- Does this Meta update change anything for my campaigns?
- What really happened: what did Meta do with GEM?
- How do these numbers compare to Meta's financial results?
- Three concrete scenarios: what changes for an e-commerce, a professional firm, and a manufacturing company?
- What changes in practice for managing Meta campaigns?
- Are Meta's numbers reliable, or should I take everything with a grain of salt?
- Why is everyone talking about this now: is the digital advertising market really growing?
- How to get started: concrete steps to make the most of GEM?
- What are the most common mistakes to avoid with GEM?
- How to tell if GEM is really improving my campaigns?
- So, is GEM really worth anything for your company?
- How we see it at SHM
- Frequently Asked Questions
- GEM (Generative Ads Model) is Meta's new unified AI system that replaces separate models for Facebook, Instagram, and Reels with a single central brain.
- Meta claims +18.9% in conversions and +12% in CTR: these are self-reported numbers, but they are consistent with the growth in advertising revenue (+27% in Q2 2026).
- The improvement happens automatically: it doesn't require manual actions, but it should be checked in campaign reports to understand the real impact on your account.
- The Italian digital advertising market is worth 6.2 billion euros in 2025 and is growing towards 7 billion in 2026: those investing in Meta are already part of this transformation.
Meta states that its new AI model for advertising has increased conversions by 18.9% across all its platforms. If you have active campaigns on Facebook or Instagram, this change is already happening in your accounts — without you having done anything.
Does this Meta update change anything for my campaigns?
Yes, and automatically. GEM is already active on all Meta ad accounts: you don't need to do anything to turn it on, but you do need to understand what to expect from your reports.
GEM — acronym for Generative Ads Model , which is the generative AI model for advertising — is the system Meta uses to decide which ad to show to which user, at what time. Until recently, Facebook, Instagram, and Reels each had their own separate system. Now there is a single central model that learns from all three together.
For those managing campaigns, this translates into a more powerful optimization system working in the background. Meta claims an 18.9% increase in conversions and a 12% increase in click-through rate (CTR, which is the percentage of people who see an ad and click on it). The numbers come only from Meta and are not verifiable from the outside, but they are consistent with the company's advertising revenue growth.
| Your situation | Does GEM affect you? | why |
|---|---|---|
| You have active campaigns on Facebook or Instagram | Yes | GEM is already active: your ads are being optimized by the new system |
| Use conversion objectives (purchases, leads, app installs) | Yes | GEM optimizes these very objectives in a unified way across platforms |
| You have a limited budget and want to maximize every euro | Yes | A more efficient system can improve the cost per conversion without increasing spending |
| Run only awareness or reach campaigns | Partially | GEM also improves CTR (+12%), but the main impact is on conversions |
| You have no active Meta campaigns | No, for now | There is no direct impact, but it is worth evaluating if the channel is suitable for your industry |
What really happened: what did Meta do with GEM?
Meta has replaced the separate optimization systems for Facebook, Instagram, and Reels with a single central AI model, called GEM, which learns from all data together and then instructs the smaller models that serve ads in real time.
The structural change is significant. Previously, every advertising surface — Facebook Feed, Instagram, Reels — had its own prediction system. Each learned from its own data, without sharing anything with the others. GEM breaks this separation: a single huge model learns from all user behavior across all Meta apps, then transfers this knowledge to the smaller models that decide in real-time which ad to show.
The new architecture is called Kunlun and replaces the previous one (called Interformer). According to Meta, Kunlun achieves double the performance with the same amount of compute, and has allowed for a 2.5x increase in model complexity compared to previous versions. All described in the technical post published on Engineering at Meta .
GEM is not brand new: the first public document dates back to November 2025. The current one is the third update in less than a year, and it consolidates previous progress into a single mature system.
An important warning: Meta's numbers don't have a stated benchmark. We don't know if they come from A/B tests, internal evaluations, what period they are measured over, or how they are weighted. These are self-reported figures from a company talking about its own product. They should be read as a directional signal, not as a guarantee for your specific account.
To understand how to measure the true impact of paid campaigns on your business, the article on conversion lift and paid campaign measurement offers a practical method independent of vendor claims.
How do these numbers compare to Meta's financial results?
Meta's financial data confirms real and consistent growth: advertising revenue increased by 27% in the second quarter of 2026, which makes the impact of GEM plausible — even if not verifiable in detail.
Meta's advertising revenue growth is documented in official filings with the SEC (the US commission that regulates financial markets). In the second quarter of 2026, Meta recorded $59.36 billion in advertising revenue, an increase of 27% compared to the same period in 2025, according to the Form 8-K filed with the SEC . In the first quarter of 2026, growth was even more marked: +33%, with 55 billion dollars in revenue, as reported in the Q1 2026 Form 10-Q .
This growth doesn't just depend on GEM, but the system is explicitly mentioned in quarterly reports as one of the main drivers. In Q2 2026, Meta's CFO attributed an 8.3% increase in ad clicks and a 15.7% increase in conversions on Facebook to GEM and user understanding models.
There is, however, a discrepancy to note: the technical post from October 2026 attributes a 5.8% conversion uplift to Facebook for targeted users, while the Q2 2026 communication mentioned 15.7% by combining GEM with other models. The two numbers do not necessarily contradict each other—they measure different populations, different periods, and different attributions—but Meta does not explain the gap. Those reading campaign reports should keep in mind that the numbers declared at the system level do not automatically translate into a +18.9% on their own account.
Three concrete scenarios: what changes for an e-commerce, a professional firm, and a manufacturing company?
The impact of GEM varies based on the campaign objective and the type of conversion you are optimizing for. Here are three illustrative scenarios to understand where improvement is most likely.
Note: the following calculations are illustrative examples. The numbers are meant to show the logic of the reasoning, they are not real client data.
| Company type | Campaign objective | Expected impact from GEM | What to check in the report |
|---|---|---|---|
| E-commerce (e.g. clothing) | Website purchases | Potential reduction in cost per purchase thanks to more precise targeting across Facebook and Instagram | Cost per purchase, ROAS (return on ad spend) week over week |
| Professional firm (e.g., consulting) | Lead generation (contact forms) | Better lead quality thanks to unified optimization across multiple surfaces | Cost per lead, qualification rate of received contacts |
| B2B manufacturing company | Qualified traffic to the site or quote requests | More limited impact: GEM is optimized for quick conversions, less so for long sales cycles | Ad CTR, time on site, contact requests |
For an e-commerce with a 5,000 euro monthly budget on Meta and a current cost per purchase of 25 euros (200 purchases per month), even just a 5% improvement in the conversion rate would lead to 210 purchases with the same budget — that's 10 more sales without spending an extra euro. This is the type of calculation to do on your actual account, not on the average numbers reported by Meta.
For manufacturing companies with long sales cycles, the direct impact is more limited. GEM optimizes quick and measurable conversions well. If your sales process takes weeks or months, the system struggles to learn fast enough. In this case, it is worth reading how AI agents can support campaigns with complex cycles .
What changes in practice for managing Meta campaigns?
GEM works automatically: there are no new settings to turn on. The practical change concerns how to read the results and how to structure campaigns to make the most of the new system.
- 1 · Meta (automatic) GEM optimizes in the background The unified model analyzes user behavior across all apps and updates which ads to show in real time. No action required on your part.
- 2 · Ads manager Verify the learning period Meta campaigns need a learning period (usually 7-14 days) to stabilize. With GEM, this process can be faster thanks to data shared between platforms.
- 3 · Ads manager Check reports by platform Compare performance on Facebook, Instagram, and Reels separately. GEM optimizes differently on each surface: you might find significant variations between them.
- 4 · Marketing Manager Update reporting for the client or management If you manage campaigns on behalf of a company, communicate this update in your next report. An improvement in conversions should be contextualized: it's GEM, not just the creative.
- 5 · Ads manager Avoid frequent campaign changes GEM learns from data: every major change (budget, audience, creative) resets the learning phase. Fewer manual interventions = more data for the system.
An often underestimated aspect: GEM optimizes better when campaigns use clear and measurable conversion goals. If you are using traffic or reach goals for campaigns that you actually want to convert, you are leaving performance on the table. This applies regardless of GEM, but with the new system, the gap between well-configured and poorly-configured campaigns tends to widen.
For those who want to dive deeper into how Meta is evolving creative tools in parallel with algorithmic optimization, the article on New Meta AI tools for advertising creativity and insights offers a complementary picture.
Are Meta's numbers reliable, or should I take everything with a grain of salt?
The numbers are self-reported and cannot be verified externally, but they are consistent with official financial data. They should be read as a directional signal, not as a promise for your specific account.
The original source, PPC Land , it says it clearly: Meta does not specify the measurement period, the method (A/B test or internal evaluation), nor how the numbers are weighted. The 18.9% is an aggregate average across all surfaces and all objectives — which means that some campaigns will have seen more, others less, some nothing at all.
That said, there are signs that make the numbers plausible. Meta's advertising revenue growth is documented and verifiable. The average price per ad grew by 12% in Q2 2026, and impressions by 14%: if the system didn't work, advertisers would stop spending.
| Meta statement | Verifiable? | Support signal |
|---|---|---|
| +18.9% conversions across all surfaces | No, for now | Consistent with Q2 2026 ad revenue growth (+27%) |
| +12% CTR across all surfaces | No, for now | The price per ad grew by 12% in the same period (SEC Filing Q2 2026) |
| +5.8% Facebook conversions for targeted users | No, for now | More specific data, but 'targeted users' is not defined in the technical post |
| GEM is the industry's largest RecSys foundation model | No, for now | Meta does not provide comparison metrics with other competitors' systems |
| Kunlun offers 2x performance with the same compute | No, for now | Internal technical data, not replicable from the outside |
Practical advice: don't wait for Meta to tell you if GEM is working for you. Look at your data. Compare conversions from the last 4 weeks with the same period last year, taking seasonality into account. If you see an improvement, great. If you don't, the problem might be in your campaign structure, your creative, or your tracking — not in Meta's system.
Why is everyone talking about this now: is the digital advertising market really growing?
Yes, and significantly so. The Italian digital advertising market is worth 6.2 billion euros in 2025 and is heading towards 7 billion in 2026. Those investing in Meta are part of a rapidly expanding market, where AI systems like GEM are rewriting the rules of the game.
The most relevant data for those managing budgets on Meta is this: 83% of online advertising investments in Italy end up on international platforms like Meta and Google, according to the Internet Media Observatory of the Politecnico di Milano . This means that Meta isn't just one of many options: for most Italian companies, it's the main channel. An improvement in Meta's optimization system therefore has a direct impact on a huge share of the country's advertising investments.
At the European level, the market reached 131 billion euros in 2025 with a growth of 10.5%, according to IAB Europe . Retail Media — that is, advertising within e-commerce platforms — has grown by 16.7%. The context is that of an accelerating market, where AI systems are becoming the main differentiator between those who get results and those who don't.
How to get started: concrete steps to make the most of GEM?
GEM is already active, but there are specific actions you can take to put your campaigns in the best position to benefit from it.
- 1 · Ads manager Check conversion tracking GEM optimizes for the conversions it can measure. If the Meta pixel isn't set up correctly or key events (purchase, lead, add to cart) are missing, the system is working in the dark. Check in Events Manager that all events are active and receiving data.
- 2 · Ads manager Use campaign objectives aligned with your true goal If you want sales, use the 'Sales' objective. If you want leads, use 'Leads'. GEM optimizes based on the stated objective: a mismatch between the objective and the expected result reduces the system's effectiveness.
- 3 · Ads manager Let campaigns run without changes for at least 7 days Every significant change resets the learning phase. GEM needs data to learn: the more stable the campaign, the better the system can optimize.
- 4 · Marketing Manager Compare data with the previous period Take the last 4 weeks and compare them with the same 4 weeks from last year (or the previous quarter, adjusting for seasonality). Look for variations in cost per conversion and CTR.
- 5 · Marketing manager Update internal reporting When presenting results to management or a client, put any improvements into context: GEM is a structural factor, not a random variation. This helps in making more informed budget decisions.
A step often overlooked: the creatives. GEM optimizes ad delivery, but it can't compensate for weak creatives. An ad that doesn't grab attention in the first 2-3 seconds won't be saved by the algorithm. For holiday and year-end campaigns, Meta has published specific guides that you can find in the article on Meta holiday 2026 guides and AI playbook for Christmas campaigns .
What are the most common mistakes to avoid with GEM?
The main mistakes aren't about GEM itself, but how you set up the campaigns that GEM needs to optimize. A powerful system on a weak foundation won't produce results.
| Common mistake | Why it's an issue with GEM | How to fix it |
|---|---|---|
| Incomplete or broken conversion tracking | GEM optimizes based on what it measures: without reliable conversion data, the system learns poorly | Check the pixel and events in Meta Events Manager before anything else |
| Changing campaigns every 2-3 days | Every change resets learning: the system never accumulates enough data to optimize | Plan campaigns with at least 2 weeks of stability before evaluating results |
| Budget too low to generate conversions | GEM needs at least 50 conversions per week to optimize effectively (Meta threshold) | Focus your budget on fewer campaigns with clear goals, instead of spreading it across many |
| Attribute all improvement to GEM | GEM is a factor, not the only one: seasonality, creative, and offer matter just as much | Use correct comparison periods and consider all factors in reporting |
| Ignoring variations between platforms | GEM optimizes differently on Facebook, Instagram, and Reels: aggregating everything hides opportunities | Segment reports by placement and compare performance by surface |
How to tell if GEM is really improving my campaigns?
There is no 'GEM active' indicator in the Meta dashboard. You need to measure your campaign results with the right KPIs and compare them to previous periods.
| Indicator (KPI) | What it measures | Positive signal | Control frequency |
|---|---|---|---|
| Cost per acquisition (CPA) | How much you spend for each purchase, lead, or target action | CPA reduction with the same budget | Weekly |
| CTR (click-through rate) | Percentage of people who click on the ad after seeing it | CTR increase without changing creatives | Weekly |
| ROAS (Return on Ad Spend) | Euros of revenue generated for every euro spent on advertising | ROAS increase over time | Monthly |
| Post-click conversion rate | Percentage of people who click and then complete the goal action | Increase indicates better quality of traffic brought by Meta | Monthly |
| Ad frequency | How many times on average a person sees the same ad | Stable or decreasing frequency with growing CTR = more precise targeting | Weekly |
The most useful comparison is year-over-year, not week-over-week. Short-term variations are often statistical noise. If after 4-6 weeks the cost per conversion has dropped and the CTR has risen without you changing creatives or budget, it is likely that GEM is contributing. For a more rigorous methodology for measuring campaign impact, the article on conversion lift and paid campaign measurement explains how to isolate the real contribution of advertising from external factors.
So, is GEM really worth anything for your company?
It is worth it if you have active Meta campaigns with clear conversion goals, working tracking, and enough budget to generate at least 50 conversions per week. In this case, GEM is already working for you and you should see an improvement in the cost per conversion in the coming weeks.
A middle ground you have active Meta campaigns but with a limited budget or mixed objectives (awareness + conversion). GEM can still help, but the impact will be more limited. It is worth optimizing the campaign structure before expecting big jumps in performance.
Not yet if you do not have active Meta campaigns or you have very long and complex sales cycles (industrial B2B, large contracts). GEM is optimized for fast and measurable conversions: in these contexts, other channels or approaches may be more effective.
How we see it at SHM
For us, GEM is an example of how AI is becoming invisible in the daily work of those who manage campaigns. The system optimizes in the background, without the ads manager having to do anything. This is exactly the kind of automation that makes sense: it frees people from the routine of manual optimization and gives back time for the work that really matters — strategy, creative, and data analysis.
We believe AI should be integrated to free companies from routine, not to replace human judgment. GEM is a good example of this principle in action: the system does the heavy lifting of optimization, but strategic decisions—which goal to pursue, which message to communicate, how much to invest—remain in the hands of people.
The risk we see, however, is that the numbers reported by Meta are taken at face value without verification. An average +18.9% across all campaigns worldwide says nothing about your specific account. Our approach with clients is always to look at real data, not vendor promises.
If you manage Meta campaigns and want to understand if GEM is really improving your results, we can help you with an analysis of your account. We work on Meta campaigns and google ads campaigns with an approach based on real data, not system averages. For those who want to understand how to integrate AI into their marketing strategy in a structured way, our service of strategic AI integration is the starting point. You can find more insights on campaigns and performance in our section Marketing and ADV performance .
Frequently Asked Questions
GEM is already active on all Meta ad accounts: it requires no manual action. The system works in the background and automatically optimizes ad delivery on Facebook, Instagram, and Reels. You won't find a 'GEM on/off' setting in the campaign dashboard.
Not necessarily. 18.9% is an aggregate average across all campaigns and all advertisers: some campaigns will have seen more, others less, some nothing at all. Meta does not specify the measurement method or the period. The only way to know the impact on your account is to compare your actual data before and after.
There are no mandatory changes, but some best practices help GEM work better. Make sure conversion tracking is correct, use campaign goals aligned with your real business objective, and let campaigns run without frequent changes for at least 7-14 days. GEM learns from data: the more stable the campaign, the more the system can optimize.
GEM is optimized for fast and measurable conversions, so it works best for e-commerce and businesses with short sales cycles. For B2B with long and complex cycles, the direct impact is more limited because the system struggles to accumulate enough conversion data to optimize effectively.
Compare the cost per acquisition (CPA) and CTR of the last 4-6 weeks with the same period of the previous year, taking seasonality into account. If the CPA has dropped and the CTR has risen without any changes to the creatives or budget, it is likely that GEM is contributing. There is no 'GEM' indicator in the Meta dashboard.
Not directly, but GEM optimizes better with budgets sufficient to generate at least 50 conversions per week per campaign — this is the general threshold indicated by Meta for optimal learning. If the budget is limited, it's better to focus it on fewer campaigns with clear objectives rather than spreading it across many campaigns with low volume each.
GEM is not a promise of results: it is a more powerful system that amplifies both well-built campaigns and configuration errors.
Related articles
Discover more articles exploring similar topics, selected to offer you a more complete and stimulating perspective. Each piece of content is carefully chosen to enrich your experience.
