Founder @Blue Water Marketing | eCommerce & Digital Marketing Nerd

Hey, I'm Chris Marrano.

I've spent that last 15 years in digital marketing. Started at Mullen in Boston running campaigns for Match.com and LendingTree. Went client-side at Harvard Business School. 
Then built Blue Water Marketing from scratch.

BWM is an eCommerce growth agency that manages Meta, Google, and Klaviyo for Shopify brands scaling from 7 to 8 figures. But somewhere along the way, the agency became a lab. 

We stopped just running ads and started building what goes behind them. AI systems that monitor every client account 24/7. Automated creative briefs generated from real performance data. 

A learning algorithm that gets smarter with every human decision. Real IP that's always running in the background.

We turned an agency into a tech company that runs ads.

That thinking became ADIQ.AI. AI-powered creative intelligence that tells DTC brands what's working before their budget does.

And after building all of this publicly on LinkedIn, the same question kept showing up in my DMs from founders and agency owners: "How do I build this for my business?"

Now I help them do exactly that. Private consulting. Full infrastructure builds. And a community of eCommerce operators who are done guessing and ready to build real systems. 

I help brands and agencies stop renting results and start owning their own IP, their own data layer, their own competitive moat.

The future of eCommerce isn't more ads. It's better infrastructure. I'm here to help people build it.

South Florida. Family first. Beach days, t-ball games, and a goldendoodle who supervises everything I build.

Agencies charge $3K-$5K for a creative strategy brief. I built a Claude skill that builds one with you in 20 minutes.

And I am giving it away.

Here is the problem with most creative strategies:

They produce 20 ads that look different to humans but identical to Andromeda.
Meta's visual recognition groups similar creatives into the same Entity ID. 

Your "fresh batch" of 15 ads registers as 2 or 3 concepts in the system.

That is why your CPMs keep climbing even when you launch new creative every week.

We solved this with a framework called P.D.A.

Persona times Pain times Desire times Awareness.

I turned it into an interactive Claude skill. But here is what makes it different from every other AI tool:
- It does not just generate and dump output.
- It thinks with you.
- It asks your opinion at every step. 
- It presents persona candidates and waits for you to pick. 
- It shows you the pain points it found and asks if they match what you are seeing. 
- It builds the matrix and lets you add, remove, or reshape before it writes a single ad.

Your expertise stays in the loop. The skill just moves faster than a whiteboard session.

Here is what the workflow actually looks like:
→ You paste a product URL. It extracts positioning and asks you to confirm. 
→ It generates 4 to 6 persona candidates. You select the ones that match your customers. 
→ It maps pain points and desires. You approve, adjust, or redirect. 
→ It builds the full P.D.A. Matrix. You pick which concept cells to develop. 
→ It writes 5 ads per concept across all awareness stages. You approve each set before it moves on. 
→ It checks every ad against Andromeda diversity requirements.

The whole thing is a conversation. Not a prompt and pray.

It is the same engine we use to plan creative sprints for 7 and 8 figure DTC brands.

Comment CREATIVE and I will send you the skill.

Chris Marrano

Right now everyone is trying to crack AI static ads the same way.

JSON prompts. 4 different LLMs. Copy-pasting between tools. Tweaking parameters. Debugging hands. Fixing garbled text. Spending 45 minutes on one ad that still needs a designer to clean up.

I know because I built that exact workflow.

I shared it publicly. 859 people wanted it.

And then I stopped using it.

Here is what most people are doing right now:
1. Write a detailed JSON prompt with product specs, persona, angle, layout. 
2. Feed it to GPT or Gemini for the image. 
3. Realize the text is wrong. 
4. Open Canva or Figma to fix it. 
5. Try again with a different model. 
6. Repeat 10 more times.

6 steps. 4 tools. 45 minutes. One static ad.

Here is what I do now:
Pick a product. Pick a theme. Pick an angle. Click generate. Launch.
3 clicks. Done.

No JSON. No prompt engineering. No bouncing between tools.

The strategic thinking is already built into the system. 

Hooks. Layouts. Visual hierarchy. What actually converts on Meta. 

Trained on what we have learned managing millions in ad spend across 150+ brands.

This is the difference between engineering a prompt and engineering a system.

One makes you feel productive. The other makes you profitable.

I put together a breakdown showing exactly how the old prompt workflow compares to the 3-click system we use now.

Comment "STATIC" and I will send it over.

Meta just made two announcements that most brands are going to miss.

Detailed targeting is now mostly a suggestion.
ASC and AAC campaigns are being phased out entirely by September 2026.

Read those together and there’s a clear pattern.
Meta is systematically removing every manual control lever advertisers have relied on for the last decade.

Audience targeting. Campaign structure. Budget allocation.

All of it is moving into the machine.

And the brands that are going to struggle the most are the ones who built their entire strategy around controlling those levers.

Because when the lever disappears, so does the strategy.

Here’s what that leaves you with.
- Your offer.
- Your creative.
- The strength of your signal.

That’s it. That’s the whole game now.

The agencies still arguing about ad set structure and audience stacking are having the wrong conversation entirely.

Meta has already decided that conversation is over.

The only edge left is what you put in front of people and whether it makes them stop.

Everything else is just settings.

What are you seeing in your accounts since targeting started getting overridden?

Two announcements dropped last week that most marketers scrolled past.

They shouldn't have.

Manus now has an Ads Manager integration. An AI agent that builds reports, researches audiences, and analyzes your campaigns directly inside the platform.

Three days earlier, Google launched Pomelli Photoshoot. Upload one product photo and their AI generates studio-quality lifestyle images, product shots, and marketing assets. For free.

Read that again.

Meta's ad ecosystem now has an AI analyst. Google's ad ecosystem now has an AI photographer.

Both platforms are quietly building the same thing: an end-to-end advertising ecosystem where you never need to leave.

Audience research. Creative production. Campaign analysis. Reporting.
All inside the walled garden. All powered by their AI. All designed to make outside tools feel unnecessary.

The question nobody is asking...

When the platforms that sell you ad space also build the creative, analyze the performance, and recommend the next move…

Who is actually running your brand's marketing?

Is digital advertising becoming a true monopoly disguised as a free tool?
What do you think the next 12 months look like?

What the f*ck is happening?! Shopify just quietly made half the ecommerce tech stack irrelevant.

Their Winter ’26 Edition dropped 150+ updates. Most people skimmed the headline and moved on.

That was a mistake.

Here is what actually matters and why it changes everything for brands and agencies.

1.	Sidekick is no longer an assistant. It is an operator.

It builds custom apps from a text prompt. It creates automation workflows in Shopify Flow. It edits your theme without touching code. It removes backgrounds, expands canvases, and generates lifestyle scenes from your product photos. All inside the platform.

The $10/month micro-SaaS apps that solved one small problem? Their value proposition just went to zero.

2.	Your products now sell inside AI conversations.

Agentic Storefronts let you show up directly in ChatGPT, Perplexity, and Microsoft Copilot. One setup in admin and your catalog is discoverable wherever AI shopping happens.

This is a brand new sales channel. And most brands have not even turned it on yet.

3.	You can now simulate buyer behavior before going live.

SimGym uses AI trained on billions of real Shopify purchases to predict how your changes will impact conversion, AOV, and add-to-cart rate. Before a single real customer sees them.

Combined with native A/B testing through Rollouts, you no longer need three third-party tools just to test a homepage layout.

4.	AI image editing is built in.

Background removal. Canvas expansion. Composition adjustments. Lifestyle scene generation. Studio-quality product shots. From your phone.

The brands that were spending $500 per product shoot can now iterate on creative at near zero marginal cost.

Here is what this means for the next 12 months.
The gap between brands that adopt these tools and brands that ignore them is going to widen fast.
Not because the tools are complicated. Because they are simple. And the brands that move first will compound that speed advantage every single week.

For agencies, the implication is even bigger.
If your value is execution, Shopify just automated most of it.

The agencies that survive will be the ones providing strategic direction, creative systems, and performance frameworks that AI can plug into.

Not the ones charging retainers to swap backgrounds and build flows.

The platform is not waiting for anyone to catch up.
Neither should you.

undefined undefined

Building AI-Systems For eCommerce | Founder@ADIQ.AI | Founder@BlueWaterMarketing

12 months ago, we were a traditional ecom agency.

Manual reporting. Copy-paste metrics into slides. Strategy calls where half the time was spent explaining what happened last week instead of what to do next.
Today, we're a tech company that runs ads.

Every Monday morning, our clients get a strategy report before anyone opens their laptop. New creative concepts. Performance trends. Visual analysis of what's working and why.

Built entirely by the automation infrastructure our team designed.
Vector stores that ingest and chunk massive amounts of client data. 

AI that watches video ads and classifies images by structure, tone, and audience. Systems that pull top performers against 90-day baselines to spot real trends.

Then it all merges into a PDF that doesn't just show numbers — it drafts new creative concepts and headlines based on what the data is telling us.

The piece most people get wrong when building this stuff: memory.

If you don't force the system to check against previous concepts, it keeps suggesting the same ideas. That kills creative performance. We built a layer specifically to prevent it.

But here's what I want to be clear about, he tech didn't replace our team.
It made them dangerous.

Our strategists spend Monday mornings reviewing AI-generated insights instead of building reports. Our media buyers focus on decisions, not data pulls.

The team got better because the system handles the work that was slowing them down.

Traditional agency → tech company with amazing humans. That's been the 12-month shift.

I recorded the full technical walkthrough showing exactly how we built it. Vector database, visual analysis workflows, report generation , all of it.

Link in comments.

What would your team do with an extra 5-10 hours a week back?

The founders scaling fastest right now are not better at Meta ads.

They are better at seeing what is happening inside their ad account without depending on anyone else.

Claude Code is a tool that runs on your computer. You talk to it in plain English. It reads your files, your exports, your data, connects to your Meta account. No dashboards. No waiting.

Here is what that looks like for a Shopify brand running Meta ads:
→ Connect via your Meta developer account via access token.
→ Ask it questions the way you would ask your strategist. "Which creatives are above our CAC target this week?" "Where is frequency spiking?" It answers with your actual numbers.
→ Build reusable audit frameworks so every review follows the same standard. 

No drift between analysts. No blind spots.

This does not replace your team. It gives you visibility you have never had before.

The founders who understand their own data make faster decisions. Faster decisions compound.

Have you tried this yet?

If you want to try this, leave a comment down below or shoot me a DM!

Most brands are running 2024 creative strategy on a 2026 algorithm.

And it is costing them everything.

Here is what changed.

Meta’s Andromeda retrieval engine now assigns every creative an Entity ID.

Think of it as a fingerprint.

If two ads share similar visuals, similar structure, similar composition, Meta groups them under the same Entity ID.

Same Entity ID means they share performance data. They compete for the same delivery. They reach the same audience pockets.

So when you launch 30 ads that are all the same product shot with different headlines?

Andromeda sees one ad. Not thirty.

This is why brands scale spend and see CPMs climb with zero improvement in reach. The algorithm thinks you are showing the same thing over and over.

It gets worse.

Meta just rolled out three new metrics: Creative Fatigue Score, Creative Similarity Score, and Top Creative Themes.

The Creative Similarity Score is essentially telling you how many of your ads the algorithm considers duplicates.

High similarity = higher CPMs. The system penalizes repetition.

And with GEM now running underneath the ranking layer at 4x the efficiency of the old models, creative fatigue is hitting in 1 to 3 weeks instead of 4 to 6.

The old playbook of iterating on a winner is dead.
Color swaps do not create a new Entity ID.

Headline changes do not create a new Entity ID.
Same UGC creator in the same setting does not create a new Entity ID.

What does?

Different concepts. Different formats. Different environments. Different creators. Different angles that the algorithm can file into separate branches of its index.

Your ad library should look like a film festival.
Not a casting call.

The brands winning right now are the ones feeding Andromeda 10 to 20 genuinely distinct creative concepts inside simplified campaign structures.

One campaign per objective. Broad targeting. Let the creative do the targeting.

Because in 2026 your creative IS your targeting.
We have been building our entire system around this shift for months, both at Blue Water and inside ADIQ.

If your CPMs are climbing and your reach is flatlining, do not touch your targeting settings.
Audit your creative library first.

How many unique Entity IDs do you actually have?
That is the question most media buyers are not asking yet.

Everyone is talking about Meta's GEM like it just dropped out of nowhere.

It didn't.

GEM has been quietly running since Q2 2025. If you manage real ad spend, you already felt the shift before anyone gave it a name.

Here is what GEM actually is in plain English.

It is Meta's central brain for ad delivery. One unified model trained on billions of user behaviors across Facebook and Instagram. It replaced a bunch of smaller systems that used to operate in silos.

Here is what that means for your ads.

Andromeda scans your creative and shortlists candidates. GEM ranks them and decides who sees what. Lattice sorts the final delivery.

Three systems. One pipeline. And your creative is the only input you control.
This is the part most people are getting wrong.

GEM does not reward more ads. It rewards more different ads.

If you upload 30 variations with the same background, same creator, same layout — GEM sees one ad. One Entity ID. One ticket into the auction.

The other 29 never get ranked.

The brands winning right now are not the ones producing the most creative. They are the ones producing the most diverse creative.

Different formats. Different hooks. Different visual styles. Different psychological angles.

That is not a volume game. That is a strategy game.

Meta's own data showed a 5% lift in Instagram conversions and 3% on Facebook Feed after GEM rolled out. But those are averages. Some accounts saw 15% gains. Others saw zero.

The difference was creative diversity.

One more thing most people are missing.

GEM learns from organic behavior too. Your Instagram posts, your Reels engagement, your page interactions — all of it feeds the model.

Which means your organic content is now training your ad delivery system whether you realize it or not.

The playbook has not changed as much as the headlines suggest.

Creative is targeting. Diversity is distribution. Signal quality is the edge.

GEM just made the rules harder to ignore.

What are you seeing in your accounts since the shift?

Everyone’s chasing Claude skills and Cowork tutorials right now.

Building custom GPTs. Sharing OpenClaw repos. Posting about their latest prompt.

Cool. But that’s not where the real value is.
The real value is using these tools to build actual infrastructure for your business.

We just finished an internal app called BWM Systems. It runs our entire agency.

Here’s what’s inside:
Agentic daily recommendations running 24/7 analysis through a worker.ts deployed on Render. It never stops. It never sleeps. Every morning our team wakes up to fresh insights on every account.

Creative strategy mapped to dynamic matrices built on our database structure. It identifies top performing angles, hooks, tone of voice, and visual patterns across every client. Not gut feelings. Real data.

A bulk ad launcher that pushes directly into Meta through the API. Creative and copy sitting in a queue, ready to go live the second a client approves assets in ClickUp.

One app. One system. No more jumping between 3 different LLMs to get work done.

The people winning with AI right now aren’t the ones with the best prompts.

They’re the ones building systems that compound.
Skills are a starting point. Infrastructure is the destination.

If you want to see behind the scenes of how we built this, drop “BTS” in the comments and I’ll show you.

Block just cut 4,000 people.

Almost half the company. Gone.

Gross profit? Up 24%.

Stock price? Up 24%.

Wall Street? Standing ovation.

Jack Dorsey didn’t blame a bad quarter.

He didn’t blame the economy.

He said the quiet part out loud:
“A significantly smaller team, using the tools we’re building, can do more and do it better.”

Then he said something every operator should screenshot:
“Within the next year, I believe the majority of companies will reach the same conclusion.”

This is not a tech story.

This is an infrastructure story.

Block did not wake up yesterday and decide to use AI. They spent years building internal tools, training teams on new workflows, and restructuring how decisions get made.

The layoff was the outcome. The infrastructure was the work.

And that is the part everyone skips.

I talk to ecommerce founders every week who want to “use AI” but have no system underneath it.
No automated reporting.

No creative analysis pipeline.

No way to go from data to decision without five people in a Slack thread.

They hear about Block and think the answer is fewer people.

The real answer is better infrastructure.

We have spent the last year building exactly this across our Meta ad accounts. AI agents pulling performance data in real time. Creative analysis running automatically. Decisions happening in hours instead of days.

Our team is small. Our output is not.

And we are now helping other brands and agencies build the same thing.

If you are running a seven or eight figure brand and you know your team needs to operate differently but you do not know where to start, that is exactly what we help with.

Comment “INFRA” and I will walk you through what we have built and how it could work for your team.

Everyone is using AI to manage their ads now.

And most of the recommendations it gives are garbage.

Not because the AI is bad.

Because there are no rules feeding it.

AI without guardrails is just a very confident guessing machine.

Media buying is not “ask the robot what to do.”

Media buying is a skill set. It requires understanding metrics, timeframes, and the actual math behind the business.

Here is what I mean.

If your product costs $22 to make, ships for $8, and sells for $65, your contribution margin is $35.

That means your maximum CAC to break even on first purchase is $35.

Factor in a 2.2x 90-day LTV and now your allowable CAC shifts to roughly $77.

That single calculation changes every decision you make.

Which campaigns to scale. When to kill an ad. How aggressive your cost caps should be.

AI does not know your COGS.

AI does not know your shipping costs.

AI does not know your return rate is 14% on that one SKU.

AI does not know your cash flow requires a 60-day payback period, not 120.

So when someone asks ChatGPT “should I scale this campaign?” and it says yes based on a 2.1 ROAS, but that 2.1 ROAS actually means they are losing $4 per order after fulfillment costs…

That is not an AI problem. That is a fundamentals problem.

The brands winning with AI right now are the ones who built the rules first.

Target CAC by product.

Contribution margin thresholds.

Break-even ROAS by channel.

Payback windows by customer segment.

Then they hand AI the data and the rules, and let it find patterns within those constraints.

AI is the engine. Unit economics are the steering wheel.

Without the steering wheel, you are just accelerating into a wall.

If you are using AI for your ads and have not mapped your unit economics first, you are not optimizing. You are guessing faster.

Salesforce just replaced half its support team with AI agents. Meta wants to do the same thing to your media buyer by December.

Here is what happened.

Benioff cut his customer support team from 9,000 to 5,000. AI agents now handle half of every conversation. Customer satisfaction scores stayed the same.

Then he went on a podcast and celebrated.
Two months before that, he told Fortune that AI “only augments humans” and that “humans are not going away.”

Sound familiar?

Meta is telling agencies the same thing right now. AI is just a co-pilot. Automation is just a tool. You still need humans in the loop.

Meanwhile they are building a system where a brand uploads one product photo, sets a budget, and Meta’s AI handles the creative, the targeting, the optimization, and the delivery. No agency required.

They want full automation by the end of this year.
Benioff showed us what the playbook actually looks like when it plays out. Step one, tell everyone AI is just a helper. 

Step two, quietly let the AI do half the work. 

Step three, cut the humans and call it exciting.

Here is what I think most people are getting wrong.
AI is not going to replace good agencies. It is going to replace agencies that were never doing real work to begin with. 

The ones running the same broad audiences, recycling the same three hooks, and sending clients a dashboard they do not understand.

If your entire value is button pushing inside Ads Manager, you are the 4,000.

If your value is creative strategy, understanding the customer, and building systems that feed the algorithm what it actually needs, you are not replaceable. Not by Meta’s AI. Not by anyone’s.
But you have to be honest with yourself about which one you are.

The gap between operators who use AI to get better and operators who get replaced by AI is going to be measured in months, not years.
We are watching it happen in real time.

What is your take. Is Benioff right that this is exciting, or is this the beginning of something most people are not ready for?

While you were arguing about which AI tool makes better ads, something happened this week that should concern every single person building with this technology.

Anthropic told the Pentagon no.

No to mass surveillance of American citizens.

No to fully autonomous weapons that select and kill targets without a human in the loop.

They didn’t do it quietly. They did it knowing it would cost them a $200 million contract. Knowing Trump would publicly blacklist them. Knowing the Secretary of Defense would label them a national security threat.

Dario Amodei said it plainly. The technology is not reliable enough for autonomous weapons. And mass surveillance of Americans is incompatible with democratic values.

Then something interesting happened.

Sam Altman told his employees he agreed with Anthropic. Said OpenAI shared the exact same red lines. Called it an issue for the whole industry. Said he trusted Anthropic as a company.

Six hours later, OpenAI signed a deal with the Pentagon to replace them.

Same red lines. Same language. Different outcome.

The Pentagon rejected Anthropic for insisting on protections. Then accepted those same protections from the company willing to take their spot.

Read that again.

Here is why this matters to you even if you never touch a government contract.

We are all heads down right now building AI workflows. Automating creative. Scaling ad accounts. Generating content. Saving time. Making money.

That is the exciting part of AI.

But the same models we use to write ad copy and audit Meta accounts are being deployed on classified military networks. 

The same architecture. The same capabilities. Different use case.

And the company that built the model I use every single day just got punished for saying there should be limits.

This is not a political post.

This is a wake up call for anyone who builds with AI and has never stopped to think about where the guardrails actually are.

Because right now the answer is: wherever the highest bidder wants them.

Pay attention to this one.

taplio

Chris Marrano's Linkedin Analytics

Chris Marrano

Performance ScoreChris Marrano's LinkedIn Performance

Chris Marrano's Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

taplio

.css-6ytdur{background-image:linear-gradient(284deg, rgba(39, 43, 72, 1) 0%, rgba(33, 47, 183, 1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Chris Marrano's Linkedin Analytics

Chris Marrano

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Chris Marrano's LinkedIn Performance

Chris Marrano's Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Chris Marrano's Linkedin Analytics

Performance ScoreChris Marrano's LinkedIn Performance

Chris Marrano's Best Posts (last 30 days)