Megan Harper
Online stores have a hard time turning browsers into buyers. Yoav Oz, co-founder of Rep AI, created a way to make online sales work like in-store sales. Using behavioral AI, Rep AI engages shoppers at the right time, increasing conversions and improving the customer experience without being annoying.
1. What inspired you to create Rep AI, and how does it solve the challenges eCommerce brands face with customer engagement and support?
Yoav Oz: The idea for Rep came about after I sold my last company, SpotAd, which was a pioneer in the use of AI for mobile advertising. As I pondered what to do next, I did what a lot of us do when we’re sitting idly at a screen – I shopped. Or, more accurately, I browsed. And browsed, and then browsed some more. But I wasn’t feeling compelled to buy anything.
At this point I thought about how normal this shopping-online-and-not-buying behavior is. Despite all of the advances we’ve made technologically, average ecommerce conversion rates exceeded the low single digits. Meanwhile, brick and mortar retailers are converting at rates of 20 to 40 percent. Why is that? It isn’t like there’s a massive difference in products or pricing that would favor physical stores. What is vastly better in person is the experience, and this was something that I knew AI could help solve.
2. Rep AI’s Behavioral AI enhances shopper engagement. Can you elaborate on how it senses disengagement and boosts conversions?
Yoav Oz: Most shoppers in a brick and mortar store don’t want a sales associate to pounce on them the second they walk into the store, and really skilled salespeople know this. So, instead of instantly offering assistance, they’ll keep an eye on what the shopper is doing from a distance. Then, based on their experience, they’ll watch for cues that indicate the shopper needs help, and only then will they approach. The shopper may not want help, and that’s OK. But if they do accept help, they’re suddenly much more likely to buy.
What Rep AI is doing is bringing this experience into the online realm. Our chat-based tool incorporates proprietary behavioral AI algorithms that analyze the shopper’s behavior as they browse. It’s trained to look at more than 500 data points, at regular intervals, to determine when a shopper is on the verge of disengaging. Then, just like that expert sales associate, it’s only at that point that the chat steps in.
Also like in a physical store, the shopper can choose whether or not they accept help. If they do, the results are very much like they are in the brick and mortar environment – once they interact with the AI sales concierge, they’re much more likely to convert.
3. How does Rep AI’s AI Concierge balance sales support and customer assistance to create a seamless experience for users?
Yoav Oz: One big difference with Rep AI is that we aren’t narrowly focused on the sales funnel. Instead, we’re trying to understand the shopper’s perspective and their individual journey. If you think about how you persuade someone to buy in human terms, the flow is something like this:
First, you learn the shopper’s situation or challenge, then you match products that will meet those needs. But it’s not a hard sell. You would present the information using a service-oriented, consultative approach. Also, the products you suggest are appropriate, and that demonstrates your deep knowledge of the catalog and brand. Lastly, if there are objections, you can address them using proof, commitment and reciprocity.
This is what our AI can do – that’s why we call it a sales concierge. As for how we achieve this, we’ve already talked about the behavioral aspect. On top of that, though, we’re analyzing cues about emotion and context so the AI can address the problem that the shopper is having at that moment. We refer to it as digital empathy, and it’s the special sauce that allows Rep AI to do more than churn out responses based on keywords it plucked out of the shopper’s inputs.
It’s all about right time, right context and right content. Which is also the answer to the question of seamlessly balancing sales and support. Even in its initial form as a sales solution, our AI was able to handle 95% of support inquiries. Now, we’ve added specific support skills, enabling merchants to automate common inquiries, like those around tracking, cancellations and returns, and that number is up to 97% on average.
4. With over 70 integrations available, how does Rep AI ensure compatibility and enhance merchants’ existing workflows?
Yoav Oz: While we think our AI concierge offers incredible value on its own, it’s even more powerful when integrated with the other tools in a merchant’s tech stack. That’s why it’s been important for us to make those integrations as simple as possible – we’re talking one click.
With all of our support integrations (e.g., Gorgias, Zendesk, Richpanel), Rep AI becomes the primary interface and first line of support. If our AI can’t resolve the problem or successfully answer the question, then shoppers can summon human support with a button click or message request. Because the support software is always there in the background, the handoff is seamless. We also integrate with social media platforms, so direct messages will go straight into our customers’ support workflows, which helps ensure no inquiry goes unnoticed.
Another area in which our integrations have had a surprisingly large impact is with tools for SMS and newsletter marketing. Because our AI knows how different segments will respond to email capture tactics based on where they are in the consideration process, it identifies the ideal time in the conversion path to present opt-ins. And because we integrate with tools used to execute campaigns, the signup form is connected, so addresses are automatically added to lists in your marketing solution.
5. Your platform promises measurable growth, like increased conversion rates and AOV. What factors contribute to these successes?
Yoav Oz: As we mentioned earlier, Rep AI’s chat is able to provide shoppers with highly personalized assistance, much like a skilled associate does in-store, and this is a great way to boost sales conversions. The simple act of re-engaging customers who would otherwise disengage, whether it’s offering general assistance or providing promotional incentives to complete the purchase, can increase conversions by 14%, on average and across all eCommerce verticals. As for average order value, our AI is adept at identifying products aligning with each customer’s individual needs and preferences, which makes it more effective at upselling and cross-selling.
But there’s more to the Rep’s power than the customer-facing chat. Because every interaction adds to our AI’s knowledge and contributes to insights that can inform improvements across a number of areas. For example, studying where visitors click, how far they scroll, and when they leave can provide a clear picture of where UX improvements are needed. Further, the analysis of conversations can uncover areas where content like product descriptions can be improved or cases in which resources like size charts need to be made more accessible.
6. Rep AI boasts multilingual capabilities. How does this feature help businesses scale globally while maintaining cultural relevance?
Yoav Oz: The ability to speak in shoppers’ native language – naturally and not robotically – is a large part of our tool’s ability to enhance customer satisfaction and drive sales, so making our chat available in different languages was an obvious priority for us. And our merchants, too, because it enables them to expand into new customer segments and even new markets without losing the personal touch that is a foundation of our tool’s power.
Our multilingual capabilities aren’t something we believe we can “set and forget,” either. We are continuously refining our language models based on interactions that shoppers are having through our technology. This helps ensure both our linguistic accuracy and cultural sensitivity are constantly improving.
7. What sets Rep AI apart from other conversational AI solutions in the market?
Yoav Oz: Hands down, it’s our behavioral AI. We built it, it’s completely unique, and it’s what enables our platform to work so effectively.
First, it allows our chat to be unobtrusive. No one likes a pushy salesperson, and the same is likely true for pop-ups that hit the second you visit a website. While you’re browsing, Rep is in the background, gathering and analyzing 500 different data points. The AI may not initiate a chat until it is certain you’re about to disengage, but it’s still learning based on your behavior.
The insights it garners are very specific to each merchant, and they go a long way toward informing areas where that merchant can optimize. Say the AI notices that people drop off after looking for return information or size charts. That tells us pretty clearly that there is something causing friction, and the merchant can address it.
Our behavioral AI also helps us minimize unneeded messaging, which contributes to better experiences for the shopper. It also allows us to price our platform differently. We don’t need or want to charge by message. Instead, we cater our packages to merchants based on their traffic and catalog size.
Moreover, our technology reduces the time to purchase by 34% through its proactive AI-driven approach. With our deep contextual understanding, we gather data at a scale 80 times greater than any other competitor in the market. This vast dataset enables us to provide brands with valuable insights to optimize and enhance their performance.
8. Can you share a success story where Rep AI significantly impacted a merchant’s revenue or operations?
Yoav Oz: One of our favorite Rep AI success stories is SNOW Cosmetics, a DTC company specializing in tooth whitening solutions. Through a stellar combo of superior product and savvy marketing, they went from zero to $100 million in sales in just three short years. While that’s great, the fast-growing sales also meant fast-growing customer support needs. Regardless of the team’s skill and tenure, they were spending so much of their days on low-level tasks that it was difficult to keep up with the busy queue.
So, when they started looking toward automation solutions, they initially thought they just needed to streamline processes for non-decision inquiries like “where’s my order” and “what’s your return policy.” It was while testing Rep AI that they had the eureka moment: this was a tool that could do far more than fix support challenges, it could help boost sales.
As a plug-and-play solution, Rep AI can be implemented in just six clicks, but it can also, equally easily, be customized for each specific merchant’s needs. In the case of Snow, Rep AI worked closely with them to imagine scenarios and train the AI accordingly. Since about 85% of Snow’s site visitors are there for teeth whitening solutions, they developed a tooth color chart. When a shopper indicates they’re looking for information on whitening, the chart pops up, and the shopper simply indicates which is closest to theirs. With that simple step, Snow was able to guide the shopper to the most suitable products, and the result was a massive increase in conversions – up to 15%, or about triple the average, according to the company.
Rep AI also worked with Snow to solve the problem of abandoned checkout. Usually, there’s no way to reach the customer once they’ve left the site, but with Rep AI, if the visitor comes back, the AI knows from the IP address and presents them with a special offer. That particular flow contributed to an astonishing 30% conversion rate.
Of course, the AI is also trained on everything from the entire website as well as any other form of collateral, so it can easily provide valuable product information, usage instructions and other answers to their questions – including “where’s my order.”
By implementing Rep AI, Snow not only improved their support channel, they also turned it into a sales channel. Without reporting, they can tell with absolutely zero doubt that the sale was directly driven by the assistance the shopper received from the AI.
They had spectacular results. Just within the first 60 days, they achieved a 33.85% conversion rate from the new abandoned cart flow and a 22% conversion rate for product pages. This, while also having almost 99% of support questions handled by the AI.
9. What challenges have you faced while developing and scaling Rep AI, and how have you overcome them?
Yoav Oz: Among the biggest challenges when developing Rep AI was that we started in 2020 – incredibly early in terms of generative AI. This means the technology was more or less “wild,” like it was not yet being used without supervision and was often unpredictable.
As a result, we had to build the product from scratch. This included training the AI to understand ecommerce, then setting up intricate guardrails during implementation in order to restrict what the AI does – these are the things that keep your AI from “going rogue” and doing things like pulling content from competitor websites. To empower our customers with the ability to customize, we also had to make sure the AI could be further trained by the merchants themselves.
10. What personal sacrifices or pivotal decisions have shaped your entrepreneurial journey?
Yoav Oz: One experience that truly shaped my journey was moving to China—it was a complete shock. A new culture, a language barrier, and intense KPIs weighed on me after convincing our board to expand into the Chinese market. I found myself as an “anonymous entrepreneur,” starting entirely from scratch—to the point where I even received a new name.
To make things even more challenging, the board and my co-founders didn’t believe in the expansion. They gave me only 5% of the budget I had requested—just to quiet me down.
Yet, after 18 months of tough negotiations and setbacks, we secured our biggest deals with Baidu and Tencent. Our China office, operating with just 20% of the resources of our U.S. office, was generating 1.5x more revenue.
That journey taught me a powerful lesson: if your vision is strong enough, no matter where you are or what cultural barriers stand in your way, you can achieve it.
11. What does success look like for you and Rep AI in the next 3–5 years?
Yoav Oz: Five years from now, the world of eCommerce and brick-and-mortar retail will look very different. I believe REP AI’s behavioral AI and unique power-up integrations with other tech stacks will not only revolutionize online shopping but also enhance in-store experiences. Customers will be able to use AI on their phones to receive personalized recommendations, real-time assistance, and seamless checkout options—bridging the gap between digital and physical retail.
From a market presence perspective, our goal is clear: to remain the #1 AI solution for eCommerce. We will continue to lead the industry by driving innovation, expanding integrations, and delivering unparalleled value to brands and their customers.
12. What does a typical day in your life as a founder look like, and how do you stay motivated?
Yoav Oz: As a second-time founder, you start to understand what works for you and what doesn’t. My day starts early. I wake up at 5:30 AM and take my dog for a walk—without my phone—so I can have a quiet moment to clear my mind. At 6:00 AM, I dive into emails and texts, and from there until 8:00 AM, I spend quality time with my daughter. After that, I hit the gym, where I also use the time to listen to Audible, podcasts, or great music for about two hours.
Work takes up most of my day, but I’m intentional about my evenings. My calendar is usually blocked from 6:00 PM to 8:00 PM for family time and dinner. After that, I shift back into work mode, handling meetings and tasks until around 11:00 PM—since our business operates on U.S. hours while I’m based in Israel.
Beyond work, I make sure to maintain balance. For example, twice a week, my wife and I have a date night, and at least once per week, I play tennis. I also set aside time each week for dinner or drinks with friends.
Building a startup is intense, but I’ve learned that maintaining a structured routine—and making time for family, health, and personal growth—is key to sustaining long-term success.
Editor’s Note
Rep AI is not just another chatbot; it monitors user behavior in real-time and provides assistance just when it is most required. Yoav Oz understands why e-commerce is struggling even if conventional retail is performing really nicely. Higher sales, less abandoned carts, and seamless shopping experience follow from this.
