💸 The $500K Hiding in Your Shopify Store

July 09, 2026 | Read online

💸 The $500K Hiding in Your Shopify Store

The revenue, the surfaces, and the AI numbers worth trusting are all closer than the headlines want you to think.

In partnership with

Join 49,914 founders and marketers are getting this newsletter today.
Welcome to the 184 new operators who joined us this week! 🤯

Hey Fastlane Insiders! 👋

Here's a pattern worth sitting with this week. The loudest numbers in ecommerce right now are the ones you should trust the least, and the most valuable ones are already sitting in data you own.

Three threads pull that together. AI shopping is no longer one assistant you optimize for. ChatGPT, Claude, Gemini, and Perplexity now split the referral market four ways, and the "ChatGPT only" playbook covers a third less of it than it did a year ago. The stats describing that shift contradict each other by a factor of 100, depending entirely on what each source quietly divided by. And underneath all of it, most Shopify brands over $1M are leaving 5% to 15% of annual revenue uncaptured in their own customer data, waiting on bundles, pricing, and retention flows that never ship.

Whether you're chasing your first $100K or scaling past $10M, this edition is a decoder: what's noise, what's signal, and where the next 5% to 15% of revenue is actually hiding. Episode 465 with Apoorva Modi (six years on Google's ads platform, now founder of Revenue Agent) breaks down the revenue hiding in plain sight. Two knowledge drops map where AI buyers actually come from and how to read the numbers before you move a dollar. And the week's headlines make the point for me: a Buy rating on Shopify that flips the AI-bypass panic on its head, and a 250% TikTok Shop launch that means something narrower than it sounds.

The connective tissue: your own corrected data beats every headline.

Here’s what made the cut this week:

🎧 This Week's Podcast: The 5% to 15% of revenue hiding in your Shopify store, and why you never capture it.

💡 Knowledge Drops: Where AI buyers actually come from, and why the stats about them disagree by 100x.

🔥 Tool of the Week: Katana, one real-time view of your stock across every channel.

📡 Industry Pulse: BofA calls the AI bottom on Shopify, Klarna files to become a bank, and Milani's 250% TikTok Shop headline, decoded.

Let’s get into it. 👇

Your Revenue Is Leaking. You Just Can't See It Yet.

Rogue resellers and copycats aren't just annoying — they're quietly draining your margins, tanking your buy box, and eroding the brand you've spent years building.

Most merchants don't find out until it's already expensive.

Here's what we've found inside brands just like yours:

A supplement brand was hemorrhaging $162K every month to foreign infringers. We recovered $131K/month — a 13,361% ROI.
A $250K revenue siphon traced back to just two or three unauthorized sellers operating across Amazon and TikTok Shop. Wiped out.
An exercise accessory brand doing $1.5M/year walked away from Amazon entirely out of frustration with unauthorized sellers. We removed them and brought the brand back.

This isn't guesswork. We map every leakage point specific to your brand and tell you exactly what it's costing you.

Brand Audit Report

🎧 New Podcast Episode! 🎧

The 5% to 15% of Revenue Hiding in Your Shopify Store

Every brand over $1M can feel the money it's leaving on the table: bundles that never launch, pricing that hasn't moved in 18 months, free shipping thresholds set too low. Knowing the leak exists is easy. Fixing it while you're buried in firefighting is why it keeps slipping to "next quarter," every quarter. This week I sat down with Apoorva Modi, founder and CEO of Revenue Agent (ex-Google ads, ex-Wella UK), and one line stuck with me: the money is already in your store.

Here's what we unpacked:

The real reason your experiments never ship. Why 20 planned bundle tests turn into one actual launch, and why the bottleneck is almost never strategy.
The eight levers where revenue quietly leaks. Pricing, promotions, bundles, retention, inventory, discounts, free shipping thresholds, checkout, and how to rank them by dollar upside instead of gut feel.
Why 5% to 15% of annual revenue is hiding in plain sight. For a $10M brand, that's $500K to $1.5M sitting in data you already own, with zero new ad spend.
Why you should buy workflows, not dashboards. The question in the agentic era isn't "what's the best tool for X." It's what will run the whole thing so your team stops being the glue between a dozen apps.
Why statistics should power your AI, not the reverse. There's a reason a language model doing your math will hand you confident, wrong numbers, and what Revenue Agent does instead.

We also get honest about how early agentic commerce still is, including why his own cofounder gave up buying shoes through an AI assistant and went back to Amazon, and what that means for where you place your bets right now.

[ LISTEN NOW → ] Skip this one and you might literally be leaving 5% to 15% of your revenue on the table.

💡 Knowledge Drops of the Week 💡

Your AI Buyers Aren’t All In ChatGPT Anymore

Eight months ago, you could get away with a "ChatGPT only" strategy. One assistant carried almost all measurable AI shopping referrals and felt like the whole market. Today, four AI surfaces split that demand, each using its own logic to decide what to recommend. If you optimized for ChatGPT last year and then stopped, you did the right work at the exact moment the market stopped being one channel.

Here's what's working now:

Treat ChatGPT as the floor, not the whole floor. It still carries most of the volume, but its share has slid meaningfully in just eight months. Leaning on it as your entire AI strategy now leaves a growing chunk of buyers uncovered.
The smallest surfaces often convert the hardest. One of the lowest-volume assistants quietly posts the highest conversion rate, and another the highest average order value. That changes the math on whether "small" should ever mean "ignore."
Your revenue stage decides your lane. Under $500K, two surfaces cover roughly three quarters of the volume for free through Shopify's native AI channels. Above that, the play changes, and trying to be everywhere at once becomes the premature complexity that quietly sinks brands.

The market restructured once this year and will likely do it again, so the durable move isn't loyalty to any one assistant. It's clean product data and a measurement setup that shows you where your buyers actually come from.

[ READ THE FULL BREAKDOWN →] See which AI surface actually matches your stage and catalog.

Why “AI Is 20% of Ecommerce” Is Almost Certainly Wrong

Adobe says AI visitors convert 42% better than everyone else. A peer-reviewed study says AI is under 0.2% of ecommerce traffic and, for a while, converted below organic search. Both are right, and that's exactly how merchants get talked into moving real budget off a single conference slide.

Here's what's working:

Ask four questions before you trust any AI stat. Denominator, definition, timeframe, and who benefits. Miss the denominator and "20% of referral traffic at one giant retailer" quietly becomes "AI drives 20% of ecommerce," inflated by roughly 100x.
Fix the broken ruler in your own dashboard. Around 70% of AI referrals arrive with no referrer and land in GA4 as Direct, and agentic orders often never fire a browser session at all. Before you decide the channel is too small to matter, make sure your setup can even see it.
Treat AI as a small, high-intent, fast-growing lane. Right now it's roughly a 1% channel, growing at triple-digit rates with above-average intent. That makes it worth systematic preparation, and completely unworthy of panic reallocation based on a headline.

The exact percentages here will be stale within a year. The four questions will still work, and so will the discipline underneath them: your own corrected data outranks any industry headline, including the ones in this article.

[ READ THE FULL BREAKDOWN →] The decoder for every AI number that lands on your desk.

🔥 Tool of the Week 🔥

Know your stock. Everywhere. All the time.

Sell on Shopify, Amazon, wholesale, and retail with cross-channel, real-time stock visibility. With Katana, you get the right numbers every time - across sales, orders and production. All the features you need without the ERP price tag or complexity. Start today and go live in as little as a week.

Get Started Today

⚡ This Week’s Industry Pulse ⚡

❝

A handful of updates land that actually move the needle. Here's what made the cut…

Shopify Gets Its First Buy Rating in a While. Bank of America reinstated coverage of Shopify on July 7 with a Buy rating and a $150 price target, arguing the year-to-date pullback created an attractive entry point as agentic commerce becomes a tailwind rather than a threat. The tell: Shopify's AI-powered merchant traffic is up 8x year over year, with orders from AI search up roughly 13x.
Klarna Wants to Become an Actual Bank. Klarna filed applications on July 6 with the FDIC and Utah regulators to launch Klarna Bank USA, an FDIC-insured industrial bank that would fund BNPL loans with customer deposits instead of partner banks. It's the clearest signal yet that a checkout option running on thousands of Shopify stores is angling to become full financial infrastructure.
A Shopify Merchant's TikTok Shop Launch Beat Goal by 250%. Milani Cosmetics, a $250M mass beauty brand built on Shopify, blew past its internal sales target by more than 250% in the first two weeks of a TikTok Shop mascara launch. Before you read that as "TikTok Shop is the channel," note the quieter move: Milani is already registered for a ChatGPT integration pending approval, treating TikTok Shop as demand-testing while building toward AI-native discovery.

Until Next Thursday

Here's what kept surfacing this week as I wrote this edition…

Every genuinely useful move this week points inward, not outward. The revenue Apoorva helps brands find is in data they already own. The AI stats you can trust are the corrected ones from your own store, not the conference slide. And the AI surfaces worth optimizing for are the ones your actual buyers use, which you only know once your measurement is honest.

The brands that win this next stretch won't be the ones chasing the loudest number. They'll be the ones who read their own data clearly and act on it. So this week, before you reallocate a dollar or add another tool, ask what your own numbers are actually telling you.

If one of these hit home, or you're already testing one of these plays, hit reply. I read every response, and they shape what we dig into next.

Keep reading your own data and owning your growth.

P.S. Missed a previous edition? Check out the archive for more growth strategies and insights.

Cheers!
Steve

Update your email preferences or unsubscribe here

7891 211 Street
Langley, BC V2Y 0L5, Canada

unauthorized hotel charge?

- do you pay on behalf of your travelers?

Most companies I talk to send their primary card numbers, which often get unauthorized charges, taking time to reconcile.

If you’d like to prevent this, ITILITE has a solution that can eliminate fraudulent use.

Just say “yes” and I’ll tell you how.

Outbound Wizards - A New Weekly Newsletter

July 06, 2026

Outbound Wizards - A New Weekly Newsletter

Issue #20

Welcome to Outbound Wizards. Issue 20.

As I write this, I realize I've given up on consistency, at least with this newsletter.

But I have good excuses.

First of all, the twins turned one.

Second of all, we were onboarding a massive Australian agency that’s shifting from Skylead (they have 80 clients and it’s a $4k MRR deal for us).

All of this meant a lot of custom work in migration and onboarding.

Finally, we are working on a new SalesRobot Agentic Mode that will do the work of getting you enough replies from cold outreach with skills that you've never seen before.

So yeah, those are my excuses.

But going forward, I will try to be more consistent 🙂

Anyway, back to LinkedIn automation news.

In this issue:

Milestones: Key metrics on our path to 10,000 users (currently at 5582, slight growth!)
LinkedIn Message Roast: We find the worst performing Linkedin Messages from real SalesRobot users and show you how we'd save the campaign (after making fun of them of course lol)
Outbound Wizard of the Week: I talk to Rishabh Gupta, product lead at Partao and freelance GTM engineer, on building foundational GTM systems for early-stage Indian startups and then handing them off.

We explore his Fortune 500 campaign targeting CFOs that forced a pivot from email-only to multi-channel after three months of flat results, plus his GCC setup service campaign that required combining Lusha, Clay, and HubSpot lead scoring to crack hyper-segmented personas.

Rishabh shares his path from biomedical engineer to first growth hire at a company he helped scale to $11M, then accidental GTM engineer after inheriting a failing contractor setup, explaining why GTM engineering is just a subset of growth marketing and predicting fully agentic GTM within two years where a human farmer manages a field of specialized AI agents.

Milestones

MRR: $85,220 (up!)

Where we stand:

[GOOD NEWS] Sales demos done (from 15 Jun 2026 - 6 Jul 2026): 25 (demo numbers continue to be low, but average revenue per demo has gone up massively)
[GOOD NEWS] Free Trials Started (from 15 Jun 2026 - 6 Jul 2026): 316 (averaging ~15 signups per day, up)
[GOOD NEWS] New subscription revenue: $3100 in new + upgrade revenue from 55 users (significant increase)
[BAD NEWS] Downgraded subscription revenue: $900 in downgrades and churn since I sent the last newsletter

LinkedIn Message Roast

Now, let’s dive into the LinkedIn message a SalesRobot user sent to 623 people on LinkedIn and got a whopping 4% response rate (average is 10%)

The Message Copy (some parts [REDACTED] to hide the identity of the user):

LinkedIn Outreach Sequence: Executive Resume Writing Service Campaign

Performance Overview

Message Response Rate: 3.33% (8/240)

Step 1: Like Their Most Recent Post

Timing: Initial action

Action: Engage with prospect's latest LinkedIn post

Step 2: Send Connection Request

Timing: [Set days/hours after Step 1]

Message: [Standard connection request - no blank message]

Step 3: First Message (Credentials + WhatsApp CTA)

Timing: [Set days/hours after Step 2]

Message:

Hi [firstName],

Thank you for accepting the connection request. Really appreciate it.

𝑨 𝑸𝒖𝒊𝒄𝒌 𝑰𝒏𝒕𝒓𝒐𝒅𝒖𝒄𝒕𝒊𝒐𝒏::

I am an MBA graduate with 15 yrs of experience as an Online BDM worked with companies like [REDACTED].

I help Senior Management professionals to build their Resume & LinkedIn Profile at par with international standards.

We are a Professional Executive Resume Writer with 5000+ Executive Resumes written since 2009 with clients from across the Globe.

Our articles were also published on various top-tier news media platforms including the [REDACTED], etc.

And I was awarded & recognised in the "[REDACTED]" in 2019 by Silicon India Magazine.

I look forward to a great association together.

WhatsApp us on: [REDACTED]
WhatsApp Direct message: [REDACTED]

Result: 3.33% response rate (8/240)

I know we've roasted something similar before, but this was so bad it deserved a rewrite.

"𝑨 𝑸𝒖𝒊𝒄𝒌 𝑰𝒏𝒕𝒓𝒐𝒅𝒖𝒄𝒕𝒊𝒐𝒏::"
(The fancy font is BACK! Double colon for double pretentiousness!)
"I am an MBA graduate with 15 yrs of experience"
(Still leading with their resume in a RESUME WRITING pitch - the irony remains delicious!)
"5000+ Executive Resumes written since 2009"
(16 years and STILL sending the same terrible outreach!)
"I look forward to a great association together."
(The most awkward partnership phrase in human history - STILL!)
"WhatsApp us on: [REDACTED]
WhatsApp Direct message: [REDACTED]"
(STILL two WhatsApp options because one wasn't desperate enough!)

This is what they should have done instead:

Step 1 - Connection Request (No blank request, no post liking):

Hi [firstName],

I noticed you've been at [companyName] for [years]+ years as [title] - impressive track record in [industry].

I specialize in executive personal branding on LinkedIn and would value connecting.

Best,

[FULL NAME]

Executive Branding Specialist

Step 2 - Message (After acceptance, 24+ hours later):

Thanks for connecting, [firstName]!

Quick observation - your experience at [companyName] is impressive, but your LinkedIn profile might be underselling you.

I looked at 5 other [jobTitle]s in [industry] and noticed they're getting 3-4x more profile views by structuring their profiles differently.

We recently helped [similarPerson] ([similarTitle] at [similarCompany]) restructure their LinkedIn presence. Within 60 days:

- Profile views increased 340%

- Inbound recruiter messages tripled

- Landed 2 interview calls from Fortune 500 companies without applying

Would you find a free 5-minute profile audit useful? I'll record a quick Loom showing 3 specific changes you could make today.

No strings attached - just thought you'd want to see how you compare.

Best,

[FULL NAME]

Executive Profile Strategist

[COMPANY NAME]

GTM Engineer of the Week: Rishabh Gupta, a Freelance GTM Engineer (ex-Growth Marketing)

1. What kind of clients do you work with?

Rishabh is a product lead at Partao and freelance GTM engineer working across D2C, direct-to-consumer brands, and B2B SaaS. He targets early-stage founders (pre-seed, seed, series A) rather than established companies.

"I don't look for founders doing a couple million ARR. I look for startup founders just starting out. My role is foundational: build integrations, workflows, set up the entire GTM engineering process, then go away. It's a foundational role rather than coming on top and optimizing."

2. What is the most creative campaign you have run?

Fortune 500 Multi-Channel Pivot:
Built an email-only GTM motion using Clay workflows pushing to HubSpot and email marketing tools. Targeting CFOs and senior executives at Fortune 500 companies. "The entire play of only emails doesn't work. Over three months I realized you have to include cold calls, SMS, WhatsApp. Multi-channel, not just email."
GCC Setup Service for HR Tech:
Client was selling GCC (Global Capability Center) setup services, helping Fortune 2000 companies build 0-200 employee operations in India. Key challenge: narrowing the target customer persona using Clay alone wasn't enough. Had to combine Lusha and other tools, bring data into HubSpot, run lead scoring, then build the outbound motion from scored leads.

In his words: "For very segmented target customers, you have to bring in multiple tools to get real data. Lead scoring first, then outbound."

3. How did you get into this space?

Education: Biomedical engineer. Then worked as software developer for a year, founders associate, then technical product management for 8 years across e-commerce and fintech.
First growth hire for a San Francisco company. "A lucky coincidence that I met the founder. When I left, they were doing $11 million per year." That credibility opened doors.
Entering GTM by Accident: Browsing forums, posted he was available for part-time work. A senior personality who established Jio Networks within India reached out. "Very impressive personality, once-in-a-lifetime opportunity." After joining, he drifted across different areas of the startup.

"GTM was something they were struggling with. A contractor was helping build the GTM motion, but that really sucked. I took ownership. Dove into Clay workflows, integrations, the entire GTM motion. That helped me understand what it means to scale a company from zero to at least $100K."
The Overlap: "GTM engineering is a subset of growth marketing. Growth marketing is a mix of product, inbound, ads. GTM engineering focuses more on outbound. I don't know why people make it out to be something completely different."

4. Where do you see this whole space going?

Fully Agentic GTM:
"I'm not sure about five years because even Sam Altman can't predict that. For the next one to two years, GTM is going to become fully agentic. For every small task, you'll have small agents. But the entire workflow has to be set up by a person who knows GTM."
The Farmer Analogy:
"A person sits there managing multiple small agents doing their tasks. You're a farmer and these are different plants. You're farming around different outputs."
Multi-Agent Coordination:
"Every agent is good at its own job, but has to learn to interface with other agents. Almost like a team of AI employees or AI freelancers."

5. What advice would he have for somebody starting as a GTM Engineer today?

You don't need a CS degree, but some technical bent helps:
"I'm a biomedical engineer who worked as a software developer for a year. You need enough technical understanding to look at APIs, how Clay integrations work, JSONs, prompting. In India, you first become an engineer, then become something else."
Multi-channel from day one:
"Email used to work five years ago. Same email today, no matter how well you do personalization or research, doesn't cut it. Include cold calls, SMS, WhatsApp. Be multi-channel."
Lead scoring before outbound:
"For segmented target customers, Clay alone isn't enough. Bring in multiple tools (Lusha, etc.), combine data in HubSpot, run lead scoring, then build your outbound motion from scored leads."
GTM engineering is growth marketing, scoped to outbound:
"A lot of people struggle because they say 'I don't have a technical background.' GTM engineering is a subset of growth marketing. If you've done growth marketing, you're already partway there."
Build foundations, then leave:
"My role is foundational. Build up integrations, workflows, set up the entire GTM process, then go away. That's more valuable than optimizing someone else's existing process."

Connect with Rishabh:

LinkedIn: Rishabh Gupta

Listen to the episode:

Spotify: The GTME Podcast by SalesRobot
Apple Podcasts: The GTME Podcast by SalesRobot

That’s it for Issue #20 of Outbound Wizards
If you found this helpful, share it with a friend. Got feedback? Shoot me a reply—I read every one of them.

Thanks for following along.

Saurav G.
Founder, SalesRobot 2.0
Follow me on LI

Update your email preferences or unsubscribe here

228 Park Ave S, #29976, New York, New York 10003, United States

Cinema Latest

Tamil Song Lyrics