Search advertising is the most efficient paid channel most brands have ever run.
It's also a ceiling.
Search captures intent. It reaches people who already know they have a problem, have already formulated a solution category, and are actively looking for someone to provide it. When those conditions exist, Search converts with extraordinary efficiency. The ROAS looks excellent. The CPA looks clean. Everyone is happy.
Until the Search volume plateaus.
This is the point most brands hit and don't understand: the Search campaigns have captured every available unit of existing demand, and growth has stalled — not because the campaigns are underperforming, but because the pool of people searching for what you sell is finite. You can optimise bids, improve quality scores, and expand keyword coverage indefinitely, but you cannot capture intent that doesn't exist yet.
The strategy that generates growth past that ceiling is the one most brands haven't run: upper-funnel PPC. YouTube pre-roll, Demand Gen across Google's visual surfaces, awareness display. The campaigns that don't capture demand, they create it. They reach people before they're searching and give them a reason to start.
This is the most skipped strategy in the average PPC account. Not because it doesn't work. Because it's harder to measure, harder to justify in a ROAS conversation, and requires a different kind of creative than the bottom-funnel ads most teams are built around.
Here's the case for it with the mechanics, the measurement approach, and the specific way to run it alongside your existing campaigns.
🔍 STRATEGY — DEMAND CAPTURE VS DEMAND CREATION
The distinction that explains why Search has a ceiling and what sits above it
Every paid media channel performs one of two fundamental jobs.
Demand capture: reaching people who already have intent. They've formulated the problem, identified the solution category, and are in active evaluation mode. Search is the purest demand capture channel, someone types a query, and your ad appears at precisely the moment that intent exists. The efficiency is real. So is the limitation: you can only capture demand that already exists.
Demand creation: reaching people before intent exists. They have the problem your product solves, but they haven't yet formulated it as a search query. They're not looking for your product because they don't yet know they should be. Upper-funnel advertising reaches them in this pre-intent state, raises their awareness of the problem and the solution, and creates the demand that lower-funnel campaigns will later capture.
The relationship between these two strategies is causal, not parallel. Upper-funnel investment in demand creation generates the branded search volume and category search growth that Search campaigns then convert. A well-run YouTube awareness programme will produce a measurable lift in branded search queries - not immediately, but over weeks and months as the audiences who saw the ads move through their purchase journey. Most Google advertisers are running Search and Performance Max, calling it a full-funnel strategy, and wondering why their pipeline dries up the moment branded demand softens. They're capturing intent. They're not creating it.
The measurement problem is real and worth naming honestly: upper-funnel campaigns don't convert directly in the same way Search campaigns do. A user who sees a YouTube pre-roll isn't going to click through and buy in the same session. The conversion happens days or weeks later, through a different channel, which receives the attribution credit. Last-click and even data-driven attribution models systematically undervalue upper-funnel activity precisely because the conversion path is long. This is not a reason to avoid upper-funnel investment, it's a reason to measure it differently.
CORE INSIGHT: Search campaigns are demand capture machines. They're highly efficient and will always be the core of a performance PPC account. But they have a hard ceiling: the volume of existing search intent in your category. Upper-funnel PPC doesn't compete with Search - it feeds it. The branded search lift, the increased direct traffic, the warmer audiences landing on your remarketing lists, these are the measurable outputs of upper-funnel investment. The teams that understand this relationship allocate budget across both functions. The teams that don't run Search until it plateaus and then wonder why growth has stopped.
→ Takeaway: Pull your branded search volume trend over the last 12 months. Is it growing? Flat? Declining? Branded search volume is the clearest proxy signal for whether your upper-funnel activity is building brand awareness in the market. If it's flat or declining while your bottom-funnel campaigns are stable, you're operating at the demand-capture ceiling with nothing feeding the pipeline from above.
📺 YOUTUBE — THE UPPER-FUNNEL CHANNEL MOST BRANDS UNDERUSE
What YouTube advertising actually produces and how to run it correctly
YouTube reaches over 2.7 billion logged-in users monthly. YouTube Shorts crossed 70 billion daily views in 2025. It is the second-largest search engine on the planet, and critically for upper-funnel strategy, it's where people spend time when they're not searching for anything. That pre-intent state is exactly where demand creation needs to operate.
The format logic matters before anything else. YouTube offers multiple ad formats and using the wrong one for the wrong objective is the most common waste in YouTube campaigns:
6-second bumper ads — unskippable, 6 seconds maximum. The right format for pure brand awareness and reach at scale. They can't carry a complex message but they can build brand familiarity efficiently and at low CPM. Best used for frequency building across large audiences.
15-30 second skippable in-stream — the workhorse format for consideration-stage campaigns. Viewers can skip after 5 seconds, which means the first 5 seconds determine whether you've earned the view. The creative rule: if your hook isn't strong enough that a stranger would choose to keep watching after 5 seconds, the ad will be skipped by the majority. You pay only when someone watches 30 seconds or engages, which means every view is a qualified impression.
In-feed video ads — appear in YouTube search results, the recommended feed, and the homepage. These reach users in discovery mode and tend to attract higher-intent viewers than in-stream, because the viewer actively chose to watch rather than being intercepted. Strong for consideration-stage content with real depth.
YouTube Shorts ads — vertical format, no skip button, appearing between organic Shorts. Requires purpose-built 9:16 creative. This is a genuinely different creative environment that repurposed landscape ads don't work in. Brands that have invested in native Shorts creative are finding it highly cost-efficient for reach.
The measurement approach for YouTube needs to be different from Search. The primary signals: branded search volume lift (track your brand's search impression volume before and during YouTube campaigns), view-through conversion rate, and assisted conversions in GA4's key events attribution paths report. Google also offers Conversion Lift studies - an incrementality tool that compares users exposed to ads against a control group, which gives the most honest read on what the YouTube campaign is actually producing.
CORE INSIGHT: YouTube advertising's primary output is not direct clicks. It's brand memory, category association, and the purchase intent that activates days or weeks later through other channels. Measuring it against the same CPA targets as Search produces a number that looks terrible — because it's measuring the wrong thing. The right measurement is branded search lift, assisted conversion volume, and incremental conversion rate in holdout tests. Brands that measure correctly invest confidently. Brands that measure incorrectly conclude YouTube doesn't work, deprioritise it, and wonder why their Search volume won't grow.
→ Takeaway: If you're going to run YouTube for the first time, commit to at least 8 weeks of consistent spend before evaluating. Monitor branded search impression volume weekly. Set up a Conversion Lift study from the start. At week 8, compare branded search volume, direct traffic, and assisted conversions against the 8 weeks prior. That comparison — not the YouTube campaign's direct ROAS — is the actual performance read.
🖥️ DEMAND GEN — GOOGLE'S MOST UNDERDEPLOYED CAMPAIGN TYPE
What Demand Gen is, where it runs, and why it belongs in most accounts
Google retired Discovery campaigns in 2023 and replaced them with Demand Gen - a significantly more capable campaign type that runs across YouTube (including Shorts), Gmail, Google Discover, the Display Network, and Google Maps. By early 2026, Google had evolved Demand Gen further: adding shoppable connected TV inventory, product feed integration with a 33% reported conversion lift, creator content partnerships, and tROAS bidding for brands that want to run it as a conversion-driving campaign.
Demand Gen is Google's visual, social-style campaign type running across YouTube, YouTube Shorts, Google Discover, and Gmail, built to reach audiences before they're searching. Unlike Performance Max, Demand Gen gives you real audience controls: custom segments, lookalike lists, and the ability to exclude placements you don't want.
The positioning in 2026 is mid-to-upper funnel with performance ambitions. Google pitches tROAS bidding on Demand Gen as a direct Meta Advantage+ competitor and for certain advertisers, the comparison is legitimate. But the more durable use case remains what it was originally built for: introducing your brand to audiences who don't yet know they should be searching for you, using visually compelling creative across Google's highest-traffic surfaces.
The practical case for Demand Gen is simple: most brands running Performance Max are already serving YouTube and Discover inventory through PMax, but with almost no control over how it's handled. Demand Gen gives you control over audiences, creative, and placement priorities that PMax obscures. For brands that want to run genuine upper-funnel campaigns, with proper creative, proper audience targeting, and proper measurement - Demand Gen is the vehicle.
The creative requirements are meaningful. Demand Gen doesn't respond well to performance ad creative, the direct response, feature-listing, discount-led formats that work in Search and Shopping. It needs brand creative: emotionally resonant, visually arresting, and built around the feeling rather than the feature. This is the investment most brands haven't made and the real reason most haven't run Demand Gen seriously.
CORE INSIGHT: Demand Gen is still the most underdeployed campaign type in the average Google Ads account — not because it doesn't work, but because running it well requires brand creative that most performance-marketing teams don't have and can't easily produce. The irony is that the brands most likely to benefit from Demand Gen — those hitting their Search ceiling — are often the ones least set up for the creative investment it requires. The solution isn't to skip Demand Gen. It's to treat creative development as the primary investment, with campaign management as secondary.
→ Takeaway: Before running Demand Gen, audit your creative library. Do you have video creative that's emotionally compelling without a direct response call to action? Do you have image creative built for feed placements — not banner ads? If the answer is no, that's the prerequisite investment. Running Demand Gen with repurposed performance creative produces mediocre results and confirms the wrong conclusion that the channel doesn't work.
📡 DISPLAY — THE AWARENESS CHANNEL MOST BRANDS MISUSE
Why display advertising underperforms for most brands and how the ones using it well are running it
Display advertising has a reputation problem. Most marketers have run display, seen mediocre ROAS, and deprioritised it. The experience is consistent enough that "display doesn't work" has become a received wisdom in many performance marketing teams.
The diagnosis is almost always the same: display is being measured against the wrong objective with the wrong creative.
Display advertising is an impression medium. Its primary job is to generate brand familiarity at scale, to ensure that when a prospect eventually reaches the Search bar to look for a solution, your brand is already in their consideration set. Measuring a display campaign on direct-click ROAS is measuring the right output from the wrong tool. It's equivalent to judging a podcast sponsorship by immediate web traffic - missing the entire mechanism by which it actually works.
The brands using display effectively in 2026 are running it with three specific principles:
Audience quality over reach volume. Programmatic display has a placement quality problem — significant portions of the Google Display Network serve ads on low-quality sites with minimal real human attention. The solution is audience targeting specificity: custom intent audiences built around competitor brand searches, in-market segments with tight behavioural criteria, and customer match lookalikes. Reach built on cheap inventory isn't reach — it's wasted impressions.
Sequential messaging. The brands getting measurable return from display aren't running the same ad to everyone indefinitely. They're running awareness-stage creative to cold audiences, consideration-stage creative to those who've engaged, and conversion nudges to those who've visited the site. Sequential creative - telling a story across multiple exposures in a defined order, produces recall and conversion rates that single-creative campaigns can't match.
Brand safety and placement controls. Managed placements on relevant, high-quality sites produce significantly better recall and brand association than broad network placements. The investment in curating a placement list is recovered in the quality of the impressions that remain.
CORE INSIGHT: Display's reputation for underperformance is earned when it's used as a cheap conversion channel — which it was never built to be. It's justified when it's used as a brand familiarity engine, measured on reach quality and brand recall rather than last-click conversions. The brands that run display well treat it as infrastructure: not a campaign they turn on and off, but a permanent presence in the visual environment their prospects occupy. That sustained presence is what produces the brand recognition that makes Search campaigns more efficient.
→ Takeaway: If you're going to run awareness display, set separate measurement criteria from your performance campaigns before you launch. Define success as: cost per thousand quality impressions, brand recall lift (via a survey study), and assisted conversion rate among exposed audiences. Don't put display ROAS in the same dashboard as Search ROAS. The comparison is misleading and will produce the wrong budget decision.
🔧 TOOL OF THE WEEK
Google's Brand Lift measurement (available through Google Ads for YouTube campaigns above a minimum spend threshold) the most honest measurement tool for upper-funnel campaigns.
Brand Lift surveys a random sample of users exposed to your YouTube ads and compares their brand recall, awareness, and consideration against a matched control group who weren't exposed. It gives you a direct, incrementality-based read on what the campaign is producing, separate from any attribution model.
The specific use: run Brand Lift from the start of any YouTube or Demand Gen campaign. Don't wait to see if the direct ROAS justifies it. Set the Brand Lift study as the primary measurement instrument and use direct conversions as the secondary signal. At the end of eight weeks, the Brand Lift data will tell you whether the campaign is doing the job upper-funnel advertising is designed to do, before you make a budget decision based on a metric it was never meant to optimise.
YOUR ONE ACTION THIS WEEK
Check whether you have a demand creation strategy or just a demand capture one.
Open your Google Ads account. List every active campaign type. Then ask: which of these campaigns reaches people who are not currently searching for what I sell?
If the answer is none of them, if everything is Search, Shopping, or Performance Max weighted toward conversion - you have a demand capture account. That's efficient. It's also limited to the size of the existing search pool.
The question worth asking: what would it take to run one upper-funnel campaign for the next 90 days and measure it properly? Not on ROAS. On branded search lift, assisted conversions, and brand recall.
That 90-day investment produces data no optimisation of existing Search campaigns can generate. And it's the data that tells you whether there's a bigger market available above the ceiling you're currently at.
