Kickoff: Monday, June 1 at 4pm EST · Live on Skool
49 of 100 seats remaining
Live + recorded
50+ SOPs included
$199 all-access
Recordings included
AGENTIC AGENCY · 5-DAY LIVE CHALLENGE · JUNE 1-5
Build a Meta Ads Agent
Trained on a proprietary formula of SOPs & data
10+ years of ad playbooks loaded into the under-the-radar agentic harness most of your competitors don't know about yet.
AI That Actually Runs Your Account.
Show up Monday at 4pm EST. Live on Skool with Chris, Ryan, and the cohort. Build alongside us, one workflow per day.
Ad production. 30 ad variations already drafted from the SOPs and winning creative in your own account — tighter than what you'd ship after three coffees and a Sunday night push.
Ad publishing. The launch queue pre-flighted. Every campaign named correctly. Every UTM verified. Every conversion event firing — before you click publish.
Optimization & learning. The agent reads its own ad data. Yesterday's winners feed tomorrow's variations. Yesterday's losers don't get a Round 2. The system gets sharper.
Analytics. Last week's reports in your inbox. Performance flagged. Anomalies called out. The three numbers that actually matter highlighted — not the 47 your reporting tool spits out by default.
Client management. This week's client update already drafted — performance summary, what shipped, what's coming next. Talking points for your Tuesday call pulled from the data. Ready for your sign-off.
You finally get your time back for the work that actually moves the needle. High-level strategy.
This is an actual operator. Always on. Already trained on the playbooks. It's what we run today. It's what you'll build in 5 days.
Most AI Agents Fail for the Same Three Reasons.
Maybe you've already tried one yourself. Maybe you've been waiting for one that actually works. Maybe you've been watching AI agents from the sidelines... Either way, they often break for the same three reasons.
1
It doesn't know media buying.
Generic agents have zero context on how real ad work actually gets done. Their outputs look polished and fall apart the moment you QA them.
2
It breaks, and it's annoying to manage.
Most agentic software was built for engineers writing code, not for operators navigating Meta. So it stalls. Then you spend your morning fixing it instead of using it.
3
It can't actually run on its own.
You're still the one doing the work — just with extra steps. Prompts to write. Outputs to check. Corrections to make. The "agent" is really just you, plus typing.
Get all three right — at the same time — and you have what we built.
Built on Perplexity Computer (Enterprise Max). Trained on a Proprietary Formula of 50+ SOPs.
Each of the three problems above has a specific fix. They have to land at the same time for the agent to actually run.
01
Perplexity Computer is the harness.
Built by the Perplexity team, but designed for operator work — not search, not coding. It navigates real browser interfaces the way a person does: clicking, dragging, typing, working through Ads Manager. That's why it doesn't fold the moment it touches Meta.
02
The proprietary formula: 50+ SOPs, 10+ years of direct-response playbooks, and $100,000+ in real ad spend.
The agent isn't trained on the internet. It's trained on what we've actually used across our portfolio — direct response, lead gen, ecom, B2B SMB. Every output is grounded in real ad work, not generic best practices.
03
The 5-day build gives it autonomy.
You configure the agent once — your accounts, your SOPs, your preferences. After that it runs on its own. You direct strategy. The agent does the work. No prompts to write. No outputs to babysit.
By Friday you'll walk out with that exact stack, configured around your accounts and your SOPs. Here's how we get there, day by day.
It's not "AI for agencies." It's Agentic Assets.
A repeatable way to turn the SOPs you already have into agent skills, and the agent skills into recurring crons that compound while you sleep.
Workflows
The verbal or written processes you already run. You probably have most of these. We just write them down once.
Skills
Workflows trained into the agent. It loads them and works like an expert who actually ships.
Crons
Recurring leverage. Workflows run on a schedule. The work shows up before you do.
Workflows become Skills. Skills are used in Crons. The crons compound.
What Changed When We Switched.
Real numbers. Real client accounts. The same playbook we're handing you in 5 days.
+20%
Margin
SaaS + process replaced
7
Offboarded
Part-timers replaced with skill files
1,000+
Ad Creatives
Produced in 30 days, on autopilot
$100K
Managed Spend
Across client accounts, month one
The 5-Day Curriculum
5 days. 5 builds. Monday through Friday, live.
Day 1 — Monday, June 1 @ 4pm EST
Foundation & Setup (1.5-2 Hours)
No slideware. No theory. We're going to install Perplexity Computer in front of you, load the SOP library, and wire the agent into a real ad account live. Then we'll teach you the paradigm: Workflows → Skills → Crons.
You'll learn:
- Setting up Perplexity Computer for operator work (the no-developer install)
- Connecting Meta Ads Manager — your account or a client's, your choice
- The Workflows → Skills → Crons paradigm: how the agent actually compounds
- Loading the proprietary SOP library into the agent's context
- Configuring permissions and safety guardrails (read-only first, then execute)
- Naming conventions and file structures that don't break under load
- The exact stack — what it costs, what's free, what's required
You'll leave with:
A fully configured Perplexity Computer instance, SOP library loaded, and the agent capable of reading your Ads Manager — ready for Day 2's production work.
Day 2 — Tuesday, June 2 @ 4pm EST
Ad Production at Scale (1.5-2 Hours)
The agent learns to produce ads today. Video workflows, static workflows, and the 45+ creative frameworks we've battle-tested in real client accounts. By the end of the session, you can hand it a brief and get 30 QA-ready variations back.
You'll learn:
- The ad production workflow: brief → research → variation → QA
- Live build: Video ad production agent
- Live build: Static and graphic ad production agent
- Training the agent on the 45+ creative frameworks
- Generating 30+ variations from a single brief without sacrificing quality
- Quality acceptance criteria — what passes review, what gets rejected
- The handoff format for your team or client
You'll leave with:
A working production agent that takes a brief and returns QA-ready video and static ad variations matching your account's proven winning patterns.
Day 3 — Wednesday, June 3 @ 4pm EST
Publishing & Launch Automation (1.5-2 Hours)
Publishing is where most agents fall apart. UTM hygiene, naming conventions, pixel verification, conversion events. Today we wire the agent to handle the entire pre-flight and launch sequence — and we set the first cron so it runs on a schedule.
You'll learn:
- The publishing pre-flight: UTMs, naming, pixels, conversions — what to check, in what order
- Live build: Meta publishing agent
- Automated UTM template generation and validation
- Conversion event verification — the agent confirms the pixel fires before launch
- Launch queue management: what ships, what waits, what gets killed
- Setting your first cron: scheduled publishing without supervision
- Rollback procedures for when something does go wrong
You'll leave with:
A publishing agent that pre-flights every campaign automatically and only ships when everything's verified. No more 10pm pixel disasters.
Day 4 — Thursday, June 4 @ 4pm EST
Quality Assurance & Performance Reporting (1.5-2 Hours)
The agent should be your eyes on the account 24/7 — flagging issues, surfacing wins, producing reports clients actually read. Today we wire that up and put it on a recurring cron.
You'll learn:
- QA automation — creative checks, brand rule enforcement, targeting verification
- Live build: Performance monitoring agent
- Anomaly detection and alerting — CPL drift, CTR drops, spend overruns
- Live build: Weekly reporting agent on a cron
- The three numbers that matter — and the 47 you don't need to send clients
- Custom dashboards: building once, viewing forever
- Audit logs for everything the agent did this week
You'll leave with:
An always-on QA and analytics agent producing weekly client-ready reports automatically, with the metrics that actually move the conversation.
Day 5 — Friday, June 5 @ 4pm EST
Client Operations & Going Live (1.5-2 Hours)
Final day. We connect everything — client communication, weekly cadence, talking points, and the full operating system. You leave with a system that runs Monday morning.
You'll learn:
- Live build: Client communication agent — weekly updates, talking points, status reports
- Wiring up notifications — Slack, email, SMS — so you know what's happening without checking
- The full weekly operating cadence — what runs when, what you touch, what you don't
- Handing the agent off to a junior or VA (if you want)
- Troubleshooting and edge cases — what to do when the agent gets confused
- Your action plan for week 1 post-challenge
- Live Q&A: how to handle your specific accounts, clients, and stack
You'll leave with:
A fully-operational AI media buying agent, configured around your accounts and your SOPs, ready to run Monday morning of next week.
Ready to build it?
$199
Bring your laptop. Leave with a working agent.
00
Days
00
Hours
00
Mins
00
Secs
Who This Is For. And Who It's Not.
You're a strong fit if you:
- Run an agency billing clients for Meta media buying
- Buy in-house for an ecom or DTC brand
- Freelance for clients and have hit the solo ceiling
- Are earlier in your career but already running real campaigns
This isn't for:
- People who have never run a Meta ad. Start with Ryan's free YouTube training first.
- People who think AI is a fad. Close this page. Save us both the time.
- People who want a beginner-level "intro to AI." This isn't that.
- People who buy courses and don't show up. The challenge is live, the value is in the live build.
You're Not Deciding Whether to Invest in AI.
You're deciding whether to keep paying for the absence of it.
Four costs you're already paying — every week, no matter what:
Overseas labor
Time zones, language gaps, mistakes baked into client work before you catch them.
Inconsistency
Simple requests take three rounds and 48 hours to come back wrong.
SaaS sprawl
Stacked monthly subscriptions wrapping LLM prompts you could run yourself.
Hours in Ads Manager
Manual publishing, manual QA, manual reporting — every week, forever.
The ones who feel none of this in 12 months built their own leverage. The rest are still buying it from someone else.
$2,300 of Assets. $199 to Take Them Home.
Here's what's in the box:
5-Day Live Training
Normal value: $500Live training on configuring Perplexity Computer to run lead generation advertising operations and complex media buying workflows.
50+ Data-Backed SOPs, Playbooks & Process Documents
Normal value: $1,000Loaded into your agent's context — covering ad production, publishing, QA, analytics, and client management: 20+ static image ad skills, 20+ video ad skills, media buying workflows ready to customize for your accounts.
Training Series on Production & QA
Normal value: $500Implementing Meta publishing routines, video ad production, graphic and static ad design, and client quality assurance.
Live-Only Bonuses
Normal value: $300Reserved for attendees in the room during the live cohort. Giveaways and prizes throughout the week. Bonus assets the recording-watchers don't get.
The Trained System
IncludedThe same agent that produced $100K+ in managed ad spend, +20% margin, 1,000+ creatives, and 7 offboarded contractors in month one.
Normal value: $2,300
Your price: $199
Savings: over $2,100 off industry-standard prices.
$199. The math is the math.
$199
Refundable if you show up and don't find it worth it.
00
Days
00
Hours
00
Mins
00
Secs
Questions Before Monday.
Yes — you'll need a Perplexity account on the Enterprise Max plan active before Monday, June 1. You don't need it the day you buy, but you need it by Day 1. We send setup instructions with your registration confirmation.
You'll still get the recordings, the SOP library, and the build files. But the Live-Only Bonuses are reserved for attendees in the room, and the value of the challenge is the live build with us. If you're going to miss more than 1-2 sessions, this probably isn't the right cohort for you.
If you can run a Meta ad account, you can do this. We're not coding — we're configuring. The technical work is Perplexity Computer's job, not yours.
Fine. The system scales down as well as up. One account, saving 15 hours a week, is still real ROI.
No. The SOPs are platform-agnostic — they're workflows, written down. You can use them with a VA, a junior buyer, or any other agent platform. Perplexity Computer is the leverage layer that runs them autonomously, but the SOPs themselves stand alone.
See the Show Up and Build Guarantee. Attend all 5 sessions, build alongside us, and if it wasn't worth $199 by Friday, email for a full refund.
Yes. You'll get a paid invoice at checkout — most appropriate category is "professional development" or "marketing automation."
Your instructors
Chris Blakeman
Lead Instructor
Leads the live build. You'll configure the agent, wire your ad account, and ship workflows that run real client work — step by step, on screen.
Ryan Magdanz
Co-Host
Co-hosts the challenge live on Skool — keeping sessions on track, answering questions, and making sure you leave Friday with a working operator.
See you Monday
$199
Live with Chris, Ryan, and the cohort. Mon June 1.
00
Days
00
Hours
00
Mins
00
Secs