The 5 to 15% of Revenue Hiding in Your Shopify Store (and Why You Never Capture It)

Steve Hutt:
Welcome back to eCommerce Fastlane. I’m your host, Steve Hutt. Today’s conversation is an important one. It is all about finding the revenue in your Shopify store that every brand and every marketer knows they are leaving on the table. There is always some kind of optimization that can happen. A lot of times you have a sense of what it might be, but you do not always know how to fix it or how to triage what needs to be improved, tested, or experimented on. That is the interesting part. We know this money is sitting somewhere in the store.

Steve Hutt:
It could be pricing, or maybe bundles that customers are already buying together. It could even be retention flows that are firing at the wrong time. Knowing the revenue is there is the easy part. You can often identify some of these things because you are so close to the product and what you are selling. But actually finding that revenue and then capturing it is where most brands get stuck.

Steve Hutt:
They do not really know what to go after first or how to decide which experiments to run, what changes to make, and how to monitor those changes. That is where my guest, Apoorva Modi, is going to help us. He is the founder of a company called Revenue Agent. The app is listed as Revenue Agent in the Shopify App Store. The nice thing about Apoorva is that he spent six years at Google. He will share more of his story in a moment, but I did creep on his LinkedIn profile this morning. He worked on the ads platform at Google for quite a while. He has also run his own DTC brand.

Steve Hutt:
He has done a lot of different things, so he really gets it. He has worked with Wella and with Coty. He is very much an operator, and now also a platform builder, who understands what it takes to improve the performance of a Shopify store. That is why he built this solution. Revenue Agent is in the Shopify App Store right now. The app is new, but he and his background are definitely not new.

Steve Hutt:
So he really understands this space. Apoorva, welcome to the show. I am glad you are here.

Apoorva Modi:
Thank you so much, Steve. That was a great introduction.

Steve Hutt:
All right, I feel great about that. To kick us off, let us talk about Revenue Agent as a product. What exactly does it do? I kind of alluded to it a bit, but I would love to hear it in the founder’s voice. At a high level, what problems are you solving today for Shopify brands?

Apoorva Modi:
Okay, so imagine you are a Shopify brand and you are doing more than a million dollars in revenue. You know in your heart, as I did when I was an operator, that there is money being left on the table from your existing customers. I am not talking about needing to do more advertising or finding new customers, even though that is important. That is not the focus here. The reality is that you are stretched. Every day of the week you are fighting fires. There are issues with employees, customers, or product, and you are always in firefighting mode. You never get around to running the experiments that would uncover that extra revenue. Often you even know what those opportunities are.

Apoorva Modi:
You might be thinking, “My bundles are not quite right,” or “My pricing has not been updated for these products,” or “My free shipping threshold is too low.” I could go on. What Revenue Agent does is use 21 proprietary statistical models to stack rank revenue opportunities for you across eight levers, like pricing, promotions, retention, and so on. Then it goes a step further. It does not just hand you a list of dollar-based opportunities; it also bridges the gap to execution. It will run the experiments autonomously when you give it permission, so the grunt work is not left to you. That is what Revenue Agent is.

Steve Hutt:
That is amazing. I went through your material and noticed how you put these eight revenue levers together. I am going to call them that: eight revenue levers your agent can scan for. I want to call them out so they are in the transcript. Pricing is an important lever that people are probably not testing as much as they should. Free shipping thresholds are important. Bundles are a big one.

Steve Hutt:
We talked about promotions and how much revenue and margin we may be giving away, or not needing to give away. I have heard stories of revenue being lost because of celebrity endorsers or UGC content where the customer was already going to buy, so why are we giving them a coupon when they are already in the funnel? I know you likely see a lot of that. Retention is important. Inventory is a really interesting one that I want to talk about in a few minutes. That felt like an odd place for a revenue leak, but it is an important one. Discounts, obviously. And then checkout and the whole checkout flow. We will get into a lot of those pieces in this episode.

Steve Hutt:
Before we jump into all of that, can you take me back a bit? What were you seeing at Google or even back at Wella that made you say, “Someone has to fix these revenue challenges”? There is clearly a problem. Founders wear a lot of hats. How did you get to, “There are all these problems, I need to fix them”? I always find it interesting how people come up with these concepts.

Apoorva Modi:
A few years ago I was operating as the ecommerce lead for Wella UK, which is a P&G brand in the United Kingdom.

Apoorva Modi:
And by the way, Wella is a professional hair care brand for those who might not know.

Steve Hutt:
Yes, Wella is big in the salon world. They are a salon product brand.

Apoorva Modi:
You are exactly right, Steve.

Steve Hutt:
Yes.

Apoorva Modi:
You know your beauty products. I knew our revenue was not optimized. For example, there were several bundles I wanted to build because I saw a lot of products being co-purchased. I knew that if I offered a small discount on those co-purchases, I could improve my top line and actually my bottom line too, even though I would take a small margin hit on the specific bundles. It was funny: there were 20 such experiments I wanted to run and I ended up running just one that year. That was because the CFO would ask for a report, the warehouse literally caught fire so I was dealing with what to deprioritize on the site, and there were always competing priorities. This was at P&G.

Apoorva Modi:
Then I went to Coty, and later to Google, where I saw from the ads side, through the ads data, hundreds of ecommerce brands struggling with the same thing. These ecommerce operators are smart. Their instincts are good and well honed. They know what they need to do; they just do not get to it because they are capacity constrained and they become the bottleneck. I started asking, “What can we build for them? How can we make this easy without demanding extra time from them to capture incremental revenue that is already in their data?” That thinking is what helped us build Revenue Agent.

Steve Hutt:
This is amazing. I completely agree with you on the human bottleneck. Founders and marketers wear a lot of hats. With constant firefighting and ongoing projects, it is hard to make intelligent decisions about what you should actually be working on. That is where your solution and the scan really help: you find these potential revenue leaks and then figure out how to improve them and start experimenting with the recommendations that come out of those scans. It is very high-tech, but also very practical. I am glad you have identified and categorized these opportunities.

Steve Hutt:
It feels like you are tackling a lot of things at once and then allowing the merchant to decide what to improve, test, monitor, and iterate on. I think that is very cool. I have a question about how Revenue Agent compares to what Shopify has been building over the last few years. Out of the box, Shopify has a lot of tools, not just Sidekick, but a variety of features to help merchants. They are not a Swiss army knife, and they are not going to build or buy every possible app. They allow the partner ecosystem to flourish while they focus on their own roadmap.

Steve Hutt:
Partners like you come in and solve unique problems. Can you contrast what your solution does that Shopify does not necessarily solve on its own?

Apoorva Modi:
That is a great question, Steve, and one we have thought a lot about. As you mentioned, Shopify has Sidekick and a bunch of other offerings. Their analytics are pretty decent, for example. What we did was observe our customers. We asked them, “Can you open your Shopify dashboard and walk us through how you use it, and what you are using Revenue Agent for?” It was very instructive. Their first question was often, “Can I not just do this with Sidekick?”

Apoorva Modi:
Tactically, we noticed they mainly used Sidekick for two things. First, they used it to augment reporting. When you go to Shopify analytics, you can see your data. If you want to pull custom views, it used to be a bit painful. Now, with Sidekick AI, you can feed it custom variables and it will pull that report for you. That is really easy and really nice. Second, we saw people use it to quickly make changes to their website. Previously, they might have needed a designer to update a theme for a new product or promotion, which takes time, especially if that designer is on contract. Now you can do those theme changes much faster with Sidekick. That is how we saw merchants using it.

Apoorva Modi:
What we did not see them using it for was detailed revenue optimization insights. Taking it up a level, Shopify is a fantastic company with fantastic engineers, but they have to serve roughly 5 to 6 million merchants. Their products have to work for everyone, just like I saw at Google where products had to work for billions of users. That means the product has to solve everyone’s most common use cases. It also means that when there is a specific area where you need to go very deep, like revenue optimization, they may not go as deep as a focused solution can. For us, revenue optimization is the starting point, and we will see where we go from there.

Apoorva Modi:
We also build statistical models. AI models, especially language models, will hallucinate and get things wrong. You want solid statistics and econometric modeling running on your data before an LLM comes in to make sense of the outputs.

Steve Hutt:
I see. That is amazing. Let us talk a bit about the broader app ecosystem. A lot of stores are running many different point solutions. When I used to do app audits, there was often a bit of duct tape everywhere. As we know, adding more apps can hurt site speed and Lighthouse scores. There are a lot of side effects from adding more and more to your store. When you look at merchant admin panels, they might have a separate analytics tool, they might use Shopify’s analytics, they might have an external one, they might be connected to a testing tool, plus all the marketing platforms and apps.

Steve Hutt:
What are you seeing on your end? Do you find that adding your solution allows people to pause or discontinue other tools because they do not need everything running at once? Can your app replace some of that complexity?

Apoorva Modi:
Steve, this is such a good question, and we were literally talking about this before the show. Here is what I see entrepreneurs and Shopify merchants doing. They know they need to improve margins because they are facing more competition than ever. Their cost of acquisition is going up on Google, Meta, and TikTok, tariffs are eating into margins, and the environment is tough. So they say, “I need to improve my yield.”

Apoorva Modi:
The first move is often, “Let me do better analysis.” So they install a dashboard to show where they are not performing well or where opportunities might be. Then they decide they need to calculate customer lifetime value, so they install a tool like Lifetimely. Then they need to send emails and SMS, so they install one of those solutions. All of these tools are good and have been refined over time.

Steve Hutt:
Yes.

Apoorva Modi:
But what happens is that the founder or their team becomes the operating layer. They are already swamped, and now they have to build more email flows, do more calculations, and look at more dashboards. That was one of the big insights that helped us build Revenue Agent. A few years ago, the question was, “What is the best tool to do X?” like email. Today, we live in the age of agentic AI. The better question now is, “What is the best tool that will completely handle a workflow for me end to end?” Because that is how you will compete in this new era.

Steve Hutt:
And that is exactly what you do. That is the unique part. I really appreciate the capacity point you made. The team is often the layer forced to execute on all these things. I will add that multiple times when I did app audits inside Shopify, I would ask the director of ecommerce or the marketing director, “Can you walk me through why you chose these specific apps?” Often everyone was in their own silo, and they did not even realize they had a $500-a-month app that no one had used in six months because the team member who championed it had left. It may have been perfect for inventory forecasting or some specific challenge at the time, but now it is just sitting there.

Steve Hutt:
Just doing a top-to-bottom app review can be an efficiency play. I think that is where you fit: instead of doing all the work and research into revenue maximization and profitability forecasting, and then manually making changes and running experiments, your system scans, surfaces answers, and then executes. That reduces cognitive load because the system is more agentic and solves a larger problem as a single solution.

Apoorva Modi:
Exactly right. My dream and vision is that this becomes your whole store operating system. Not only do we execute actions across the eight levers you mentioned, but eventually it should be so seamless that you can just chat with it on your phone, get insights, and have it take care of changes for you.

Steve Hutt:
Wow, what a great time to be an operator. It is really wild. Tell me what happens when someone goes to the App Store and installs Revenue Agent. I believe it is free to install right now, and we will talk at the end about some opportunities for listeners. Walk us through what happens once a merchant installs it. What is happening in the background, what usually gets revealed, and what are the typical next steps?

Apoorva Modi:
It is quite simple, as you said. You run a Shopify store and you install the app. It is free to install. It takes about five minutes to scan five years of your data. If your store is newer, it just scans however much data you have. After the scan, it runs those statistical models in the background. We worked with top data scientists in ecommerce and econometrics to build 21 proprietary models that stack rank opportunities for incremental revenue from your existing customers and your existing data.

Apoorva Modi:
It works across the eight levers you mentioned earlier, like pricing and retention. Then it orders opportunities in descending order, starting with the largest revenue impact and the easiest wins, and moving down to more complex ones. You can click on each opportunity to see the methodology, the actual numbers, and the specific products impacted. It also shows the tradeoffs. For example, if you discount a product, you may make more money but take a margin hit. It lays that out so you as an owner or manager can make a transparent decision.

Apoorva Modi:
It then sets up the experiment for you. It shows you the exact experiment to run: cell A keeps things as they are, cell B sends, say, 50% or 30% of your traffic to the new price or bundle. If you click “approve with guardrails,” it runs the experiment for you. If you see something you do not like or you run an experiment and decide you do not want it, you can click “undo” and it will roll the experiment back.

Steve Hutt:
So this is running in the background. The scan happens, recommendations are generated and sorted, and then you go through them one by one. One example might be a bundling opportunity that shows the incremental value or annual revenue impact of creating a specific bundle. That might increase profit, average order value, and lifetime value. You have already done the statistical work to figure all of that out. It reminds me a bit of RFM—recency, frequency, monetization—understanding customer data and predicting next likely purchases. Buy product A and there is a typical timeline to buying product B.

Steve Hutt:
It is interesting that you have all of this organized into one solution and then you execute on it. It might say, “Here is the impact of bundling,” and then your system goes out and actually creates those bundles, and they appear on the PDP for that product.

Apoorva Modi:
Exactly right.

Steve Hutt:
It is an incredible time to be able to run tests like that. So a merchant installs the app, gets these tests and recommendations. How do you price the product? The next question from people is, “Okay, there is clearly significant upside. How is it priced relative to that upside?”

Apoorva Modi:
Steve, just to clarify, you mean how we price the app itself?

Steve Hutt:
Yes, the app. It is free to install and includes a lot of these recommendations across the eight revenue levers. Once those opportunities are visible and you want to start running experiments, how is it priced? Is it based on revenue increases or a flat rate?

Apoorva Modi:
For now it is a flat rate. As I said, the scan is free and it can be refreshed monthly for free. If you are a small store, please take advantage of that scan. If we can add value, we are doing something right. If you want autonomous experimentation and advanced insights refreshed daily or weekly, you can get that for $99 a month.

Steve Hutt:
So it is very affordable. You can do the free scan, and then the next tier up is a weekly refreshed opportunity scan, and from there you can go further into more of an enterprise solution if you are looking for more opportunities and more guardrails at scale.

Apoorva Modi:
Exactly. We have a couple of enterprise customers. One of the largest supplements retailers in the US is an enterprise customer. They often need custom work. Some of those enterprise customers are not on Shopify, so we bring the statistical models to whatever platform they use, like Salesforce Commerce Cloud. One more note on pricing: typically we see stores find between 5% and 15% of their annual revenue as incremental opportunity.

Apoorva Modi:
So if you are a $10 million revenue store, you might see between $500,000 and $1.5 million in incremental revenue opportunity. That is why we priced it at $99. If you can get, say, $500,000 in incremental revenue at the low end of that range, it is a no-brainer. That is what we want it to be.

Steve Hutt:
Yes, it really is a no-brainer. I kind of knew the pricing ahead of time from looking at the Revenue Agent app listing and website, but I wanted clarity. You open the door with a free revenue scan across these eight areas. Merchants and founders often do not know what they do not know. They know there is a problem and there is opportunity. They just do not know which lever to pull, when, and why, and what the experiment should look like.

Steve Hutt:
People do point solutions, dashboards, CRO tools, and so on. You have put it all together more holistically under these eight revenue levers and the scan. I think that is amazing.

Apoorva Modi:
Thank you, Steve. I heard a great quote about AI that really stuck with me. Someone said, “I do not want AI to do the strategic and creative work so I have more time to do the dishes. I want AI to do the dishes so I have more time to do the strategic and creative work.” That really lit a lightbulb for me. Whatever we build needs to make life better so that the owner or ecommerce manager can spend more time on what really matters: governance, deciding what products to build, which customers to go after—not all the grunt work of setting up experiments and filling in PDPs.

Steve Hutt:
Yes, I totally get that. I have been talking a lot about humans and AI. All the efficiency gains that come from AI-powered solutions and agentic systems are great, especially when they have the right models and guardrails. But that should free up time for humans to do human things: creative tasks, brand-building, and being the human behind the product.

Steve Hutt:
That is one reason agentic commerce is still so early. There are still “wars” around things like UCP, the universal commerce protocol. Meta is working on their version, Stripe is doing their thing, Google has their approach.

Steve Hutt:
I know Shopify and Google have been doing their thing too. It is interesting because people are still a bit nervous about buying directly inside AI interfaces. It is great to be discoverable in AI through tools like Perplexity and Gemini. That is great for the research phase. But even ChatGPT decided not to allow purchases directly in-app because of all the complexity around fraud and the full commerce stack. So they moved to letting people research and then click out to the brand store to purchase.

Steve Hutt:
We have diverged a bit, but it is an important concept for brands to think about. Agentic commerce is still early. Discovery and discoverability in AI models is critical. Actually buying inside an app is another step. In your solution, I think the power is in finding the opportunity and increasing revenue through experimentation.

Steve Hutt:
That is where the lift happens, and your pricing is unreal for what it delivers.

Apoorva Modi:
Thank you, Steve. Agentic ecommerce is definitely top of mind. I was at the Lead Summit in New York last week and everyone was talking about it. The first thing they said is, “Get your structured data right,” because that is what makes you more shoppable. The second part is that models like ChatGPT, Claude, and Gemini need to figure out, and are figuring out, how to make the buying experience seamless because there is so much at stake.

Apoorva Modi:
My co-founder is going to Goa, a beach town in India, and he was trying to buy waterproof shoes. He tried using Claude and ChatGPT and said the experience was terrible, so he went back to Amazon and vendor websites. You are right: that space is evolving and it will be fascinating to see how we enable and thrive in it.

Steve Hutt:
This is amazing. What do you see as the next steps for people listening today? People listen to this podcast for a reason: to understand the lay of the land and what is new and exciting. You are in the early days with this product, but you have a lot of experience behind you. You are a current and past operator. Knowing all that, what do you see as the next steps for listeners? And is there a sweet spot of merchant where your solution is most impactful?

Apoorva Modi:
Great question, Steve. For those listening, I would say this: for any tool you are using, for any area of your work life, ask which workflow it can take over instead of just asking what it does or what dashboard it provides. That is the first thing. Second, our ideal customer profile is anyone doing over a million dollars in sales. Even at around half a million, depending on your product, we can help, because we need enough transaction volume to provide good statistical recommendations with high confidence. And yes, I would say the next step is simply to install the app. Steve, can you remind me of the full question again?

Steve Hutt:
Yes, I was really asking about next steps. Someone listening might be thinking, “I want to try this. I want to install the app.” We have already walked through how quickly it can be installed and what happens next, but I want to make sure people understand that it is free, what it is going to show you, and what happens after that. If you like what it does, how everything is ranked by dollar impact, and how the changes work, what are the next steps?

Apoorva Modi:
Sure. I will save the new customer offer for your next question. The immediate next step is to install the app. As Steve and I said, it is totally free. If you are a small store, just enjoy the scan. Look at the opportunities it surfaces. Even if you do not go for the paid version, go run that retention flow it suggests for your top customers. It will identify those customers for you. You can tag them for free.

Apoorva Modi:
If you are a larger store doing more than about $100,000 a month or roughly a million a year, you can subscribe to the paid version for $99. That unlocks autonomous experimentation and in-app actions. Our hope is that the app helps you generate additional revenue from the data you already have.

Steve Hutt:
This is amazing. I will make sure everything is in the show notes. The URL is revenueagent.app. You can also go to the Shopify App Store and search for Revenue Agent. It will pop right up. As we said, it is a new solution, but it already has a lot of five-star ratings. People clearly see the upside. I think this is great.

Steve Hutt:
Once again, thank you for sharing. I feel blessed to have the opportunity to interview people like you who are building interesting, impactful solutions. This is not just about efficiency gains; this is about revenue gains through experimentation and scanning.

Steve Hutt:
I love the eight revenue levers you have identified. You are giving people the intentional space to say, “We ran the scan. Here is what we found. Now, what do you want to do next?” Then the app goes out, runs those experiments, and you can see the revenue lift. That really sets people up for success. I commend you for giving merchants that intentional space and path.

Apoorva Modi:
Steve, thank you. You are very kind to give startups like ours a platform to share what we are building. I also love eCommerce Fastlane and the support you have given us. For all the listeners of this show, we would love to give you a month of the Pro tier free, on us. Please reach out. Steve will share the link. For the first 30 listeners, our team will personally walk you through your scan and identify the top three opportunities for incremental revenue. We would love to support the listeners of this show.

Steve Hutt:
Lovely. I will put that in the show notes. If you are listening right now, jump on it. Only 30 people will get that personal audit where someone from your team will review the scan and offer specific recommendations on next steps. At the end of the day, the upside is for everyone. The customer wins because they get the product, bundles, and pricing that fit them. The merchant wins because revenue, LTV, and basket size all go up. And your company gets a small amount to keep building and improving the product.

Steve Hutt:
This is great. Thanks again for recording. I wish you continued success. This has been excellent.

Apoorva Modi:
Thank you, Steve. It was such a pleasure to talk with you.