrator (profit, margin, pipeline - not just platform conversions)
  
  

Multiple speakers flagged substantial tracking loss in these harder to define multiturn user journeys.

We live in a world where many conversion signals never reach platforms, so it’s become critical to employ server-side approaches, enhanced conversions, and better consent implementation to restore more measurable attribution.

Accurate data has never been more important, yet never so elusive with the standard toolkit

AI performance improvements are increasingly “gated” behind data maturity. If your tracking is leaky, your CRM/offline signals are missing, and your conversion definitions are messy, you can’t out-bid your way to growth. You need to out-instrument competitors.

4) Profit > ROAS: the measurement reality check

The accepted ROI measure of campaigns based just on ROAS needs a re-think.

• ROAS targets can be self-sabotaging - lowering a ROAS target can sometimes increase total profit because it unlocks incremental volume at acceptable margin.
  
  
• “MOP the ROAS” was a term coined by one speaker. This stands for "Margin, Optimised, Performance" i.e., align bidding and targets to actual business economics, not blended vanity metrics.
  
  

That pulls PPC into the finance sphere: cost of goods, returns, customer type, and even cash-flow awareness can (and should) shape bidding logic and automation. This is often easier to quantify with ecommerce brands, but still should be mapped for lead gen style B2B clients whose longer sales cycle conversions often get left out of the targeting signals.

Using PPC To allocate investment

A mature PPC function is less about “winning auctions” and more about allocating capital. The question becomes: where does the next marginal pound create the most profitable growth, not just the most attributed revenue?

5) Creative is the new targeting basis (and storytelling is the new optimisation lever)

Multiple paid social talks aligned to one core truth: creative now functions as a major targeting influencer. Algorithms use engagement and response behaviour to find and expand audiences, meaning your messaging clarity and creative differentiation often matter more than micro-targeting.

What “wins” increasingly looks like content:

• first-person narratives (“I” voice), raw stories, customer stories, founder insights
  
  
• simple language, single emotions, clear urgency justifiers, guarantees to remove friction
  
  
• using external truth sources (reviews, communities, Reddit) to mine angles people actually care about
  
  

The deeper implication

If search is becoming less keyword-dependent, then brand preference and message resonance do more of the conversion work. Creative doesn’t just “improve CTR”, it trains the distribution system and builds memory structures that later reduce CPA across channels.

6) Declining search term visibility is a thing. And this makes wasted spend harder to spot.

One of the most actionable reminders was that many Google ad accounts leak 10–30% wastage spend, and that's down to an ever-increasing proportion of hidden search term data (sometimes up to 50%). This makes the PPC manager's job harder to spot where poor intent is creeping in.

When match types loosen (and we all know that's a thing, with phrase match now practically behaving like broad match) and query visibility declines, you can’t rely on the old comfort blanket of “we’ll just review the search terms and clean it up.” You still should review what’s available, but you also need other ways to spot poor intent.

The warning signs to watch out for (especially when you can’t see the queries)

• Rising CPA without a clear creative/landing page change can indicate the platform is expanding into weaker intent pockets you can’t fully see.
  
  

• CPC inflation on “visible” terms alongside a disproportionate CPA increase suggests relevance is degrading in the “invisible” portion too.
  
  

• Lead quality dilution (sales feedback, down-funnel conversion rate drops) is often the earliest indicator that optimisation is being fed the wrong signals.
  
  

What to do about it: today's way to go about negative keyword management

• Use Google’s own keyword suggestions as intent clues for themes you should exclude (if Google suggests it, Google is willing to match you to it).
  
  
• Apply shared negative lists at manager account level where sector intent patterns repeat (particularly for regulated or high-CPC verticals).
  
  
• Don’t just “neg what you see”- neg what you can predict based on shifts in CPA, lead quality, and user intent modifiers such as informational phrases like "how does" or commercial hooks such as "get quotes".
  
  
• Pair negative hygiene with conversion definition hygiene: avoid spammy micro-conversions (e.g., button clicks), and make sure the platform is learning from meaningful outcomes. Think about grouping campaigns by intent so you can split out conversion actions to the various stages of the research journey.
  
  

The deeper implication

As the platform obscures more of the matching layer, “account hygiene” becomes less about admin and more about risk management. Negative keywords and intent controls become one of the few remaining levers to protect efficiency while you transition budgets into broader visibility and demand creation.

7) The lines are blurring now, with PPC, SEO, and PR becoming a merged visibility framework

Search algorithms now evaluate more than your landing page. They assess credibility from wider signals: content, reviews, social presence, and brand mentions. Reddit and LinkedIn were called out as particularly prominent places AI crawlers (and journalists) source narratives from, shifting PR away from “links” and toward “make noise, earn attention, build authority.”

Some practical insights from these talks include using high performing paid search terms and ad copy learnings to shape SEO content maps, organic meta data and social posts. Compare PPC converting themes with Search Console data, and create a loop where paid informs organic and organic improves paid eligibility.

The reality of this merged entity

The winners won’t be the brands with the best PPC tricks. They’ll be the brands with the most consistent and credible visibility footprint - the kind AI systems can safely recommend and buyers can trust.

The strategic takeaway

BrightonSEO didn’t reveal a single “hack.” It confirmed a shift in operating principles:

• Visibility before intent (because AI and no-click SERPs compress journeys).
  
  
• Profit before ROAS (because blended metrics hide what’s scalable).
  
  
• Signals before spend (because AI only optimises what it can learn).
  
  
• Story before targeting (because creative is how algorithms find buyers now).
  
  
• Hygiene before scale (because query opacity makes waste easier to miss, so negatives and intent controls matter more, not less).
  
  

PPC is no longer a solus channel. It’s a growth platform spanning data, creative, measurement, and brand, operating inside AI-driven distribution engines. And the job is to be the architect of the inputs that create profitable demand, not just optimise for the diminishing number of conversions you can still measure.