Frontline Sam

125 posts

Frontline Sam banner
Frontline Sam

Frontline Sam

@DoualleSamuel

I run a b2c performance marketing agency; growth, startup, performance. 🇫🇷 earning in $ & £, spending in aed and €.

US/SA/EU Beigetreten Şubat 2019
882 Folgt133 Follower
David Herrmann
David Herrmann@herrmanndigital·
This week, like many other weeks in 2026 (2025 / 2024 / 2023 / 2022...) Meta ads simply broke. You probably haven't heard a peep from Meta on it, but it's been brutal. We're talking turning ad accounts off brutal. Some of you were probably 100% immune. Others had no idea because your teams suddenly tried shifting their creative / offers. Others work with agencies that probably set it and forgot it. Whatever it may be... I sound like a broken record here venting. But this just simply is becoming more and more unacceptable at the sheer level of these bugs. There feels like no recourse. No post-mortems. No true apologies. Just simply, break then kill a brand's performance for around a week, get a small refund, then move on to the next time it breaks. I'm tired guys. I know X is here to post about AI automations and how awesome I am, but the reality is this week (much like many other weeks prior) I find myself being therapist to CMOs and brand teams trying to figure out is this a bug, a creative issue, or the business. More and more it's a bug. I want Meta to be stable, but every year from Feb - May it's just a mess. Do we simply just pull budget during Meta's annual bug season? I'm out of answers personally but on behalf of the brands that pay me to manage advertising across every ad platform, I'd like some answers from Meta because it's the only one plagued with these issues at this scale.
English
71
35
465
115.3K
Frontline Sam
Frontline Sam@DoualleSamuel·
@drewfallon12 how does a brand triple its ARR from 100m to 300m in 3 months?
English
2
0
3
1.1K
Drew Fallon
Drew Fallon@drewfallon12·
UNILEVER TO ACQUIRE GRUNS FOR UNDISCLOSED SUM IN HUGE VMS SPLASH Grüns, one of the best stories in CPG of the last decade, has been punctuated with a massive deal to Unilever. The VMS M&A wave is in the early innings! Financial terms weren't disclosed, but given Gruns' disclosed revenue, growth rate, and Unilever's annual stated M&A budget - it was definitely a big one. The Numbers  - Founded: August 2023 by Chad Janis (confirmed a chad)  - $100M ARR: May 2025 (month 21)  - $300M ARR: August 2025 (month 24)  - Last valuation: $500M (April 2025 funding round led by Headline)  - Profitable within 14 months of launch Price undisclosed, but the $500M valuation set the floor 12 months ago. With $300M+ ARR at the time of deal announcement, rough benchmarks: - Huel acquired by Danone for ~3.4x revenue - 2026 - Liquid I.V. / Unilever - ~2.6x revenue - 2020  - Dr. Squatch / Unilever ~3x revenue - 2025 - poppi // PepsiCo ~3x revenue - 2025 Unilever has been making an all in effort into personal care and wellness, with brands like Liquid I.V. & Nutrafol set to pass $1B in sales, UL is putting together a formidable wellness portfolio. Plus, they spent $1.5B on Dr. Squatch last year. Early investors include Sugar Capital, Vanterra Capital, and more.
Drew Fallon tweet media
English
49
21
551
208.1K
Amir D
Amir D@starks_arq·
We made this viral Kitkat x Druski video in just 20 minutes... I just made a guide covering the complete workflow: - Agent pipeline - Storyboard flow - Video models And I'm giving it away for free Comment “DRUSKI” and RT to receive the full guide
English
508
246
607
60.6K
Frontline Sam
Frontline Sam@DoualleSamuel·
@brandonkuipers @Meta @nicolamen Probably, your ad account has been flagged as operating in the health & wellness category (weight loss), which results in pixel restriction and sometimes in account suspension. Were you using a 3rd party provider like Popsixle, Customerlabs or Taghero?
English
0
0
0
113
Brandon Kuipers
Brandon Kuipers@brandonkuipers·
Fell victim to the @Meta ban wave last week. Lost my appeal, and my account - which I've had since 2008 - is gone forever. Over $30M spent since 2012. Done roundtables with @nicolamen. Given talks at Ad Rep training. Featured in Meta case studies. Interviewed on several client panels. And I still can't get a single human to review my account.
Brandon Kuipers tweet media
English
123
14
510
141.4K
VEO
VEO@vrexec·
I’ve been sharing this opinion now for years. Putting money in a tax deferred retirement account is insane. Tax rates by the time you retire will be astronomically high — so high that they will offset any tax-advantaged returns come 20-40 years from now. All you need to do is see the writing on the wall. The global trend is crystal clear. Debt is out of control. Not to mention all kinds of ridiculous under-investment in infrastructure, environment, etc. I haven’t saved for retirement since 2020. Pay your taxes today and keep the liquidity… don’t let someone else tell you when you’re allowed to spend your money anyway.
Elizabeth Warren@SenWarren

Today, I'm introducing my wealth tax — and more than 50 members of Congress are joining me. It’s time for the government to start working for American families, not just the ultra-rich.

English
45
4
74
43.7K
Michelle Lim
Michelle Lim@michlimlim·
To get started: 1. Sign up for Flint at tryflint.com 2. Flint clones your design system, and pages go live on your domain. 3. Add the Flint MCP to your Claude or send an API request to start launching pages More info here: tryflint.com/product/api-mcp
English
2
0
6
1K
Michelle Lim
Michelle Lim@michlimlim·
This is the era of the one-person marketing team. Launch dozens of high-performing landing pages from Claude and Clay. Using @tryflint’s MCP and API.
English
32
26
158
56K
Peter Quadrel
Peter Quadrel@Peter_Quadrel·
NanoBanana 2 just made your static ad agency obsolete. And I just open sourced the entire tool. Drop your product page URL. It pulls your logos, product images, fonts, colors, and brand voice automatically. Builds a full brand guide for you. Then generates ad creatives at scale using nearly 4,000 high-performing ad templates across 8 niches. It dynamically matches the best templates to your brand and brief. Here's what makes it different: → Instant resizing Get any ad in 1x1, 4x5, 9x16 with one click. No regeneration. No broken text. → Highlight-to-edit See an issue? Highlight the area and tell it what to fix. → Multiple brand profiles Run different brands or segments from one tool. → Auto persona building from real customer reviews → Multiple QC loops on briefs and final assets Catches AI-isms before you do. → Upload your own templates or use ours Runs locally. Just needs your Claude and Google API keys. This is the lite version of what we use internally. You get the full finished tool AND the open source code to make it your own. Creatives still design the system, this handles iteration and scale. Want a copy to download? 1. Like this post 2. Comment "AI" Will DM you the tool along with a tutorial shortly after.
English
2.5K
240
4.4K
288.5K
Frontline Sam
Frontline Sam@DoualleSamuel·
@brivael @NicolasLefebv Meta suspend les comptes qui se connectent a Claude (ou autre) pour forcer les advertisers a utiliser Manus
Français
0
0
0
51
Brivael
Brivael@brivael·
Bon exemple concret : un système agentique Meta Ads à effets composés. L'idée c'est pas de remplacer ton expertise, c'est d'en faire le moteur d'un système qui scale. Le principe : Tu construis un agent qui tourne en boucle (quotidienne ou hebdo) avec 4 capacités : → Veille concurrentielle : il va scraper les ads libraries des concurrents, analyser les angles, les hooks, les formats qui marchent. Tu branches ça sur Perplexity ou un connecteur de recherche pour de la prospective marché. → Analyse & idéation : à partir de cette data, l'agent te propose des angles créatifs, des audiences à tester, des itérations sur ce qui performe déjà. → Allocation budget : il ingère tes métriques réelles (ROAS, CPA, taux de conversion, CPM par audience) et propose des réallocations. Pas en aveugle — avec ta validation. → Boucle de correction documentée : c'est LA clé. Chaque fois que tu corriges une décision de l'agent, tu documentes pourquoi. Cette trace devient du training data implicite. L'agent apprend ton raisonnement, pas juste tes métriques. Pourquoi ça crée des effets composés : L'IA mange la data du réel en continu (conversions, ROAS, vues, CTR). Plus elle tourne, plus elle intègre tes corrections, plus ses propositions s'affinent. Au bout de quelques semaines, t'as un système qui a ton expertise encodée + la capacité d'exécution d'un agent qui ne dort jamais. Résultat : expertise humaine × vitesse d'exécution machine = tu outplays tous les acteurs qui font encore du media buying "à la main". La cerise sur le gâteau : tu construis une interface dédiée où tu corriges les décisions de l'agent en temps réel. Chaque correction est un event tracé — pourquoi tu as changé l'audience, pourquoi tu as coupé cette créa, pourquoi tu as shifté le budget. Tu crées un vrai système de record : un historique structuré de ton expertise appliquée. C'est plus un prompt statique, c'est un système qui accumule ton savoir-faire de manière persistante. Plus tu l'utilises, plus l'écart se creuse entre toi et ceux qui n'ont pas cette boucle. Pour le stack, tu peux partir sur Claude + des connecteurs MCP pour les sources de data, un orchestrateur simple (même un cron + script), et un doc structuré pour la boucle de feedback. Pas besoin d'over-engineer au départ — le compound effect vient de la boucle, pas de l'infra.
Français
4
1
12
6K
Brivael
Brivael@brivael·
Dans 12 mois, y'aura deux types de prestas : ceux qui ont encodé leur expertise, et ceux qui cherchent un CDI. Tu fais de la presta en freelance ou via une agence, t'es pas mauvais dans ce que tu fais, mais tu sais pas encore comment intégrer des workflows agentiques dans ton quotidien ? Ce que je vais te dire, c'est probablement l'edge le plus massif que tu pourras prendre dans les 12 prochains mois. Je parle beaucoup avec des gens qui ont une expertise sur un sujet serviciel, et souvent la connaissance métier d'un domaine peut être un edge aujourd'hui. Avant d'aller plus loin, deux arguments pour ancrer ce que je vais dire derrière. Premier argument : YC a mis les "AI-Native Agencies" dans ses Requests for Startups (ycombinator.com/rfs). Leur thèse : les agences du futur ressembleront à des boîtes de software, avec des marges de software. Elles vont découpler le revenu des heures facturées en produisant à coût marginal quasi nul. Le plus gros accélérateur du monde dit "construisez ça maintenant". Deuxième argument : Foundation Capital a publié en décembre un article qui appelle ça "AI's Trillion-Dollar Opportunity" : le Context Graph (foundationcapital.com/ideas/context-…). La prochaine génération de plateformes à mille milliards sera construite en capturant ce que personne n'a jamais stocké systématiquement — les Decision Traces. Le pourquoi derrière chaque décision, pas juste le quoi. Et c'est exactement ce que t'as dans la tête. Quelqu'un qui fait du consulting, des ads, du SEO depuis des années — il a accumulé des milliers de micro-décisions. Il sait instinctivement quel angle créa va marcher, quel budget allouer, quel positionnement choisir. C'est 10 ans de knowledge et de wisdom compressés dans son intuition. Le problème c'est que cette intuition, elle scale pas. Elle est enfermée dans sa tête. Il vend son temps, il est le bottleneck. Le move c'est pas "utiliser l'IA pour aller plus vite". C'est d'industrialiser cette expertise artisanale. Et voilà comment ça marche concrètement. Tu commences par créer un système agentique. Au début, l'agent prend des décisions vanillées — génériques, pas terribles, comme un junior à qui tu viens de filer le brief. C'est normal. Ensuite, tu viens corriger chaque décision. Tu lui expliques pourquoi il aurait dû choisir cet angle plutôt qu'un autre, pourquoi ce budget est trop agressif pour ce type de client, pourquoi ce positionnement marche pas sur cet ICP. Tu log le contexte, ton raisonnement, le résultat attendu. Et tu fais ça en boucle. Ce qui se passe, c'est que tu distilles progressivement tes compétences et ton knowledge dans le système. Chaque correction l'affine. Chaque exception traitée devient un précédent. Tu construis ton propre Context Graph — sauf que le tien est nourri par 10 ans de terrain, pas par des données génériques. Un système conçu pour UN client précis, avec SON marché, SON ICP, SES contraintes. Au bout de quelques semaines, le système prend des décisions que t'aurais prises toi-même. Et il les prend sans oublier, sans fatiguer, en continu. Résultat : ce que tu faisais de manière artisanale pour 3-4 clients, tu le fais maintenant pour 15-20, avec une qualité égale voire supérieure. Tu industrialises ton expertise sans la dégrader.
Français
5
8
70
20.5K
Frontline Sam
Frontline Sam@DoualleSamuel·
@PhilKiel what's the difference between Google referral and branded search in this context?
English
1
0
0
134
Phil Kiel
Phil Kiel@PhilKiel·
We're still speaking to brands spending $500k - $1mil per month who aren't: - measuring funnel impact. - running TOF campaigns correctly. Two slides we give away as a starting point 👇
Phil Kiel tweet mediaPhil Kiel tweet media
English
6
1
22
3.3K
Pyrate
Pyrate@CEOLandshark·
@RexSteps For "masculine shadow work" - there's none that I know. That's what I'm trying to point out
English
3
0
6
452
Pyrate
Pyrate@CEOLandshark·
Think of something your nemesis/competitor would do that would crush you, leaving you feeling like "how couldn't I see it" or feeling completely snubbed, etc. And then find a way do/be it. This is real self-actualization shadow work.
English
6
13
242
13.2K
AE
AE@scalingszn·
Most ecom founders & affiliates are still just testing AI tools, without understanding how to properly leverage LLMs to turn context into cash. Context profiles aren’t just a branding exercise. They’re one of the most valuable assets you can build & leverage in your business, if utilized properly. Here’s how to use them across 4 profit levers to drive more conversions & lower CPAs, all without guesswork: 1. Develop Ad Creatives That Actually Convert Generic UGC won’t cut it anymore. You can map buyer psychology into a “hook + angle bank” based on: • Internal struggles • Vanity desires • Skepticism/objections • Functional outcomes From there, generate scripts that sound exactly like how your ICP thinks, using real customer language as the foundation. → “I used to wake up with tension headaches, here’s what fixed it in 3 days.” → “I thought X were a scam until I tried this...” Those nuances matter. And context profiles bring them to the surface. 2. Email Flows That Sell Without Actively Trying To Sell We use context profiles to build email sequences rooted in problem-first storytelling. → “Meet Taylor. She was in pain for 8 years…” → “Why 14,000 people like you switched from X to this.” By basing these on objections and emotional triggers pulled from ICP docs, your post-purchase and winback flows can consistently generate more revenue. We’ve seen it work for us. Try it out. 3. Support Scripts That Drive Retention Support isn’t just about solving problems. It’s a hidden lever for brand trust & long-term LTV. We tag past tickets by ICP archetype and generate AI-powered replies with tone, product knowledge, and empathy built in. → Returns become exchanges. → Confusion becomes upsell opportunities. → Tickets become testimonials. 4. Acquisition Assets That Prime Cold Traffic to Buy Using ICP profiles, we’ve structured VSLs, advertorials, and listicles that feel native to the customer: → “5 Tools That Helped Me Fix My Posture After Years at a Desk” → “Top 7 Wellness Products Women Over 40 Swear By” Context lets us soft-sell the product by selling the transformation first, boosting CVR on cold traffic significantly in most of the verticals we’ve tested. LLMs aren’t just for shortcuts. They’re for scaling outputs that speak to your exact target, but only if you give the inputs that matter. Want our full ICP Template + plug-and-play prompt stack to generate creatives, emails, support flows & landing pages? Like this post & comment “System” and I’ll DM it to you.
English
156
8
265
24.3K
Frontline Sam
Frontline Sam@DoualleSamuel·
A clear limitation of using ChatGPT as a tool to understand my blind spots and psyche is that _besides telling me what I want to hear, it is not designed to ask questions back. Sometimes, the best way to reach a better understanding is to have first my questions challenged and reframed, not directly answered.
English
0
0
0
191
Frontline Sam
Frontline Sam@DoualleSamuel·
The first AI cult might look a lot like the church of scientology: - start the onboarding by asking for a confession and provide a very personalized audit of the prospect’s emotional or spiritual state. - present a final state of being to reach - sell a personalized solution to get there, in levels. - create a ranking, ideally vs other members - create a full universe with its own vocabulary, founder story, myths, avatars, tasks and rewards - use elements from the initial confession to both play on the member’s shame and guilt as well as his/her aspiration to level up to guarantee retention - monetize every steps, until the member reaches the never-ending final level
English
0
0
1
179
Frontline Sam
Frontline Sam@DoualleSamuel·
Random thoughts on how AI will impact knowledge workers: - We will keep valuing work that is the result of effort, energy applied and creativity. Anything created by AI without a clear human touch will be considered 2nd class. - A contradiction of human nature is that we crave for efficiency but also tend to create new inefficiencies. Therefore as jobs are destroyed, human organizations will find new ways to employ people, even just for the sake of it. - The corporate managerial class will need to hire humans to justify their own job: if your job consists only in supervising the output of AI tools, it becomes obsolete. - Yappers and storytellers will remain untouched. Operators will become orchestrators. A personal rolodex will be more valuable than ever. - We will keep valuing and paying for human expertise for any activities impacting directly other humans (art, sales, marketing, coaching, schooling, branding…). - Humans are addicted to humans. Just like we can both enjoy fictional stories vs real life stories when going to the movies, a % of the workforce will be comfortable working mostly with AI agents while others will always tend to prefer working with humans
English
0
0
0
134
jacob sappington
jacob sappington@jsappington·
I’m looking to hire an online fitness coach…who should I be talking to? Looking for someone who doesn’t follow cookie cutter setups, who knows and understands fitness and nutrition, and has done client work before. Goals: Mild fat loss followed by lean bulk
English
39
0
20
7.3K
MONTE
MONTE@fromzerotomill·
There’s a guy in Romania Runs a faceless AI +18 disney IG profiles Auto-generates content with Midjourney + GPT Voice with ElevenLabs $300K/month Zero human contact Just ChatGPT, Stripe, and a MacBook Reply w/'AI' and ill send you guide on how to do it
English
876
66
1.3K
116.9K
Sim
Sim@SimScaler·
You don’t need “better creatives" You need a CREATIVE SYSTEM The guys printing $1M+/month aren’t just filming random UGCs They’re building ad factories Launching 10+ variations/week Iterating based on data Organizing segments like LEGOs I just made a guide breaking it all down RT + Comment “CREATIVE” and I’ll send it (Must be following so I can DM)
Sim tweet media
English
472
248
458
49.2K
Jeddi
Jeddi@antinertia·
i have a Tiktok AI farm making 10M views while i sleep here are the steps - create 200 ai ugc with arcads.ai - distribute on 20 accounts over 5 days - all of them get hundreds views - that’s a guaranteed 100,000 views - 2% will go viral and get 1M+ views it's all on autopilot comment "ai ugc farm" to get the full guide
English
1K
105
1.2K
178.6K
Cas.Fyn
Cas.Fyn@FynCas·
I don’t go viral because I’m lucky I go viral because I post 100+ times a day using AI Quantity beats everything It makes me $1,500 a day and the best part is it costs less than a BigMac Comment + RT “MAC” and I’ll DM you the exact system
Cas.Fyn tweet media
English
236
111
211
29.3K
Kody Nordquist
Kody Nordquist@KodyNordquist·
This company (NewForm.ai) just launched the first AI Performance Marketing Agent. They’re calling it WillBot. WillBot is like having a media buyer, creative strategist, and data analyst all in your pocket. How it works: WillBot syncs across your ad platforms (meta, tiktok, google, etc) and your other marketing data sources With a single prompt, it can analyze trends, generate charts, and provide full reports synthesizing data across platforms and data sources On top of that, WillBot can watch your ads. WillBot scours your entire data stack, finding winning scripts and can generate unlimited high-converting scripts in any language, for any demographic. WillBot lets one marketer do the work of 20. WillBot operates on natural language, just like chatgpt. Ask it anything you'd ask a human, and it’ll deliver answers within minutes. Most AI marketing solutions SUCK. It’s because LLMs are trained on random data that does not respect your brand voice and has no regard for performance. WillBot understands not just what’s working for marketers, but what’s working for YOU. Traditional solutions charge $10k+/month for: Ad performance analytics, campaign/creative strategy, creative optimization and iteration, ad translation/cultural adaptation, etc... WillBot does this for 1/30 of the cost. The era of agentic marketing looks to be finally here… If you made it this far, I’ve got a gift for you The team behind willbot asked it to analyze 100m+ worth of ad spend and create a full report on what is working right now. It includes: 50 best performing hooks (& the top ad types for each industry) Most optimal meta & tiktok campaign/account structure Cherry on top: 1000+ UGC creator database Comment WILLBOT and I’ll send you the link.
English
335
43
440
67.4K