Google Demand Gen has quietly become one of the most misunderstood campaign types in the Google Ads stack. 

Many marketers still think of it as Discovery with a new name, or a top-of-funnel add-on that only works with big budgets. 

That view is more outdated than ever.

And with Google releasing a flurry of major Demand Gen updates towards the end of 2025, a lot of marketers are currently scratching their heads about making their Demand Gen strategy successful for 2026.

So, to rectify this, we masterminded a live session with Demand Gen expert Thomas Eccel to break down: 

• What actually changed inside the Demand Gen algorithm
• Why performance shifted for so many accounts in late 2025
• And how marketers should rethink their Demand Gen strategy going forward

From bidding behavior to creative signals and budget thresholds, the session made one thing clear:

Demand Gen is no longer playing a supporting role.

This recap focuses on the biggest structural changes first, before getting into how marketers should adapt their strategy in practice.

Watch the full session on-demand here, or keep reading for the key highlights and written takeaways:

Timestamps:

0:10 - Intro
04:50 - [Icebreaker poll] ChatGPT ads
08:24 - Session agenda
10:37 - Thomas' Demand Gen playbook
11:16 - Major Demand Gen algorithm update for 2026
24:16 - Demand Gen setup guide
27:54 - Demand Gen billing models
35:48 - Avoiding PMax / Demand Gen cannibalization
41:58 - Exclusive data: Demand Gen IVT
43:04 - How Demand Gen IVT compares to other campaigns
51:48 - Thomas' script to reveal Demand Gen channel data
54:00 - Google Maps Channel (New DG feature)
54:56 - Location option for presence/interest (New DG feature)
56:11 - Branded searches (New DG feature)
57:34 - YouTube engagements goal (New DG feature)
58:40 - Bonus Demand Gen Resource
59:56 - Live Q&A with Thomas Eccel
1:09:40 - Final thoughts

Demand Gen’s algorithm shift: from contextual to interest-based

One of the most important updates to Demand Gen happened quietly in late 2025. Google fundamentally changed how the algorithm decides who sees your ads.

Earlier versions of Demand Gen leaned heavily on contextual matching. Ads were placed based on the content a user was browsing at that moment, whether that was a YouTube video, a Discover feed, or Gmail. That approach kept Demand Gen firmly in the mid-funnel.

The new version works differently. The system now starts with the user, not the content. It identifies what a person is likely to buy based on behaviour across Google properties, then selects the most relevant creative to show them.

Thomas described this as a clear move toward an interest-based model that feels much closer to Meta or TikTok:

“Demand Gen is not a contextual matching algorithm anymore. It’s an interest-based algorithm that first identifies what a user is likely to buy, then finds the best creative to show them.”

This shift explains why many advertisers saw Demand Gen suddenly driving more front-end conversions, even when they had not changed their setup. The campaign type moved closer to the mid-to-low funnel by default, which altered attribution patterns and made Demand Gen appear more competitive with Performance Max.

For Demand Gen strategy, the implication is simple. You can no longer treat it as a passive awareness channel. It actively competes for conversion credit and influences lower-funnel performance, whether you intend it to or not.

Key takeaway

Demand Gen now behaves like an intent-shaping engine, not a placement-based discovery tool. Strategy needs to reflect that change.

The signals that now matter most for Demand Gen performance

As Demand Gen moved toward interest-based matching, the signals that feed the algorithm became more important. Creative, copy, and product data now act as targeting inputs, not just presentation layers.

For ecommerce advertisers, Thomas was clear that product feeds sit at the centre of Demand Gen performance. Titles, descriptions, attributes, and imagery all help the system understand what you are selling and who should see it.

“If you have a very good product feed, you send Google the most accurate picture of your product. That’s what allows the algorithm to match it to the right audience.”

The same principle applies to lead generation. Clear ad text, structured value propositions, and messaging informed by high-performing search terms help the system classify your offer correctly. Vague copy slows learning and pushes the algorithm toward less relevant audiences.

First-party data also plays a critical role. Customer match lists, high-LTV segments, and repeat buyers teach the system what quality looks like. These inputs shape lookalike expansion and improve early-stage efficiency.

Creative engagement signals complete the picture. How users interact with videos, especially shorts versus long-form, feeds back into delivery decisions. Demand Gen rewards variation and volume, giving the system room to test and adapt.

Key takeaway

Strong Demand Gen strategy starts with signal quality. Feed structure, copy clarity, first-party data, and creative variety all influence who sees your ads.

Can Demand Gen work with small budgets? Yes, but only in the right scenarios

Demand Gen is often dismissed as too expensive for smaller advertisers. Thomas pushed back on that assumption, but with clear caveats.

He defined a small Demand Gen budget as roughly €50 per day for a single campaign. At that level, success depends less on spend and more on context. Demand Gen works best as an add-on when search volume is capped, creative is proven, and core performance channels are already stable.

“I see Demand Gen as an add-on strategy. It’s not a replacement for your core performance campaigns.”

Low-budget Demand Gen makes sense when search can’t scale further due to limited demand, or when strong Meta creatives already exist and can be repurposed. In these cases, Demand Gen extends reach without pulling budget from lower-funnel channels.

It becomes risky when advertisers use it as a substitute for search, shopping, or Performance Max. Testing Demand Gen before validating creative elsewhere often leads to wasted spend and noisy data. The algorithm needs signals, and weak inputs are amplified at low budgets.

Demand Gen bidding strategies matter here as well. Maximise conversions typically outperforms max clicks for smaller budgets, as it avoids paying for low-intent traffic that looks cheap but converts poorly.

Key takeaway

Demand Gen can work on smaller budgets, but only when it complements existing performance. Creative readiness and clear role definition matter more than daily spend.

Google’s Demand Gen setup guide: useful guardrails, not a strategy

Google quietly added a new setup guide for Demand Gen campaigns inside the Google Ads UI. At a glance, it looks like another attempt to steer advertisers toward Google’s preferred defaults. In practice, it is more nuanced than that.

The guide checks fundamentals such as conversion goals, bidding setup, audience size, and estimated performance. For less experienced advertisers, this reduces the risk of launching a Demand Gen campaign with broken tracking or an audience that is far too narrow to spend.

Thomas sees value in it, with an important caveat.

“These guides are never fully neutral, but this one is actually quite useful if you understand it’s giving you estimates, not answers.”

The most helpful element is the estimated impressions and reach indicators. These give an early sense of whether your setup can realistically deliver with the budget you have. What it cannot do is decide your Demand Gen strategy for you. Creative quality, audience intent, and business context still sit outside Google’s line of sight.

Key takeaway

Use the setup guide as a sense check, not a green light. It can catch obvious issues, but it won’t replace strategic thinking.

CPC vs CPM in Demand Gen: how bidding really changes behavior

One of the most overlooked aspects of Demand Gen is that it does not operate on a single billing model. Some placements charge per click, others per thousand impressions. That distinction matters far more than most accounts reflect.

CPC placements, such as Discover image ads and Gmail, charge for engagement. CPM placements, including YouTube video and Discover video, charge for attention. Optimising both the same way leads to inefficient spend.

Thomas frames this difference very clearly:

“If you optimize for CPC, you’re paying for engagement. If you optimize for CPM, you’re paying for attention.”

For CPC-driven placements, the goal is to reduce low-quality clicks. Pricing upfront, eligibility qualifiers like “B2B only”, and neutral CTAs such as “see pricing” help pre-qualify users. Fewer clicks at higher intent usually improves efficiency.

For CPM-driven placements, every impression needs to work harder. Early hooks, proof points in the first few seconds, captions for muted views, and frequency control all lower effective CPM over time. Channel controls and audience pruning matter more here than headline tweaks.

This is where Demand Gen bidding strategies intersect with creative decisions. The billing model should shape how ads are written, designed, and sequenced.

Key takeaway

Align creative intent with how you are charged. CPC and CPM placements reward very different behaviors.

Demand Gen and Performance Max: managing overlap and cannibalization

As Demand Gen moved closer to the mid-to-low funnel, overlap with Performance Max became unavoidable. Both campaign types now compete across YouTube, Discover, and Gmail, often targeting similar users.

The risk is not just double exposure. It is distorted attribution and slower learning. Performance Max frequently claims the final conversion, even when Demand Gen created the intent earlier in the journey.

“Demand Gen can warm up the user, then PMax comes in at the last second and takes the credit.”

When this happens, Demand Gen appears to underperform in standard reports, despite driving assisted conversions. Conversion data becomes fragmented, which slows optimisation and makes budget decisions harder.

Mitigation starts with role clarity. Decide which campaign owns prospecting and which owns remarketing. Use exclusions to prevent both campaigns chasing the same users. Different creative also helps. Demand Gen should focus on narrative and education, while Performance Max leans into product and offer.

Assisted conversion reports are essential here. They show how Demand Gen contributes even when it does not close the sale.

Key takeaway

Demand Gen and Performance Max can work together, but only when roles are defined and overlap is actively managed.

Invalid traffic and Demand Gen: understanding the risk profile

Demand Gen’s reach across upper-funnel inventory brings a trade-off. Exposure to lower-quality traffic is higher than in search or Performance Max, particularly when display inventory is included.

Thomas sees clear patterns when analysing engagement metrics:

“When I look at GA4 engagement time, Display is usually the worst. That’s where I see the most invalid traffic.”

Zero-second sessions, abnormal micro-conversion rates, and poor on-site engagement are common red flags. These signals often point to automated or low-quality traffic rather than genuine users.

The solution is not to avoid Demand Gen entirely. It is to control where ads appear. Starting with Google-owned properties like YouTube, Discover, and Gmail significantly reduces risk. Display can be layered in later, once performance and quality are proven.

Demand Gen strategy should treat traffic quality as a variable, not a constant. Monitoring engagement metrics outside Google Ads is critical.

Key takeaway

Demand Gen can drive quality demand, but only when inventory and traffic signals are actively controlled.

New Demand Gen features marketers need to factor into their strategy

Google continues to add new surfaces and controls to Demand Gen, often without much fanfare. In 2026, several of these updates materially affect how Demand Gen strategy should be planned and evaluated.

One of the biggest additions is Google Maps. Promoted pins are now part of the Demand Gen inventory, giving brands visibility at moments of local intent. For location-driven businesses, this introduces a new testing opportunity that previously required separate local campaign types.

“You can now have a Google Maps-only Demand Gen campaign. That’s something we’ve never really had before.”

Location targeting has also improved. Demand Gen now supports presence-based targeting, similar to search. While this reduces accidental international spend, Thomas still recommends explicit exclusions for non-target markets to stay fully in control.

Another meaningful update is branded search conversions. This new conversion action shows when users search for your brand after seeing a Demand Gen or Performance Max ad. It does not replace revenue metrics, but it adds a valuable layer for measuring lift and intent creation.

Finally, YouTube engagement goals give advertisers the option to optimize for subscriptions, likes, or channel interactions. This is most relevant for brands actively building an audience, rather than pure lead or revenue generation.

Key takeaway

New features expand what Demand Gen can do, but each one needs a clear purpose. Default activation creates noise, not performance.

What a strong Demand Gen strategy looks like in 2026

Demand Gen has matured into a core part of the Google Ads ecosystem. It shapes intent, feeds lower-funnel campaigns, and competes for conversion credit. Ultimately, treating it as a simple awareness channel leaves performance on the table.

A winning Demand Gen strategy in 2026 starts with clarity. Clear signals, clear creative, clear roles alongside Performance Max, and clear expectations around attribution. Budget size matters less than structure, intent, and input quality.

Thomas summed it up best toward the end of the session:

“Demand Gen is here to create intent. Performance Max is here to capture it. When you let each one do its job, everything works better.”

For marketers willing to invest in creative, control inventory, and align Demand Gen bidding strategies with how the platform actually charges, Demand Gen becomes a powerful lever rather than an expensive experiment.

The opportunity is there, but the difference comes down to how deliberately it’s used.

Huge thanks to Thomas Eccel for joining us, and for dropping a seriously invaluable guide for anyone looking to optimize Demand Gen strategy going forward.

Make sure to watch the full session for additional tips, tricks, and insights from Thomas. 

Get links to all the bonus resources from the session below - and be sure to keep an eye out for our next webinar to help you stay on top of your marketing goals.

Bonus resources from the session:

• Lunio - Free Demand Gen traffic audit
• Follow Thomas on LinkedIn
• Follow James on LinkedIn
• Download Lunio’s 2026 Global IVT Report
• Video Summary of the 2026 IVT Report

• Thomas’ Demand Gen Playbook
• Demand Gen algorithm update announcement
• Running Demand Gen w/ low budget?
• New setup guide for Demand Gen
• CPC vs CPM billing for Demand Gen
• Avoiding PMax / Demand Gen Cannabilisation
• Lunio’s 100k Display placement exclusion list
• Thomas’ Google Ads Christmas Wishlist
• Claire Atkin, Check My Ads, IVT Report LinkedIn post
• Lunio’s IVT report covered in MediaPost
• James interviews Rand Fishkin about AI
• Thomas’ Demand Gen script to reveal channel data
• New DG feature: Google Maps channel
• New DG feature: Location Targeting options
• New DG feature: Branded Searches
• New DG feature: YouTube Engagement goals
• Bonus Resource: Teo’s DG Audience Strategy Deck
• Follow Teo on LinkedIn