UCDP #83 – AI Inventory Merchandising for Dealers w/ Sanjay Varnwal of Spyne AI

 In this transcribed episode of the Used Car Dealer Podcast, Zach talks with Sanjay Varnwal, co-founder and CEO of Spyne, about how AI-powered imaging and automation are changing vehicle merchandising—and why faster “time-to-live” listings directly impact dealer margins through holding cost.  

anjay shares how his background across product and engineering roles in e-commerce and travel helped him bring an outside-in playbook to automotive retail, then explains why Spyne focused deeply on one problem: helping everyday dealers compete with studio-level online merchandising—without the studio-level cost.

The conversation covers why inventory speed is everything, how premium visuals and instant publishing can compress go-live timelines, why Spyne expanded beyond photos into conversational AI (Vinny/Vini) and a more unified workflow, and where AI is truly delivering value vs. just hype. It ends with rapid-fire advice for dealers on reducing holding cost and tightening inventory discipline.

Zach: Thanks. All good. Zach here, and today’s guest is Sanjay Varnwal, co-founder and CEO of Spyne, an AI-powered automotive retail platform helping dealers create studio-quality vehicle listings and manage digital merchandising at scale.

Spyne has rapidly expanded into the U.S. dealer market, using computer vision and automation to transform how dealers photograph, market, and sell vehicles online. Sanjay joins us to talk about AI’s growing role in automotive retail and how dealers can leverage technology to move inventory faster. So Sanjay, thanks so much for joining me on the podcast today.

Sanjay: Thank you, Zach, for inviting me for the amazing podcast.

Zach: Yeah. So Sanjay, before founding Spyne, you worked in product and technology roles, including at Oyo.

What experiences there shaped how you approach building a startup in the automotive tech space?

Sanjay: Yeah, I mean, it is strange because I worked in product and engineering for all my career before starting Spyne, but not specifically in the automotive sector, right? So that also was very beneficial because I was able to bring in an outside-in perspective in this particular industry with the way the other industries were functioning, whether it was real estate or e-commerce. So I worked in some of the leading e-commerce firms in India, Oyo, which was a leading real estate firm, equivalent to Airbnb, then Expedia, equivalent of travel booking portal in India, and some of the e-commerce firms, including Amazon as well, globally. So I saw this entire internet commerce evolving from the very beginning in those last 12 to 15 years.

So I brought in a lot of new practices that these industries were following, which helped me also realize the kind of practices that could be brought in the automotive space. So we started with merchandising as a core statement in this horizontal e-commerce ecosystem, where the core goal was, can we get the seller to go online with the best quality visuals of their products on marketplaces, social media, their own websites, everywhere. And then gradually we figured out that, OK, we are able to make progress across all these industries, but all these industries are very, very deep on its own, like the way the visuals will be shot in real estate will be different than the model shoot, will be different than the food shoot.

And similarly, it will be different than the automotive shoot. So we decided to pick one category, and that was automotive, because we already were working with some of the leading auto marketplaces in India, some of the leading companies which were selling cars to the consumers in India. So we’ve got like very, very deeper exposure in the automotive sector, and this merchandising is a problem in India.

So when we were deciding on which sector to pick, which industry to go very deep in, so we zeroed on automotive due to X number of reasons, like US and Europe being extremely huge markets and completely online markets were one of the big decisions why we went into automotive. And also the ticket size, like so automotive is a fast moving inventory, ticket size is huge. And a lot of dealers, commerce is dependent on how beautifully they can showcase their cars online, like how premium is their merchandising, which Carvana, CarMax, some of these companies were able to do amazingly well.

So we thought that, hey, can we enable 99% dealers, which are not able to, I mean, afford those kind of studios for the photo shoots, can we enable them with this, this photo shoot, photo studio equivalent technology that can help them compete with Carvana or CarMax in terms of digital presence. And this worked out well. And that’s how my journey started, like, so from the, from the companies, which we started building for in India, we start, we took this particular product to Europe, South America, North America, Canada, some of these places, and then the journey started about three, four years back.

Zach: That’s really impressive. And just a ton of growth in three to four years, but kind of going back to that earlier point where you saw the auto market was the right opportunity. What was the early moment when you realized that AI powered imaging could really fundamentally change how vehicles are merchandised online?

Sanjay: So there are two, three aspects of this particular problem, right? So if you see, a dealer’s business is all about managing their inventory speed, right, the turnover time in which the inventory can be moved in their yards, right? So because dealer’s business is all about capital rotation business.

So let’s say they’re selling about 100 cars, they need to purchase 100 cars as well. But they can’t keep those cars for their own consumption, they will have to sell it. Because every day the car is not selling, they’re incurring loss.

That is called as holding cost in our industry, in the automotive industry, right? So typically, if you see, holding cost of a car is about $40 to $80 per car per day. And dealers do a margin of over 2,000 to 4,000 dealer dollars per car in general. So every day the car is not selling, dealers are losing $50 a day, right, which is cutting from their gross margin profile of that particular car, right?

So the idea was, how can we maximize the selling potential of that particular car in the earliest possible time? And we started looking at the factors.

And one of the primary factors was the dealer’s ability to go live instantly the moment the vehicle is acquired with premium-looking merchandising, right? So these were the two axes on which we started thinking deeply. How can we enable dealers go live immediately the moment the car is acquired to all the digital channels, which is marketplaces, social media, their own websites, everywhere. And with the kind of visuals, which is very, very premium in nature, right?

So if I give you an example, so let’s say you take Carvana, CarMax, some of these companies, the moment they acquired the car, they ensure that the car is live like instantly using some form of imagery or something.

And then they have all those mechanical studios in their yards, where the car is driven to, there’s auto shoot that happens, and the entire system is auto connected to their websites and all the platforms. And so they won’t take more than two to three days to get their inventory online with the best quality merchandise, which is shot in a studio.

If you see a normal dealer, the dealer, when they acquire a car, they do the inspection, then they do the reconditioning or refurbishing of the car.

And then they do the photo shoot using the photographers before going live. This entire process takes about, for the best of the dealers, about three to four days. But for an average dealer, it takes anywhere between 10 to 14 days.

And 10 to 14 days versus three to four days is a good like seven to eight days gap, where they are losing out on almost 400 to $500 of their margins.

So we started looking at, can we get the dealer go live with the inventory the moment they have acquired the car through stock imagery? Let’s say we have certain very realistic, almost near realistic kind of images that they can go live with like instantly with the marker that, okay, this is not the real image.

So right now, how it happens is dealers list their inventory without images or one very like a stock, like not the real image, but a one single stock image they go live with.

And when the consumer see those images or without images vehicles, they do not get any confidence. No leads are getting generated on those kind of vehicles that they go live with.

So we started creating near photorealistic images for those cars, even before the inventory is reconditioned, like reconditioning happened.

But that is the first shoot that we enabled the dealer to do with a wrap. And then after the first shoot, the reconditioning happened and then they did the final shoot.

So we compressed the go live time from seven to 14 days to about like maybe two to three days for a normal dealer.

So that impacted a lot of margins for the dealers because of which the product started working very well for us.

Zach: Really impressive. And for dealers hearing about Spyne for the first time, maybe listening, how would you describe what the platform does in the day to day workflow of their dealership?

Sanjay: Yeah, so we fundamentally are looking at this metric of cars getting sold in the fastest possible time.

This is the core metric that we are looking at. And we are building all the products targeting this particular metric. So we started with studio that started looking at this metric as a how do we get your inventory live in the earliest possible time so that you can start generating leads.

Those leads can get you enough opportunities to get them to closer. After this, after this product, we started looking at another leg of this particular problem, which was, hey, once your car is online, can we get this car generate more opportunity for you without you being online always, right?

So we introduced human like bots in voice and chat and email on the dealers like website profile, dealers marketplaces and dealers phone lines. So there’s a human like bot like you and me, which we call Vini, which has complete information about the car, its history, financing, and every other information that you want to know about that particular car, right?

As a consumer, when you tap on the button, hey, I want to know more information about this car, there’s another person like agent that can answer any of your question.

It knows you deeply. It knows the car deeply. It fetches all the information from inventory systems, CRM systems, DMSs and stitches to create a very, very seamless conversation with the customer.

And its goal is to get that customer book an appointment for that particular car and also does the follow ups to ensure that the appointment actually happens.

Zach: Why is it important to build more of kind of like the full automotive retail suite?

Sanjay: Yeah, so if you see today’s automotive ecosystem, right, the way the dealers use their softwares, these are isolated softwares that dealers use from different companies like Cox Automotive would be offering maybe VIN solutions, which does the CRM job for the dealers, it will be offering, let’s say vAuto, which is inventory system for the dealers, that will be offering different software.

So all these softwares are, are does the work like beautifully, but in its own domains, right. So it does, these softwares are connected with 20 different interfaces that the dealers would have to use to make their business function smoothly.

So we thought that here, I mean, all these softwares does the job amazingly well, when seen in isolation, but when you see like everything connected, the dealers have to apply a lot of mind share to make things work for them, right. And there’s a lot of training involved as well to make to understand the workings of different software, right.

So we thought, hey, why not stitch the experiences together in one cohesive unit that basically helps the dealer target this one fundamental metric, which is can I move the car faster in our lot, right.

So we started looking at which kind of softwares we should be building. So we built like merchandising, we built syndication. Now we after syndication, which is publishing the cars, cars across the platform, we connected that with conversational AI bots, which can do real time conversation with the customers, and engage that customers to bring the appointments to set the appointment and bring the customer to the store.

And so, so these, these, these are the puzzles that we are we are solving, which is letting the customer, letting the dealer solve this entire journey of sourcing to selling and servicing the cars. So we are trying to connect the dots, which will give the dealer a holistic experience to see at their business, right.

Zach: Like when I was building Selly, my CRM company, we had to do a lot of integrations, a lot of companies didn’t even have APIs, you know, we’re doing integrations at this point. And I’ve noticed the automotive ecosystem is far less integrated than some other, you know, verticals in the industry. And it’s kind of a two sided challenge.

So you’re building more of a retail suite, but you also have to still kind of integrate and touch points with some of the other like automotive systems. And how does your company do that?

Sanjay: See, in this entire industry, the dealers call the shots, right. So, so and because this entire software ecosystem is designed in a way where most of these large umbrella companies offer different software suites in isolation to the dealers.

So the dealers at one given point of time would use some software from company one, the other software from company two, the and so on, like they’ll be using maybe 30-40 different point solutions to do their jobs anyway. So this already happens in the industry, we are looking to capture this entire sourcing to selling journey of the car in one unified interface that may also fetch some information from the other, other, other mediatory software that the dealers are using, right.

So, and it is a dealer, right. So if they like your software, they will ask the other software vendors to integrate with you. Right. So that’s how the business function and that’s so we created this merchandising, which was way, way better than the any existing merchandising experiences, right.

So despite some of the bigger companies offering the existing merchandising solutions, dealers pushed us deeper into, into the bigger company software ecosystem and got us integrated, right.

So and this is this is like true for any other software solution that the dealers use, if they like it, they will make every single effort to use it. And dealers, once they tell the vendors, okay, I need to use this software, just expose the API, they will do it.

Zach: So that’s how the software started penetrating in the ecosystem. And you’ve expanded into the US market, you know, as you mentioned earlier, but I was curious, you know, as we go back a little earlier, what differences have you noticed in US dealerships and maybe like other markets and how they operate and kind of use your technology?

Sanjay: Yeah, it’s, it’s very interesting when when you see US versus any other markets in the world, right. So in the US, I think, I can give you an example of the differentiation between US and Europe markets, and Europe markets, maybe some other markets as well.

In most of the markets, we see OEMs calling the shots. Right. So, so the, so the BMWs, the Toyotas, the Hondas, some of these companies basically dictate how the dealers are going to use the software and sell the cars.

Right. So, and there’s a lot of influence that the OEMs have with the dealers on on on on not the influence, basically, there is a mandatory system that they have created with the dealers have to use in the other markets. Majority of franchise dealers in the other markets work this way.

Whereas when you see the US markets, it is the dealers who are autonomous in their in their business functioning right now. I mean, the OEMs recommend them, but it is it is completely onto the dealers which software they want to use.

Right. So and these are like giant companies, $30 billion enterprises. So you you have to work in partnership with them.

Right. So we have seen, I mean, US market, completely decentralized and run by some of the really large dealership groups, but but also having huge fragmentation. So almost 60 70,000 60,000 plus independent single dealers working in the in the US automotive ecosystem along with maybe 10,000 part of the groups. So that’s how the differentiation is.

And as I said, like the groups basically dominate the way the cars are sold. So that’s the one of the core differentiation that we have seen.

And, and the second is, of course, the because the business is so complex, buying to selling of the car every day matters because every day they are losing $50 a day on the car.

Right. So they, they have to manage this entire inventory stay time with with very high precision. They use a lot of softwares across the entire entire journey of the car buying to selling.

And this is where we see a we see huge penetration of softwares happening in this entire ecosystem as well. So we have giants like Cox Automotive, CDK, Solera, Reynolds and Reynolds doing billions of dollars of revenues just in automotive software space.

Right. So which is not true in the other markets right now in the world.

So, yeah.

Zach: And I remember when I said in NADA, almost every booth was advertising AI of some kind and AI is becoming a big buzzword, in automotive, retail and other ecosystems. Where do you think AI is genuinely like delivering value today? Whereas, you know, it might be hype in some cases. Just curious, your perspective.

Sanjay: Yeah, I mean, AI has really become a buzzword. You were right that NADA, they’re like two halls in NADA. One hall, which is the main hall, all the big companies were there.

And the second hall was termed as new age tech hall, where all the AI companies were established. We were having the booths. We were in the second hall.

But the buzz was amazingly high in the second hall. So the dealers, the consumers, the partners, everyone was looking at what is the new thing that has emerged in the AI space.

Right.

And buzz is like real, absolutely real. In the last one year, the trend has been multiplying. And the next three years, if you see, it will just like go deeper and deeper in the dealers ecosystem.

So next few years, and the use cases wise, I think AI will touch almost every part of the vehicle’s sourcing to selling journey and servicing journey.

So, but few places where I think AI will deeply impact is where the consumers, actually the dealers interact with the software to do certain things. Some of these things will become agent led, like the software will do the job on their own, because of the agent tech transformation of the software, and will let like complex decisions left for the dealers to take.

It will take like easier decisions that, okay, this is good to do. So it will do automatically all those things. It is not to do.

So it will not do like. So there are absolute positive decisions, absolute negative decisions. And then there are middle of middle tier decisions where they are not sure what to do.

So these are the decisions that they will leave to the dealers. Okay, so hey, I think you should do this. So dealers is now, I mean, they will take the decision.

So this is the agent tech transformation of most of the workflows of the dealer, where the dealers were interacting with the software will convert into AI led transformation. This is one.

The second could be this entire, this entire like BDC operations where the consumers are, consumers are, are, are, are connecting with the dealers and trying to get more information about the car and set up like real appointments for them.

And so this is one of the areas which is evolving so, so fast, where the, where the bots are becoming more and more human like. And it has as deep as information as anyone, like maybe 10x deeper information about the car, about the customers, right in fingertips. So it can hold that conversation in such a big depth that the car appointments, chances of appointments happening is, goes higher, higher up like multiple times, right?

So this is one of the areas that we see could become like a completely automated.

Of course, the selling in person will not change. So there’ll be like a people at the store who will, who will deal with these, these additional appointments and then convert those appointments into real sales. So there is this accessory selling happening, there’s this finance selling happening.

So these sales will keep on happening by the people, because cars will still be like a high ticket value, high ticket like a transaction. So, and most of the high ticket transaction need like a human interaction, at least like in-person interaction, test drives, in-person interaction. So some of these areas will be like, will be more consultative, in-person, will be not prone to AI attack, but anything which is digital will, will go through with transformative changes.

Zach: Yeah, I think. And I had, you know, two future questions for you. One, Spyne has talked about becoming almost like Shopify for automotive dealers in a sense.

What do you think that vision looks like longer term, like as we look out and, you know, Spyne kind of develop this platform?

Sanjay: So we are very much, as I said, we are very much centered around the core objective of the dealers, which is, which is reducing the turnaround time of the, of the vehicle, which ties up to the capital that they have invested in the car and in their business, right? So the more capital they have available to purchase more cars, they’ll be able to increase the transaction like buying and selling of this particular car more and more.

So we want to build, we want to rethink this entire ecosystem of dealer software and, and center around this particular core objective of can, how can they maximize their business by, by, by selling low, but, but extremely high demand cars, basically by buying low, extremely high demand cars, and then selling higher, maybe increasing appointments, maybe increasing leads.

So these all are the metrics around which we are looking to build the software and we want to be known as fundamentally the company which can help the dealer sell more cars faster.

So this is what we are looking to do.

Zach: And if you were advising a dealer today on where to invest in technology first, what would you prioritize?

Sanjay: I think most of the functions which are, which are right now getting bottlenecked because of the discipline or the availability are the, are the, are the, are the instant wins that the dealers could, could, could be, should be focusing on, right? So availability in terms of offline art calls, weekend calls, holiday calls, holiday meetings. So these are the areas where, where, where it is an instant win, like no brainer wins, because, because these are the lost opportunities you should be targeting and, and getting in.

Second is the, the areas which need like a very, very high operational excellence and discipline, which is probably the outbound, outbound ecosystem. Like, so dealers have tons of leads coming every single month for different types of cars that they could have in their inventory, right? So, and there are like a lot, many use cases or the dealers could engage with the customers, but they don’t have a means to engage right now, right? So, so the AI can not only target this entire lost opportunities in terms of inbounds, but also the hidden opportunities which are in the CRMs, which are in their DMSs, hidden, I mean, multiple layers behind the softwares right now, right? So these two are the obvious opportunities where I think these two are no brainers. They are not doing it right now.

And if something can generate like additional value out of it, it is a win-win for everyone.

Zach: Awesome. And my last round, rapid fire questions, just three questions, and I really appreciated your time today, Sanjay.

I’ve learned a lot kind of on the merchandising side. But first question, what’s one dealership process that AI will completely automate in the next five years, in your opinion?

Sanjay: I think all the digital processes that dealers are doing right now, BDC, I think is up for disruption big time. Yeah, all the call center operations are up for disruption big time.

Zach: Definitely. And what’s one mistake dealers make when it comes to digital merchandising?

Sanjay: I think they don’t really look deeper into this holding cost per day right now. So, losing every single day without the inventory going live is costing them $50 a day, right? So they are paying us maybe $3 to $4 for the merchandising right now, but they are losing like $50 every single day the car is not going online right now.

So this is like the extremely low hanging fruit that they should be capturing. We help them capture immediately.

Zach: And last rapid fire question, if you could give one piece of advice to dealers trying to sell more cars online, what would it be?

Sanjay: I think have a very strong grip on this holding cost that each of these vehicles are having right now, how many days the cars are sitting at the lot.

So divide all the cars into four buckets, cars at risk, cars at getting to risk, cars green, like maybe the entire timeline could be sold. The cars which are just entered to mid stage to the final stage after which. So this is one of the operational discipline that most of the softwares are helping the dealers focus on, but there are a lot of optimizations that the dealers can do to further improve this particular thing, right? So questioning every single thing that they do around this particular can help them improve their business significantly.

Zach: Very well said. And Sanjay, it’s just been a pleasure having you on the podcast today. One last thing I saw in the background, is that a Jeff Bezos quote on the wall?

Sanjay: Absolutely. So I used to work at Amazon. So Amazon is like the gold standard for customer centricity, right? So we are building our company with this particular ethos right now.

Zach: Love that. Well, thanks again, Sanjay, for being on the podcast.

Sanjay: Yeah, thanks a lot for inviting me, Zach. Thank you.