How A.I. Is Changing Car Buying Forever | Aharon Horwitz
发布时间 2024-01-11 10:00:32 来源
AI is making its way into car dealerships faster than ever before. But will we see robots on the sales floor? Or will the real impact happen behind the scenes? My guest today is our own Horowitz, the CEO of Fullpath, a cutting edge customer data platform that is helping dealerships innovate using AI. We discussed how AI is changing car buying, whose job may be up for grabs, how dealerships are predicting car shopper behavior, what will it be like to buy a car in five years from today, and much more? Don't forget to click subscribe so you never miss an episode.
AI正在以前所未有的速度进入汽车经销商。但我们会在销售场地上看到机器人吗？还是真正的影响会发生在幕后？我今天的嘉宾是我们自己的Horowitz，他是Fullpath的首席执行官，这是一个尖端的客户数据平台，正在帮助经销商利用人工智能进行创新。我们讨论了人工智能如何改变汽车购买，哪些工作可能被取代，经销商如何预测购车者的行为，从今天起五年后购车会是什么样子，以及许多其他问题。别忘了点击订阅，这样你就不会错过任何一集。
What's up everyone, this is Car Dealer Ship Guy. You're listening to the Car Dealer Ship Guy podcast, which is my effort to give you access to the most unbiased and transparent insights into the car market. But before we get into the show, this episode is brought to you by Autofyne. Autofy helps progressive dealers like you sell smarter, not harder, on your dealership website and now in your showroom too. Autofyse solves the everyday problems dealers actually face, like bottlenecks at the sales desk, customer distrust, and decision overload. And their all new showroom solution includes deal estimation, desking, lender routing, and an F&I menu. All of this in one powerful platform that bridges the gap between the CRM and the DMS. Dealerships with Autofyne can manage the floor more efficiently, fast track the yes, and make better lender decisions, enabling them to sell cars faster with higher satisfaction and more profit. In fact, deals with Autofyne taking average of 28 minutes from customer check into loan approval and dealers are making $411 more backend PVR per deal. Go to autofy.com slash CDG to learn more. That's autofy.com slash CDG and start selling smarter today.
大家好，我是汽车经销商的人。你正在收听的是汽车经销商的播客，我努力为你提供最客观透明的汽车市场洞见。但在我们开始节目之前，本期节目由Autofyne赞助。Autofy帮助像你一样进步的经销商在汽车销售网站和展厅中更加智能地销售，而不是更加努力。Autofy解决了经销商实际面临的日常问题，例如销售台上的瓶颈、客户的不信任和决策超负荷。他们全新的展厅解决方案包括交易估算、交易桌面、贷款人路由和F&I菜单。所有这些都在一个强大的平台上，弥补了CRM和DMS之间的鸿沟。使用Autofyne的经销店可以更有效地管理场地，快速追踪交易并做出更好的贷款决策，使他们能够更快地销售汽车，获得更高的满意度和更多的利润。事实上，使用Autofyne的交易平均只需28分钟，从客户审核到贷款批准，经销商每笔交易的后端PVR增加了411美元。请访问autofy.com/CDG了解更多信息。立即开始更加智能地销售吧。
Last time you were here, it feels like it was like a century ago when it comes to just the advancement of AI and just dealership technology. You brought tons of insight. It was super well received. And so I'm excited to dig back in. Before we get into AI and all that stuff, can you give me tables, have you explained to us what you do? What is full path and what do you actually do?
上次你在这里的时候，AI和经销技术的进步感觉就像是一个世纪以前的事情。你带来了很多见解，得到了极高的评价。所以我很兴奋地要继续深入探讨。在讨论AI和其他相关内容之前，你可以向我们介绍一下你的工作吗？full path是什么，你具体从事什么工作呢？
Yeah, full path is a customer data platform for car dealers and a marketing automation platform that lives on top of the customer data platform. So we call ourselves a CD XP customer data and experience platform. The experience being the way that we can use data to create great customer experiences and bring more shoppers and buyers to the dealerships.
是的，Full Path 是为汽车经销商提供的客户数据平台，同时也是在客户数据平台上运行的营销自动化平台。因此，我们将自己称为 CD XP（客户数据和体验平台）。体验是指我们可以利用数据来创建出色的客户体验，并吸引更多的购物者和买家来到经销商店铺。
Okay, so what is a customer data platform?
好的，那么什么是客户数据平台？
客户数据平台（Customer Data Platform）是一种用于整合、管理和利用客户数据的技术工具。它提供了一个集中存储、整合和分析客户个体数据的平台，使企业能够更好地了解客户并提供个性化的服务。
客户数据平台的主要功能包括数据集成、数据清洗、数据标准化和数据分析。通过将来自不同渠道和来源的客户数据进行整合和清洗，企业可以得到准确、全面的客户画像。这些数据可以包括个人信息、行为记录、交易历史和社交媒体互动等。通过对这些数据进行分析，企业可以获得深入洞察，了解客户的需求、购买偏好和行为模式，从而进行精准的营销和个性化推荐。
客户数据平台的目标是帮助企业建立更好的客户关系，并提供更满意的客户体验。它使企业能够更好地了解客户、预测其行为，并在不同的渠道上提供一致的个性化服务。通过客户数据平台，企业可以更加有效地管理客户关系，提高客户忠诚度和购买转化率，从而实现业务增长和市场竞争优势。
Oh, that is a great question. Or let me ask differently, right? If you take two dealerships, one has a customer data platform, one doesn't. What's the difference? So, a, the best metaphor in my mind is like that picture we've probably all seen the meme on Twitter of like the iceberg, what you see above the water and what's below the water. Right. So imagine a customer or dealership, okay, with like 150,000 records in their CRM, whatever, tens of thousands of their DMS with an advertising platform with an email platform with all that stuff. Right. And imagine two dealerships like that, exactly the same amount of data, okay, that sits in their system. All right. In one of them, they've done all the work to say, okay, like, here's a guy named Jack, a woman named Jill, right. What has she done? What does she like? What cars did she looked at throughout her history? What does she bought? What emails did she open? And then the other one, there's just records and there's the massive data, okay. The first dealership that hasn't organized underneath that dealership, there's just a huge amount of data structured and ready to go to make really personalized outreaches and engagements to Jill. And then the other dealership, they say have the same data, but they have no way to leverage it to give a personalized outreach. It's going to cause Jill to really want to come in and buy a car at a higher rate or service there. So very simply what a customer data platform is, it's like it goes to all the silos, it's like a truck, right, all the data silos, the CRM, the DMS. It picks up all the data. You know, once a day, once an hour, however often as you can up to real time, it takes it to a central processing place. It organizes and cleans it. And then it says, do all the marketing and the sales operations. Hey, here you go. Here's the information about Jill. Here's what she likes. Let's write the email this way. Let's have a conversation this way. Let's talk about the vehicle in that way. And it's a lot of software that is involved in doing that, but very simple. All it does it. It creates what they call like a, you know, a pretty clean data layer for the dealership. And then enables the dealership to do just smarter outreach and engagement.
哦，那是个很好的问题。或者让我换个问法，对吗？如果你拿两个经销商来对比，一个有客户数据平台，一个没有，有什么区别呢？好吧，我脑海中最好的比喻就像我们可能都在Twitter上看到过的那张冰山的图像，上面是水面上所能看到的，下面是水下的部分。对，想象一个有150,000个记录在他们的CRM中的客户或经销商，有数以万计的DMS、广告平台、邮件平台等等。想象两个完全相同数量的这样的经销商，他们的系统中都存有这样的数据。其中一个经销商已经做了所有的工作，可以说是，在经销商之下，数据是结构良好的，可以进行个性化的接触和交流。而另一个经销商，他们可能有同样的数据，但却没有办法运用这些数据来进行个性化的接触。这样做将使得客户从而真正想要到店购车或保养车辆的概率较高。所以，一个客户数据平台非常简单地说，就好像一个卡车，它会前往所有的数据孤岛，比如CRM、DMS，将所有数据收集起来。可以是每天一次、每小时一次，根据你的要求，甚至是实时更新的频率，将数据传输到一个中央处理地点。它将会进行组织和清理，并将信息提供给营销和销售部门。嘿，这是关于Jill的信息，这是她喜欢的东西。让我们这样写邮件，这样交流。针对这辆车，我们用这样的方式谈论。这涉及到很多软件，但很简单，它的作用就是为经销商创建一个干净的数据层，使其能够进行更智能的接触和交流。
I looked at your crunch base, saw that you launched in 2015. What happened in 2015 that spawned this opportunity for customer data platforms. And I have, I have a suspicion, but I want to hear from you.
我看了你们的crunch base，发现你们在2015年推出。在2015年发生了什么事情，使得顾客数据平台有了这个机会呢？我有一个猜测，但我想听听你的意见。
Like what changed in the world. We actually only came into automotive later, like 17, 18 we pivoted in. And you know, by 19, we were really becoming fully automotive oriented. But the whole space of customer data platforms actually emerged like a little earlier, like out of auto, I don't know, 2012 or something like that.
世界发生了哪些改变。实际上，我们在汽车行业方面起步较晚，大约在17、18年才开始转向该行业。而关于客户数据平台的整个领域实际上更早地出现在汽车业之外，大约是在2012年左右的时间。到了19年，我们真正开始全面面向汽车业了。
So what changed? What happened? So what happened was that there were all these disparate systems and a lot of them were like legacy repository style systems still, right? And, and you know, you had everything just sitting all over the place. CRM's had some data. Emails had some data. Email systems had some data. You know, MailChimp had some data, right? And there was a need to create a unified customer idea across all these platforms, right?
那么发生了什么变化？发生了什么事情？发生的事情是，存在着许多不同的系统，其中许多仍然是遗留的储存库式系统，对吧？你知道的，所有的数据都散布在各个地方。CRM有一些数据，邮件中有一些数据，电子邮件系统中有一些数据。你知道的，MailChimp也有一些数据，对吧？需要在所有这些平台上创建一个统一的客户身份。
And a lot of these systems that existed to do that at some point woke up and said, Oh, wow, we actually are the data layer. We are the core pub that has all the different insights. And we can actually leverage that for customers. So that was what emerged there. And you saw a really interesting motion out of auto.
很多早期存在的这些系统突然醒悟过来，意识到自己实际上是数据层。它们是具备各种不同见解的核心发布物，我们可以为客户利用这些见解。这就是那里出现的情况。你可以在汽车领域看到一个非常有趣的动向。
You saw like Salesforce go out and buy companies and set up these CDPs. And they bought Data Ramma in order to build connection connectors into databases. They built other things and eventually came out with a customer data platform that unified all the data silos. You saw the same thing with other with other vendors.
你可能见过Salesforce这样的公司外围并收购其他公司来建立这些客户数据平台（CDP）。他们购买了Data Ramma以便将连接器链接到数据库。他们还建立了其他功能，并最终成功推出了一个统一了所有数据孤岛的客户数据平台。你也可能在其他供应商那里见过类似的情况。
So it didn't really exist in auto. And we were sitting around the office in 2022. We just read a book called a few of us here called Play Bigger. I don't know if you read that, Play Bigger. It's a good book. It talks about like categories as defining a marketing initiatives, kind of create a category.
所以在汽车行业中它并不存在。我们在2022年坐在办公室里时，刚读完了一本叫《我们中的一些人都在玩大》的书。不知道你是否读过，《玩大》是一本很好的书。它谈论了类别如何定义营销计划，创造一个类别的概念。
And we had this platform that was doing a lot. It was doing marketing automation. It was doing data collection. It was the most connected platform in the industry as far as we knew. And it actually had the levers to do stuff with it. So you didn't just have the data. You could take the data and send a smart email automatically or make smart ad or whatever it might be. And we were like, huh, we're just getting lost in all the noise. Maybe we could create a category.
我们有一个非常强大的平台。它能进行市场自动化，进行数据收集。据我们所知，它是行业内最具连接性的平台。并且它实际上可以通过一些控制手段来进行操作。所以你不仅仅拥有数据，还可以利用这些数据自动发送智能电子邮件或制作智能广告，或者做其他任何事情。于是我们意识到，在这些嘈杂的信息中我们正变得越来越迷茫。也许我们可以创造一个新的类别。
And actually, I don't like our VPC sales at the time and Elana, Shop Tire, VPC marketing, we were sitting around. And Mike was like, oh, we keep talking about CDPs out of auto, like, you know, tealium and blue shift and all these CDPs. Why don't we just bring CDP to auto? And I think Elana was like, all right, it's be a CDX be customer data experience platform. And that was it. We were like, great, this works. We announced it. And, you know, released the CDP. And that's what we've been doing.
实际上，我不喜欢当时我们在VPC销售部门和Elana、Shop Tire和VPC营销部门围在一起时的情况。迈克说，我们一直谈论着关于自动化的CDP，比如tealium和blue shift以及其他的CDP。为什么我们不把CDP引入到汽车行业呢？我想Elana说，没问题，就将其称为CDX——客户数据体验平台。我们当时就觉得这个主意很好，于是宣布了这个决定并发布了CDP。这就是我们一直在做的事情。
So pretty much the way I look at it is like tech adoption in the dealership space, right? It was horrendous for many years. It's gotten a lot better. You see the opportunity of all this tech stuff. You say, hey, let's extract this data and leverage it to make the sales process more efficient, make it more profitable, reduce expenses, enhance customer experience.
我个人的看法是像车商领域的技术采用一样，很多年来都非常糟糕。不过现在已经有了很大的改善。你会发现所有这些技术的机会。你会说，嘿，让我们提取这些数据并利用它来使销售过程更高效，更有利可图，降低成本，提升顾客体验。
Right. So like, if I'm a customer in 2010 versus a car buyer today, right, and correct me from wrong here, please, I can get a much more personalized message across the entire ecosystem, whether it's to buy a new car to service my car. It was something along those lines, but super personalized to me, you know, given my mileage on the car, maybe where I live, my, who knows what. I mean, you could tell me where the limit is drawn, but that's fundamentally what is happening today.
没错。就像，如果我是一个2010年的顾客和如今的汽车买家相比，对吧，请纠正我如果有错，我可以在整个生态系统中获得更个性化的信息，无论是购买一辆新车还是为我的车辆提供服务。这与之相仿，但是对我来说是超级个性化的，你懂的，根据我的汽车里程、可能的居住地、我的种种情况等等。我是说，你可以告诉我哪里有限制，但这基本上就是今天正在发生的事情。
Yeah, when it's done, right? Any data that the dealership has gathered, that's legitimate to be used, right? It's opted in in some way. Plus any third party data that's legitimate to be used can be put into some sort of mix to create a better marketing and sales experience. And I'll give you a couple examples.
是的，等它完成时，对吧？任何经销商收集的数据，都可以合法使用，对吧？它会以某种方式选择加入。此外，任何合法使用的第三方数据可以混合在一起，以提供更好的营销和销售体验。我给你举几个例子。
I was talking to one of our longest, uh, stand customers, uh, I got him, uh, Steve Guebara. He's the GM of a dealership called Zott Ford in Detroit. And he was telling me this crazy story. He said, I always had this dream, you know, since like Syrian, and, and okay, Google, uh, Google, Android, uh, emerge right Apple Syrian, and then the ability to talk to Google that just like I asked my phone sometimes. Hey, like, how much that was on the ground or how long is it going to get to work or what's the temperature outside. And it gives me an answer. That's a good answer.
我跟我们最老的一位合作伙伴之一进行了交谈，他是史蒂夫·古埃巴拉。他是底特律的一家名为Zott Ford的经销商的总经理。他告诉我了一个很疯狂的故事。他说，我一直有一个梦想，就是自从有了像Siri这样的语音助手以及Google和Android的出现后，可以像问我的手机一样向Google提问。比如，我会问手机，外面有多热，或者去上班需要多长时间之类的问题，然后它会给我一个准确且满意的答案。
I can also ask it questions like, Hey, it's the last day of the month. Who wants to buy a card today? Right. And I want my phone to tell me, Hey, Paul, these four people. And you can make a sale. Okay. So. He was actually in Israel where we're all of our R. D. is and he was hanging out with us. He told this story to one of our product managers, a guy named. And I don't know, five, six months later, Alec calls him and says, Hey, Steven, open up your email. And Steven's like, okay, why? He's like, I just sent you your phone's answer to the question of who's going to buy a card today. Okay. And so, Steven opens up the email and he sees a list of five people. He says, what is this? Alex says, I used our algorithms, all the data we're collecting from all your systems. And I can tell you that these people we haven't talked to in forever, we have shown no visible signs to you of wanting to buy them. We'll buy a car in the next week or two or already buying part today. I don't know. So, Stephen's like, okay, let me try it. So he gets one of his sales guys to call them. And according to the way Stephen told me, he's like three, they couldn't reach two. But three of them were bull's eyes. They were absolutely either just bought a car within the past 48 hours or replying to buy a car in the next two, the next two weeks. And so that level of just insight into the data where you can build a picture and create an essence like a new category of lead.
我还可以向它提问，比如说：“嘿，今天是月底，谁想今天购买一张卡？”对，然后我希望我的手机告诉我：“嘿，保罗，这四个人。你可以完成销售了。”好的，所以他实际上在以色列，在我们所有研发工作都进行的地方，他与我们一起度过了时间。他把这个故事告诉了我们中的一位产品经理，一个叫作的人。我不知道，大约五、六个月后，亚历克给他打电话，说：“嘿，史蒂文，打开你的邮箱。”史蒂文问：“好的，为什么？”亚历克说：“我刚刚给你的手机回答了今天谁要购买卡的问题。”史蒂文就打开了邮箱，看到了一个名单，上面有五个人的名字。他问：“这是什么？”亚历克说：“我使用了我们的算法，从你所有系统收集到的所有数据，我可以告诉你，这些人我们很久没联系过，你们也看不到他们有意愿购买车辆的迹象，但他们将会在未来一两周内购买或已经购买了车辆。”史蒂文说：“好的，让我试试。”于是他让他的销售人员给这些人打电话。根据史蒂文告诉我的方式，有三个人无法联系上，但其中三个命中了目标。他们要么刚在过去48小时内购买了一辆车，要么准备在未来两周内购买一辆车。这种对数据的深入了解让你可以绘制一个画面，并创造一种新的潜在客户类别的本质洞察力。
Now, dealers have, as you know, they have phone up and they have showroom up and that, you know, this up concept, which I guess was an old concept of like, Hey, someone's walking in the store, let's get up. Right. I guess that's what it was next up. So they have these up concepts. They call the lead and up. There's like a showroom up, which is someone who walks in the door. There's a phone up. Someone who makes a phone call and then there's an internet up. Someone who fills out an internet lead. Suddenly, there's a CDP up, a customer data platform up. We've done alpaby. We've taken data. We've turned into actual, actionable intelligence on someone who might want to buy a car. And that's not just good for the dealership. It's actually good for the consumer.
现在，经销商们，就像你所知道的那样，他们通过电话和展厅与客户联系，这种更新的概念，我想是在店内有人走进来时，他们就与客户联系。对吧，我猜这就是下一个概念。所以他们有了这些概念。他们称之为潜在客户和来访者。有展厅来访者，也就是走进门的人。有电话潜在客户，也就是打电话咨询的人。还有互联网潜在客户，也就是填写网上咨询表格的人。突然之间，又有了顾客数据平台潜在客户，我们通过收集数据，将其转化为关于可能想买车的人的可操作的智能信息。而这不仅对经销商有好处，对消费者也是有益的。
That's actually the first thing I thought about. I mean, as a customer, it's much better for me kind of doing the work for me. Listen, if you call and say, Hey, I'm just checking in. Want to just say hello. I haven't talked for a while. Happy to share anything we've got going on in our specials or we're not that person might appreciate it. If you call and try to like pitch them hard, we probably want to appreciate it. So again, it depends on how the dealership approaches it. But I do think that's the baseline of like why the data works. Now, you then take it to the next level. You say, Okay, great. Like we can give these insights, but what dealership has the time to act on all those insights or what dealership has the human power to act on all those insights, very few or none.
这实际上是我最先想到的事情。我的意思是，作为一位顾客，如果我能够自己完成这些工作，那会更好。听着，如果你打电话说：“嘿，我只是来看看，想要问候一下。我们已经有一段时间没说过话了。很高兴分享我们特别推出的任何信息，或者我们不太确定那个人可能会感兴趣。”如果你打电话并试图强行推销，我们可能会不感兴趣。所以再次强调，这取决于经销商的做法。但我确实认为这是数据能发挥作用的基本原因。而且，你可以进一步提升。你可以说：“好的，我们可以提供这些见解，但是哪个经销商有时间去处理所有这些见解，或者哪个经销商有足够的人力来处理这些见解，很少或没有人。”
So then you say, Okay, email programs. I don't know. There's a lot of them in our industry, right? Friends of ours competitors like outsell, like four eyes, right? So, you know, great companies, full path. We have something called audience activation. You have these email programs. I know that, you know, Sally is interested in these five cars. I know that Sally has this amount of equity in our loan. I know when our last payment date. I know everything that the dealership has allowed to know. And I can then use that to construct campaigns with certain amounts of urgency in emails and I can actually write those emails automatically and inject all sorts of relevant information. And those can go out with full automation and the dealership can just wait and collect those leads.
那么，你说，好的，电子邮件程序。我不太清楚。我们行业有很多这样的程序，对吧？像Outsell、Four Eyes这样的竞争对手公司。这些都是很棒的公司，经验丰富。我们有一个叫作"受众激活"的东西。你有这些电子邮件程序。我知道，Sally对这五辆车感兴趣。我知道Sally在我们的贷款中有多少股权。我知道最后还款日期是什么时候。我知道经销商允许了解的一切信息。然后，我可以利用这些信息，构建带有一定紧迫感的电子邮件广告，我可以自动撰写这些邮件并注入各种相关信息。这些邮件可以全自动发送出去，经销商只需等待并收集这些潜在客户。
Then I say, well, hey, Sally's about to walk in the door. And there's a salesperson sitting there. I can suggest here's four cars that Sally's visited and look like her current deal that might be a zero payment raise flip, you know, something really nice. And more than that, here's a talking points that compare her existing car to the cars that you've kept all that already in your system. We're not even talking like future. This is existing. But how? Like, what is that information? What is the source of all this information? So, a, you know, like, it depends how you want to approach it. Like, we do different things.
然后我说，嗯，嘿，Sally马上就要走进门了。旁边坐着一个销售人员。我可以建议这里有四辆汽车，Sally已经去过，看起来像她当前正在谈的那笔交易，可能是一个零付款升级的机会，你知道，非常不错的选择。更重要的是，这里有一些论点，可以将她现有的汽车与你们系统中已有的汽车进行比较。我们甚至不仅仅是在谈论未来，而是现有情况下的信息。但是，这些信息来自哪里？因此，你可以根据你的喜好进行不同的方法。
So one, the source of all the information is the cleaned up version, the hygienized version of the CRM, of the DMS, of website track, website browsing data and form fills, all the stuff that happens going on, all the data information that's created. That's what we have to be really good. So full path, we've never announced this. We'll probably announce it in an ADA, but anyone who's listening will get an early announcement. We're connected. I think we just counted last week. So this might be out of date. We have 198 data sources that we can currently ingest, right? Websites, CRMs. Not all of those have APIs, APIs, computer to computer, talking languages, right? But we figure out ways to get the data in.
首先，所有信息的来源都是经过清理的版本，即经过清洁化的CRM、DMS、网站追踪、网站浏览数据和表单填写等，即发生的一切，所创造的所有数据信息。这就是我们需要非常熟练的地方。整个路径，我们从未宣布过这一点。可能我们会在ADA上宣布，但是任何倾听的人都将得到提前公告。我们已经连通了。我想我们上周刚刚统计过。所以这可能已经有点过时了。目前我们能够接受198个数据源，对吧？网站，CRM。并非所有这些都有API，即计算机之间的通信语言，对吧？但我们找到了一些方法来获取数据。
Once you have that data in, you have to decide what to do with it. Some of it's just basic stuff. If a column says first name and then the next one says first underscore name. Those are both first names. Put them together. You know, you have, you know, the same person. If it's the same, you know, first name, email, phone, et cetera. But some of it's much more complex, a typo and an email. Is it the same person or is it a different person? And there's all these ways to resolve those things. We use all these different algorithms we've developed and others have developed.
一旦您获取到这些数据之后，您需要决定如何处理它。其中一些只是基本的内容。例如，如果一个列写着“姓氏”，而下一个列写着“姓氏下划线”。它们都是姓氏。将它们合并在一起。您知道，它们表示同一人。如果涉及到相同的名字、邮箱、电话等信息。然而，有些情况更为复杂，比如邮箱中的拼写错误。这是同一个人还是不同的人？解决这些情况有多种方法。我们使用了我们自己研发的算法以及其他人开发的算法。
When you want to do stuff that's like proactive, you want to say give talking points to a salesperson, they're going to probably use GPTs. Because the generative capabilities of a GPT where you can give it information about a vehicle. You can give it trim and give it all the stuff. You can say, hey, compare it to this. And then output some talking points about where the benefits or the disadvantages are. The GPTs are pretty good at that. We have found. So that's the type of stuff where you're going to use a GPT. So everything has its own little tool box that you're going to want to use.
当你想进行积极主动的事情时，比如给销售人员提供谈论要点，他们可能会使用GPTs。因为GPT具有生成能力，你可以给它关于车辆的信息，比如配置等等。你可以要求它与另一辆车进行比较，并输出一些关于优势和劣势的谈点。我们发现，GPT在这方面表现得非常好。所以这就是你使用GPT的情景。每件事情都有它自己的工具箱，你会想要使用它们。
I'd have to imagine that like the AI, just the AI, the rise of AI has been very quick. At least, let's just say the way it's gotten mainstream. So you know, Chad GPT and all that. I'd have to imagine that for you, that's like just the golden opportunity. Given what you're doing and you're adjusting all these data sources, how are you? I mean, two part question here. Like, how are you leveraging AI today? And then how are you going to leverage it in the future to just do all the everything you're doing currently, but better, faster. Like, tell me about that.
我必须想象，就像人工智能一样，单单人工智能的崛起是非常迅速的。至少我们可以说，它的大众化进程是如此之快。你知道的，查德GPT等等。我必须想象，对于你而言，这就像是一个黄金机会。考虑到你正在处理和调整所有这些数据源，你是如何利用人工智能的？我的意思是，这是一个两部分问题。首先，你如今如何利用人工智能？其次，你将如何在未来利用它来使你目前所做的一切变得更好、更快。告诉我一些相关情况。
Yeah, well, it's a great question. I mean, I actually think the what's gotten most of the buzz has been the new type of AI, or the generative AI. What does that mean? So up until now, there's lots of forms of like what people call AI, which is like, you know, just a way to use computer algorithms to take a look at different problem sets and data sets and try to like, you know, make an intelligent prediction that maybe isn't obvious from the data, right? And there's all these algorithms, you know, collaborative filtering algorithms. There's random forest algorithms, right? There's all these fancy mathematical frameworks that aren't really important for us to talk about. So we've been using a lot of those. And a lot of those are also off the shelf. I'm not saying we, we like are doing anything particularly special. We just use them. They're well known algorithms.
是的，这是个很好的问题。我是说，我实际上认为引起了最大反响的是新型人工智能，或者说生成式人工智能。那是什么意思呢？直到现在，有很多形式的被称为人工智能的方法，就像是一种使用计算机算法来查看不同的问题集和数据集，并试图从数据中得出一个可能不明显但智能的预测的方式。有各种各样的算法，协同过滤算法，还有随机森林算法。有很多这些高级的数学框架，但对我们来说并不重要。我们一直在使用很多这样的算法，而且很多也是现成的。我并不是说我们做了什么特别的事情，我们只是使用它们而已。
So, for example, a simple example would be, you know, the way that Netflix or Amazon can suggest what other things you might want to buy. They might use what's called a collaborative filtering algorithm. So we've been using those things frankly since the company was founded. We've had to, we've used those to suggest, you know, cars in our chat, or we use those, but all these different tools. The one that's creating all the buzz is like a new form, or it's not new, but it's called generative pre-trained transformers, right? What they call generative generative AI, it's based on what are known as neural networks, which is an AI model, it's meant to mimic the way the brain works, and it comes from like the field of deep learning.
所以，比如说，一个简单的例子就是，你知道，Netflix或Amazon可以推荐你可能想买的其他东西。他们可能会使用所谓的协同过滤算法。事实上，自从公司成立以来，我们一直在使用这些工具。我们用它们来推荐聊天中的汽车，或者使用它们来进行各种不同的工作。目前备受关注的是一种新形式，或者说并不新，但被称为生成式预训练转换器。它被称为生成式AI，基于所谓的神经网络，这是一种模拟大脑工作方式的人工智能模型，来源于深度学习领域。
And so when you look at that, that actually existed quite a long time ago, meaning they were trying to build these different, you know, ways to handle uncertainty and have computers actually like say the right thing, even in an environment where it doesn't know all the answers. One of the earliest attempts at this was like to create a language, a simple language that had never existed before, to train a computer on it, and then have it almost like expand the language in a logical way that would work, right? So that's been going on for a while. In 2017, Google researchers published a paper about GPTs, and that sparked this like new generation, right, which came to fruition most recently with OpenAI's GPT.
所以当你看到这个时候，实际上这种情况很早就存在了，意思是他们一直在尝试构建各种处理不确定性的方法，并让计算机在不知道所有答案的情况下说出正确的话。其中早期的尝试之一就是创建一种从未存在过的简单语言，让计算机对其进行训练，然后让它以一种符合逻辑的方式来扩展这种语言，以使其有效。所以这个已经持续了一段时间。在2017年，谷歌的研究人员发表了关于GPT的论文，引发了这个新一代的研究，最近由OpenAI的GPT实现了这个目标。
So we actually find this to be really, really powerful, and we think it's going to change a lot of things. You know, as we experienced it, it has transformed, for example, hundreds of dealers that are using our GPT-based chat for consumers. They're, you know, we're processing every day thousands and thousands of shoppers, giving them great conversations, experiences, getting them in as leads.
我们实际上发现这真的非常非常强大，我们认为它将会改变很多事情。就像我们亲身经历过的一样，它已经改变了数百个使用我们基于GPT的消费者聊天工具的经销商。他们每天处理成千上万的购物者，为他们提供良好的对话体验，将他们转化为潜在客户。
You know, as was famously covered on Twitter, we had a group of pranksters come and also do silly things with it. But I think that's sort of the anomaly. That's like the fun buzz around a new technology, and that's going to fade away. We're seeing like everyone adopting GPT for chats because you really can let normal shoppers have a great experience and, you know, and whatnot. So we see it like in a consumer-facing way. That's cool.
你知道的，就像在Twitter上引起了轰动一样，我们曾经有一群恶作剧者也来用它做一些傻事。但我觉得这只是个例外，这就像是对新技术的有趣关注，而这种兴奋感会渐渐消退。我们看到大家都开始采用GPT进行聊天，因为你真的可以让普通顾客获得很好的体验以及其他种种好处。所以我们看到这种应用在面向消费者的方面非常酷。
Transformative is going to be in the way it works for businesses behind the scenes, and we're doing a lot there. Tell me more about that, though. How is it going to push the envelope in this industry, right? Like we have people have consistently, you know, kind of joked about the auto industry that's so far behind in tech, right? Now AI is sort of at the forefront of all this kind of technological revolution. I mean, shit, I'm using AI in things I'm doing. I'm testing like my social clips, right? Like we've trained my voice so that when we do my social clips, I can actually, you know, write a quick hook and use it with AI. Now it's not perfect yet. Like I'm still recording it myself. But the point is it's like it's crazy the capabilities that I see. It's actually bringing real efficiency.
在商业运作的背后，Transformative将以一种方式改变，我们在这方面做了很多工作。不过，你可以告诉我更多吗？它将如何推动这个行业的创新边界呢？就像我们一直在开玩笑说汽车业在科技方面远远落后一样，现在人工智能正处于这场技术革命的前沿。我的意思是，我正在使用人工智能进行一些工作。例如，我正在测试我的社交视频剪辑。我们已经训练好我的声音，这样在制作社交视频剪辑时，我可以快速编写一个吸引人的开场白，并与人工智能一起使用。当然，现在它还不完美，我仍然需要录制自己的声音。但重点是，我看到了它的惊人功能，它真的带来了高效率。
So tell me, like how does that actual actually parlay to the automotive industry specifically? When it comes to like dealers, I like, let me pose a positive this question to you. When do you think a dealer is going to hire its first AI employee? You laugh at that. You may not. No, no, I'm not laughing. I think that that AI is coming quite soon, and you know, we're all going to be part of that. But what I'll tell you is we got pitched by some startup here that offers AI developers is in their AIs that write code. You train it, you list that. The other thing that will work in your code repository, it's insane.
那么告诉我，这实际上是如何与汽车行业联系起来的呢？当涉及到经销商时，让我向您提出一个积极的问题。您认为经销商什么时候会聘请第一位人工智能员工？您可能在笑，也可能不是。不，不，我没有笑。我认为人工智能很快就要到来了，而且我们都将成为其中的一部分。但我可以告诉您的是，我们被一家初创公司推销了一些提供人工智能开发者的服务，他们的人工智能可以编写代码。您对它进行训练，然后将其列入您的代码存储库，这太不可思议了。
So if I want to create an AI agent that can do, for example, inventory analysis for dealers and work with the new car director or with the use car director, and they can literally talk to it, ask questions and it's sitting on not just the dealers own inventory, but frankly, like a national set or an OEM set, right? You can pull all the inventory that's on the boats and on the trains, all the inventory is on the ground, and you can use all sorts of ways to decipher all the codes of the different pieces that can really be an asset to the inventory manager. So that type of thing we see happening quite soon, there's all different sorts of different ways that's going to take place. So I actually think dealers are going to be open to it in particular internally, and then there's the more innovative dealers who want to also use it externally and help customers and leverage it for customers. So, you know, it's going to be a little bit of a stop start adoption, but it's inevitable that it will be adopted.
所以，如果我想创建一个能够为经销商进行库存分析并与新车总监或二手车总监合作的人工智能代理，他们实际上可以与其交流，提问，并且它不仅局限于经销商自身的库存，而是可以涵盖全国范围或原始设备制造商的设备。你可以获取所有船上和火车上的库存以及地面上的所有库存，并使用各种方式解密不同产品的代码，这对于库存经理来说是一个真正的资产。我们预计这种情况很快就会发生，有各种不同的方式将会出现。所以我认为经销商在内部尤其是会对此持开放态度，而且还有一些更具创新性的经销商希望在外部使用它来帮助客户并为客户提供支持。所以，它的实施可能会有一些起伏，但无疑它将会被采用。
But where do you think is going to be the biggest impact, right? Like when you look at the opportunities today, like do you think the biggest impact will be with consumers or will it be behind the scenes for me, you know, as a dealer or a business in general, I'm going to be leveraging it for decisions. I think that pound for pound, the biggest impact is going to be in the next five to six years for the business behind the scenes, but I think ten years down, it's going to be for the consumers. So it's not going to be operating in a world where you either have personal agents, you can set up AI searches, you can have AI do things for you, and that's going to be quite transformative. Again, I'm not like a huge booster of technology, and you know, it's just obvious when you are in our space that this type of tool is going to be adopted. It's hard to see like how you can see it differently once you're actually building stuff with it. And that's, you know, that's I think a pretty powerful statement. So we're going to see that.
但是你认为会有最大的影响力的地方在哪呢？就是说，当你看到今天的机会时，你认为最大的影响力会在消费者身上还是在幕后，比如像我这样的经销商或者一般的企业，我会利用它来做决策。我认为在未来五到六年里，对于幕后的企业来说，影响力会是最大的，但我认为十年后，会是对消费者的影响力。所以，它将不再运行在一个你可以拥有个人代理，你可以设置AI搜索，你可以让AI为你做事情的世界，这将是非常具有变革性的。再次强调，我并不是一个技术的狂热者，但当你在我们这个领域里时，明显可以看到这种工具将会被采用。一旦你真正开始用它来构建东西，就很难以不同的方式看待它。这是一个非常有力的陈述。所以我们将会看到这一点。
I will say, and this is like our message to dealers right now, is that if you want to be ready for that, the most important thing is to get your data house in order. So most dealers, if you walk up to them and ask them, Hey, what's your data strategy? Right? They don't really think about that. They don't have an answer for that. That's obviously outside of auto. Like the only thing people are thinking about right now, right? What's my data strategy?
我要说的是，这就像我们给经销商们的一个信息一样，如果你想要为未来做好准备，最重要的事情就是整理好你的数据。所以，大多数经销商，如果你走过去问他们：“嘿，你们的数据战略是什么？”他们通常不会考虑这个问题，也没有答案。这显然与汽车行业无关，因为现在人们唯一关注的就是：“我的数据战略是什么？”
So you need to be able to figure out how do you get your data organized? How do you get it in in a proper format? How do you clean it up? And then once you have that set, you can use AI or you can use automation, right? Which often is running some maybe prior version of or some other version of AI machine learning to do magic. And it takes a lot of the doing load off the people and lets them focus on executive decision making and creativity and to get inside the box.
所以，你需要能够弄清楚如何整理你的数据？如何以适当的格式导入数据？如何清理数据？一旦你整理好了数据，你可以使用人工智能或自动化来处理，对吗？这通常是运行一些之前的版本或其他版本的人工智能机器学习来完成神奇的工作。这样可以减轻人们的工作负担，让他们专注于执行决策和创造力，并深入了解事物。
So just as an example, if I have a challenge, you know, I have a, let's say an AI agent in the future. It's monitoring my systems. It notices that maybe a competitor is doing something. Right? It comes to me and says, Hey, I see this happening. Here's what I suggest. Should I do it? Yes or no? You say, yes, boom, new ads are created, new emails are sent out, new analyses are created and you're able to execute with like a lot of intelligence and speed. So the other thing that we talk a lot about in full path is like sensor to solution loops and so I guess like this is where like our, you know, unfortunately, you know, the like military analogies come in, right?
所以，举个例子，假设我遇到了一个挑战，比如说未来有一个人工智能代理在监控我的系统。它注意到竞争对手正在做某些事情。然后它来告诉我，嘿，我发现这个情况。这是我的建议，我应该去做吗？你说是，然后新的广告就会被创建，新的邮件被发送，新的分析被生成，你可以以高智能和高速度来执行。所以，我们在全路径中还经常讨论的另一个问题就是传感器到解决方案的循环，我想这可能就是我们不得不使用军事类比的地方了，不幸的是。
But like if you, if you let's give you, here's a random example. If someone's, you know, launching a rocket somewhere in the world and you need to take that rocket out of the sky, how fast can you see that rocket launch and trigger something to take it out of the sky? Okay. So we actually have people who work or whose jobs, you know, we're building algorithms to take rockets out of the sky. Right? That's the type of thing that, you know, you have to think about those things.
但是，就像如果你，如果你让我们给你举一个例子。假设有人在世界上某个地方发射火箭，并且你需要迅速发现火箭的发射并触发某些操作来将其击落。好的。实际上，我们有人在从事这样的工作，他们正在构建算法来攔截火箭。对吧？这就是你需要考虑的那些事情。
So anyways, the point is what we look at a lot is when there's an anomaly in the data or there's something happening. Do we have a way to surface it for the dealership and quickly apply a solution? Simple example. A good equity position buyer showed up on the website. She previously opted in. So, you know, we have, it may be done something. So we know that that's happening. The dealership does it. But the person had opted in. They, you know, submitted their lead at some point in the CRM or what have you. And now they're on the website and they're in a good equity position and they're looking at a car that we want to sell. Can we really quickly close that loop for the dealership and both make sure a salesperson knows. Make sure call happens. Make sure the right email goes out. Make sure they're in the right at audience. That's a very important loop to close because that really could be a sale at a very high likelihood.
简而言之，我们经常关注的是数据中的异常情况或者某些事情发生时，我们是否有办法将其及时反馈给经销商并迅速解决问题？举个简单的例子，一个有良好资产价值的买家出现在网站上，她之前选择了订阅。因此，我们对此有所了解。经销商也了解这个情况。但这个人已经选择了订阅，他们在某个时刻在CRM系统中提交了线索。现在他们在网站上，并且对我们想要销售的汽车感兴趣。我们能否快速地为经销商关闭这个循环，并确保销售人员知晓情况，确保电话联系进行，确保发送正确的电子邮件，确保他们位于正确的目标受众中。这是一个非常重要的循环需要关闭，因为这很有可能成为一笔非常有可能的销售。
So that's, you know, a simple example or complex examples might be, wow, it's interesting. The cost per lead on ionics is going up in the market. Can we see whether competitors may be doing something and do we need to go compete or can we find a way to ride someone else's investment in the market and take their, you know, take, take, get benefit of their traffic. Of their spend to kind of wake up awareness about ionics by doing certain email campaigns or certain other advertising campaigns that will enable us to get that that interest and capture it at a lower cost for our, for our client.
所以那就是，你知道的，一个简单的例子或复杂的例子可能会让人惊讶，很有趣。在离子类产品方面，每个潜在客户的成本正在市场上上涨。我们能否查看竞争对手是否在做某事，我们是否需要去竞争，或者我们能否找到一种方法来乘坐市场上其他人的投资，并利用他们的流量获益，通过某些电子邮件活动或其他广告活动来唤起人们对离子类产品的兴趣，以更低的成本为我们的客户捕获这种兴趣。
Another example would be, you know, data showing that there's a certain increase in the velocity of a certain model that's being sold. What does that mean? What do we need to do? Can we trigger certain things automatically, right? Do we need to drop certain, you know, prices raised certain prices. So all that, all that stuff should happen as much as possible quickly automatically with the dealer knowing what's going on able to make decisions where necessary.
另一个例子就是，你知道，数据显示某个型号正在售卖的速度有一定的增长。那意味着什么？我们需要怎么做？我们能否自动触发某些事情呢？我们需要调低某些价格，提高某些价格吗？所以所有这些，尽可能地快速、自动地发生，经销商知道正在发生的事情，并能根据需要做出决策。
What's been the industry's appetite on this? Like I'm listening to you speak and it's fascinating just how things are changing and evolving. What are you seeing as far as dealers, just the automotive industry in general? What is the appetite for all this new software, all these new tools? Like are people, are people listening? Like I'm listening to you right now and I'm just, you know, intellectually curious. This like fascinates me.
关于这方面，业界的态度如何？当我听你说话时，我着实被事物的变化和演进所吸引。对于经销商和整个汽车行业而言，你见到了什么情况？对于所有这些新软件和工具，人们的兴趣如何？他们在倾听吗？就像我现在在听你说话，我只是感到好奇，这让我着实着迷。
Dealer, what's the sell cars, right? And obviously this is, this is a part of that. But like, what's the appetite for this all, all of this? Yeah, like it's just follows the same patterns that any tech adoption cycle follows, right? There's obviously early adopters. And there's some sort of mainstream, then there's your latecomers, right? And, you know, sub-segments. It's like, you know, reading like a, like, a, Clay Christensen book in business school or something, right? It basically is what it is. It's the, it follows the pattern.
经销商，销售汽车，对吗？显然，这是其中的一部分。但是，这一切的市场需求如何，大家对这个有多感兴趣呢？就像任何科技采纳周期一样，它只是遵循相同的模式，对吧？显然有早期采纳者，然后是某种主流市场，再之后还有后来者，你知道的，细分市场。就像在商学院里读克莱克里斯滕森（Clay Christensen）的书一样，它基本上就是这样。它遵循着这个模式。
So I think right now we're in the point where early adopters are already in the game and they're doing the work that it takes to get, get, get ready. I think we're on the verge of going mainstream with things like customer data platforms.
所以我认为现在我们正处在早期采用者已经进入游戏并且正在进行准备工作的阶段。我认为我们正处在像客户数据平台这样的东西即将成为主流的边缘。
What I'll tell you is auto is a tough industry for data. It more resembles like a commodities, like a, like a, like a, I don't think about, like a run to your state, right? Where they have oil and they pull the oil out of the ground and, you know, take control of the oil. You have to pay for the oil. So data's like that. There's like all these countries that control a lot of data and sometimes even like the country doesn't control their own data. They have to pay a third party for access to their own data. So there's dealers who've like given their data to other places and can't even get it. And then you've got situations where, you know, there's a big pipe coming out with, with oil, right? With dealer data. Sorry for the metaphor here, but there's a big pipe coming out. And I want to go now and hook up, hook up another pipe so we can join the data. I love the metaphors, the metaphor. Right. And I'm like, Oh, wow, they, they build all their pipes like this and the other our pipes are smaller. Wait. So I have to now build a connector to like size my type to their pipe. And then, Oh, wait, guess what? That other country that has some, that other, you know, area that has some of your data, your data is now split into three different places. Guess what? They use railroad traps and take the trains. So I have to take it from the railroad tracks into one thing. By the time you build enough, you know, architecture to pull all that, all that data into one pool, right? You then realize, Oh, country A labeled the same data differently than country B and you're got a mess in the pool. You've got a clean. So that that's a lot of the challenge for the mainstream. Give some good metaphors, man. Thank you. And I never know if there's similar metaphors. I, I, I have no idea. No idea that part of my turf. My wife was an English reader. So I should, Oh, my goodness.
我告诉你的是，汽车行业的数据是非常艰难的。它更像是商品，就像，就像，就像…不管怎么想，就像一个能源供应国家。他们拥有石油，并从地下开采出石油，然后掌控石油。你必须付费购买石油。所以数据也是如此。有很多国家控制着大量的数据，有时甚至一个国家都不能控制自己的数据，他们必须向第三方支付费用才能访问自己的数据。所以有一些交易商将他们的数据交给其他地方，甚至无法获得数据。然后还有一些情况，有一个巨大的管道供石油流出，对吗？我的意思是交易商的数据。对不起，这个隐喻可能有些拗口，但有一个巨大的管道供数据流出。我想去连接另一个管道，这样我们就能共享数据了。我喜欢这个比喻。然后我意识到，哇，他们建造所有的管道都是这样，而我们的管道都比较小。等等，我现在必须建造一个连接器，将我的管道与他们的管道匹配。然后，哦，等等，你猜怎么着？那个拥有你的一部分数据的另一个国家，你的数据现在被分散在了三个不同的地方。你猜怎么着？他们还使用铁路运输，并会中途更换火车。所以，我必须将数据从铁路上取下来，放到一个地方。当你建造足够的架构将所有数据汇集到一个池子里时，你会意识到，啊，A国家和B国家对同样的数据命名方式不同，你的池子里有一团糟。这就是主流面临的很大挑战之一。说了些好比喻，伙计。谢谢。我从来不知道有没有类似的比喻。我没有任何头绪。这是我妻子的地盘，她是英语学者。所以我应该…天啊。
Point is I, I don't know how auto is going to how fast the transition can go because the early adopters are willing to do. Do things to get the data organized. The mainstream is going to have to either work with like a full path who has done a lot of work to build all those weird connectors to the pipes or wait until vendors. And others make more standardized APIs. And I think that's, that's sort of the dilemma here is, you know, how do we do things that in a sense might be bad for a specific vendor, but better for the industry. And, you know, one of our, one of our mentors once said like a lot of this, like when an industry has to evolve, like it's sometimes fair. It's good for the species, but it's bad for the individual animals sometimes. So, you know, if a vendor's making all this money off of forwarding the data and we're saying, Hey, you got to open the data up for, for the dealers to be successful, your own customers. But they say, wait, but everyone pays me for that now. Well, sure, but we're going to make you start naming name soon. No, we're not going to get it. I know.
重点是，我不知道自动化的速度能有多快，因为早期采用者愿意进行一些工作以整理数据。主流业务必须要么与那些已经为管道连接器建立了大量工作的零售商合作，要么等待厂商和其他人提供更标准的API。我认为这种困境的核心在于，我们如何做一些事情，这些事情在某种程度上可能对某个特定的供应商不利，但对整个行业更有利。我们的一位导师曾经说过，当一个行业必须发展时，有时对于个体动物来说是不公平的，但对于物种来说是有益的。
因此，如果供应商只是通过转发数据来赚钱，而我们却要求他们为零售商的成功开放数据，即他们自己的客户，他们可能会说，但是现在大家都为此付费呢？嗯，没错，但我们很快会要求你公开具体的资源。不，我们不会接受。我知道。
Well, well, what are, what are like, give me some examples of like the actual business functions that you think are going to are going to be impact or actually are like, are being impacted by the data. Are you like, are you being impacted right now? You know, buy, buy your platform, buy this data. Sure, like, I'll just give you a simple example. Like, imagine, like, let's take marketing. Okay. Outreach. A lot of what a dealer needs to do is to reach out, reach out to people and make them aware of what, what opportunities there. One of the most common questions I get in my DMS is like, I just got one right before this call and I get it all the time. Right. People ask me like, I'm looking for more ways to, you know, source clients, obviously, and, you know, to get more outreach and blah, blah, blah. So just putting that out there. It's one of the top questions. Yeah, for sure. It's really important. It can make or break a month. Right.
嗯，嗯，你认为哪些实际的业务功能会受到数据的影响或已经受到了影响，能给我一些例子吗？你是否正在受到影响？我指的是通过你们的平台和数据，你正在受到影响吗？当然，我来给你举一个简单的例子。比如，想象一下市场营销。好的，展示。经销商需要做的很多工作就是联系人并让他们意识到有什么机会。我在我的数据管理系统(DMS)里经常收到一个最常见的问题，就在这个通话之前，我也收到了。人们问我，像我正在寻找更多的方式来吸引客户，显然，想要更多的推广等等。我就是想说明一下，这是一个最常见的问题之一。是的，当然。这非常重要，它可以决定一个月的成败。
So, well, like most dealers think of marketing mainly to their, when they say you like database marketing, they're usually really looking at their DMS data. Right. DMS for those that don't know is like, I guess it's like a dealer ERP. It's a dealer ERP. Yeah. It's like the core accounting, you know, source of truth for what happens in the dealership. What cars are bought? What, what cars are serviced? So dealers look at those and say, okay, like, let's do like a loyalty program and like look for people who fit certain criteria and they do an outreach.
所以，嗯，就像大多数经销商主要将市场营销对象定位于自己，当他们提到数据库营销时，通常是在查看他们的经销管理系统（DMS）数据。没错，DMS对于那些不了解的人来说，就像是一种经销商企业资源计划（ERP）系统。它是经销店内发生的事情的核心会计、真实来源。哪些车辆被购买了？哪些车辆被维修了？所以经销商会查看这些数据，然后说，好的，让我们做一个忠诚计划，寻找符合某些标准的人，并进行外联活动。
But I'll ask a dealer, well, what about someone who was a lead seven years ago and you haven't heard from since. Is that person in the market? They're like, I asked no idea. I'm like, well, you have data on that person. That person might have been on your website today. Do you know? Like, I don't know. Like, well, we can find out. Let's use this automation program that we're hooking to see around in the stand and suddenly starts to send emails, relevant emails, because it goes back seven years and says, here's the thing that this person was most passionate about or most interested in it. It sends a nice, gentle talk. It's had a huge amount of people from seven years ago. Do some action.
但是我会问一个车商，那么，对于一个七年前担任过销售主管的人，你们从那时起就没有再听到他的消息，他还在市场上吗？他们说，完全毫无头绪。我说，你们有这个人的数据。这个人可能今天还在你们的网站上。你们知道吗？他们说，不知道。我说，好吧，我们可以找出来。让我们使用这个自动化程序，它会搜索七年前的数据，并发送相关的电子邮件，因为它会找到这个人最热衷或最感兴趣的事情。它会发送一封友好的讲话，然后会有大量来自七年前的人采取行动。
And as they do some action, they're starting to create connections in the data. And now I know, oh, wow. That person's kind of identified. They've raised a hand and I can see they were on the website looking at an F-150 three weeks ago. And suddenly we can say, hey, there's a person from seven years ago. You should probably be a touch that person because they're looking at F-150. Would we want to send them an email? They're like, yes, I would. And I can't send. I can't go through 10,000 records manually now. Oh, no problem. It'll automatically write the email automatically send it and then handle the back and forth, maybe, and drop a lead in your CRM. So that's the type of functions that are already being handled pretty easily.
当他们进行某些操作时，他们开始在数据中建立连接。现在我知道了，哇，那个人有点被识别出来了。他们举起了手，我可以看到他们三个星期前在网站上看了一辆F-150。突然间我们可以说，嘿，这是一个七年前的人。你可能应该联系一下他，因为他正在看F-150。我们想给他发封邮件吗？他们会说，是的，我想要。但是我不能手动地对10,000条记录逐个处理。没问题。它会自动地写邮件、自动发送，并处理来回沟通，或许在你的客户关系管理系统中添加一个销售线索。所以这些功能已经相当容易地被处理了。
And what's coming up next? Like, what do you think is the next couple steps there? We think that the one of the biggest challenge for dealerships is like making the old school new school. So you've got, you've got great people sitting in dealerships, but in particular after COVID, like selling cars is not the easiest thing in the world. And people don't necessarily remember some of the ways that you sell cars from before. And so we actually think that finding ways to really integrate with their workflow, maybe they're in CRM, maybe they're wherever they are. And, you know, every case a little different. And giving them insight on a shopper or someone walking in the door saying, hey, here's some things you could say. Here's some opportunities you could offer really arming them with those things could be very valuable to dealerships and then making sure that whatever happens there in that moment in that interaction is translated through into the market and is, you know, brought back when next time the person comes in.
接下来是什么？你认为接下来是哪几个步骤？我们认为经销商面临的最大挑战之一就是如何将传统方式与现代方式结合起来。在经销商中有很多优秀人才，但尤其是在COVID之后，汽车销售并不是一件容易的事情。人们不一定记得以前的汽车销售方式。因此，我们认为真正融入他们的工作流程，可能是他们的CRM系统，可能是其他地方，寻找方法并为他们提供洞察一位购物者或前来咨询的人，告诉他们一些可以说的话，一些可以提供的机会，装备他们这些东西对经销商非常有价值，并确保在那一刻的互动中所发生的一切都能传达到市场，并在下次来店时保留下来。
So we think like the most important thing right now and what's going to be all the work of the next year or so is going to be just connecting all those dots and making sure that the customer journey is as connected as seamless as possible. And how does that compare to what it is like today? It's pretty shattered, pretty siloed. Also, like, it's hard to ask someone who's not used to like opening a screen and looking at it to do it. It's got to be in their workflow. But right now, most CRMs that we look at, we do a lot of CRM analysis, we do a lot of DMS analysis, but most CRMs we look at could have 30 to 50% of their shoppers as duplicates or near duplicates.
所以我们认为，目前最重要的事情以及未来一年左右的所有工作，都将是连接所有这些点，并确保客户旅程尽可能地无缝连接。那么与现在相比，它是什么样子呢？它相当分散，相当独立。此外，要求一个不习惯开启屏幕并查看内容的人这样做是困难的。它必须融入他们的工作流程中。但是目前我们看到的大多数CRM系统，在我们进行CRM分析和DMS分析时，他们的购物者中可能有30%到50%是重复或接近重复的。
Right. So you, you have like four instances of Sarah. And no one's doing the work to merge those together in an intelligent way or trying to figure out is it all the same person. And then you look at like a shopping experience of Sarah and she's getting just the generic email about trucks. But she has no interest in trucks. She's looked, she's told you what she's interested in by her behavior. So she's still getting the truck email dealership down the road that has a CDP and hasn't hooked into their outreach arms, their engagement arms, right? They're not offering her a truck. They're offering her what she wants, right?
没错。所以你有四个Sarah的实例。但没有人在以智能的方式合并它们，或者试图弄清楚它们是否是同一个人。然后你看看Sarah在购物体验中，她只收到有关卡车的普通电子邮件。但她对卡车不感兴趣。通过她的行为，她已经告诉过你她感兴趣的内容。所以她仍然收到来自附近车行的卡车电子邮件，这些车行没有与他们的外联部门和参与部门进行连接。他们没有向她提供卡车，而是提供她所想要的东西，对吗？
Another example where I think it's going to go is like collaboration. So if I'm a shopper right now and I go to a dealership and I click on an ad for like a lease offer and I come to the website, I browse, I convert, whatever.
我认为还有一个例子是合作。 所以，如果我现在是一个购物者，然后我去经销商那里，点击了一个租赁优惠的广告，然后进入网站，浏览，最终购买。
Full path CDP knows what car that person is interested in. They know what lease offer ties them. Now, when they're on the website, even before they convert, they clicked in on that ad. I already had most of that information.
全路径CDP知道这个人对哪种汽车感兴趣。他们知道哪种租赁优惠适合他们。现在，当他们在网站上时，即使在他们转化之前，他们也点击了那个广告。我已经拥有大部分这些信息了。
When that chat pops up to say, hello, how can I help you today? Right? If it's a full path chat, it's going to relate to what we know about the person from that initial interaction. We don't give them too much data to chat, but we give it enough.
当那个聊天弹窗出现并问道“你好，今天我能帮助你什么吗？”对吧？如果是一次全路径的聊天，它会与我们从最初的互动中了解到的个人信息有关。我们不会给他们太多聊天所需的数据，但我们会提供足够的信息。
I have that data even if they're not using me. Let's say they're using a competitor chat. I would love to hand that data off to the competitor and say, hey, mention this lease special when you say hello. It doesn't have to be full path. Right. So that ability to work like interoperability between vendors, I think is going to be really important for dealers. It just doesn't mean it's not.
即使他们不使用我，我仍然拥有那些数据。假设他们正在使用竞争对手的聊天服务。我很乐意将那些数据交给竞争对手，并说：“嘿，当你打招呼时提及这个租赁特价”。这并不一定要完全转变。对于经销商来说，能够在供应商之间实现互操作性的能力将非常重要。这并不意味着它不重要。
I was going to say, does any of that exist today? Not really. Like, even when people, people who are saying it exists, it doesn't really exist. Any integrations are not really integrations. There's, yes, sure. I can go get the info, but, but, or I can give the info to someone, but do they really know what to do with it? And that's the real thing.
我本来要说的是，这些东西今天还存在吗？实际上并没有。即使那些声称它们存在的人，其实它们并不存在。任何整合都不是真正的整合。当然，是的，我可以获取信息，或者把信息提供给别人，但他们真的知道如何处理吗？这才是真正的问题。
Like, one of the principles we came up with as we were building our company was we only, we only talk about data or claim and integration when we can do something with the data. If there's a data source that I can stick in my database and a table, but I can't do anything with it for the dealer, who cares, right?
就像在我们建立公司时提出的原则之一一样，我们在讨论数据或主张数据整合时，只会谈论可对数据进行实际操作的情况。如果有数据源可以放入我的数据库和表中，但对经销商来说却没有什么作用，那谁会在意呢？
Like, but a lot of people think that having data matters. It doesn't matter if you have data matters. If you have data and you could do something valuable for it with the customer.
喜欢，但很多人认为拥有数据很重要。事实上，拥有数据并不重要。只有当你能通过客户的数据为其提供有价值的服务时，拥有数据才有意义。
By the way, in a lot of ways, this is just going back to like pre pre-internet times in some ways. If you think about it, like dealers, dealers were always community hubs. Dealers were always rooted in local, you know, local environment. The dealers there knew everybody, right? And most of the time families bought from the same dealer year after year of year. They were relationships.
顺便说一句，在很多方面上，这就像是回到了互联网之前的时代。想想看，像汽车经销商这样的企业，一直以来都是社区的中心。他们总是紧密联系着当地环境。那些经销商认识每个人，对吧？而且大多数家庭每年都会从同一个经销商购买汽车。这是一种关系的建立。
You kind of knew their kids, you knew what was happening, you knew when the birthdays were coming. Like, there's just a lot of awareness. And then as the internet emerged and in particular, like endemic sites, like, you know, cars not common and car viewers, et cetera, came out.
你有点了解他们的孩子，知道发生了什么事，也知道生日什么时候到来。就像，这种意识很多。然后随着互联网的崛起，特别是像汽车网站之类的流行网站，你知道的车型和车辆查看者之类的工具出现了。
Like dealers just became like phone numbers and prices and the relationship room. And I think a lot of what dealers want to do now is take back those relationships and they have now the tools to do it.
就像经销商们只是变成了电话号码、价格和关系的角色一样。而我认为，现在经销商们想要做的主要是重新建立这些关系，并且他们现在有了实现这一目标的工具。
So they can compete now against the big players because of these CDPs and power that is going to emerge from AI capabilities.
现在，通过这些CDP和即将涌现的人工智能能力，他们能够与大型参与者展开竞争。
More than that, I think it actually gives dealers a fighting chance to compete against emergent competitors who would love to dominate the whole industry. The Amazon's the Walmart's one day, right? That's where things are going to go down the road, where you're going to see like, finally, these big players find a way to get in on a car business or meaningful way.
而且，我认为这实际上使得经销商有了与希望主导整个行业的新兴竞争对手竞争的机会。亚马逊和沃尔玛总有一天也会进军汽车行业，对吗？这就是未来的发展方向，你会看到这些大公司最终找到一种方式来参与汽车业务或发挥重要作用。
I think dealers will have to, if they embrace these tools, CDPs, AI, they actually will be able to compete in the future, even though they are a fragmented ecosystem. I want not.
我认为经销商们如果能采用这些工具，如CDP和人工智能，他们实际上能够在未来竞争，尽管他们是一个碎片化的生态系统。我不希望他们放弃这些工具。
Let's take your vision for a second. When you're successful, let's say five years out, even longer if you need. But what does that, what does that car buying experience look like and feel like? How has it changed?
让我们先设想一下你的理想景象。假设五年后，或者更远一点，如果需要的话。那么那个购车体验会是什么样子？它会有何变化？
So the, the car buying experience is going to change it because certain things won't change. People are going to want to actually feel a car out. They're going to want to drive a car. They're going to spell the meaning. There's an element of a car that is very tapped out.
所以，购车经验将会发生变化，因为某些事情不会改变。人们会希望实际感受一辆车。他们会想要亲自驾驶一辆车。他们会领悟汽车的含义。汽车具有一种非常枯竭的元素。
And I think people are going to want to get into their cars. And I think dealerships will be around. I do think a lot more of the research and stuff is going to be done through tools like AI.
我的意思是，我认为人们依然会想要使用自己的汽车，并且汽车经销商依然会存在。但我认为，更多的研究和其他事务将通过人工智能等工具来进行。
And I think for us, what our vision is, is that the dealerships, like we think about the, getting the dealerships ready, is that their data is organized. They have tools in the AI to leverage it for whatever the consumers want and however the consumers want to engage. Right.
我认为，对于我们来说，我们的愿景是将经销商机构准备就绪，确保他们的数据被有序整理。他们拥有人工智能工具，可以根据消费者的需求和消费者的参与方式进行灵活运用。
There's a dealer, David Long, who said the golden rule is dealership is like communicate the way the dealer wants to communicate. Right. That's the, you know, it's a golden rule plus plus. I think it was what it was. Right. You know, a dealer, dealer coined golden rule, which is if the shopper wants to talk by phone, talk by phone, if they want to text, let them text.
有位经销商叫 David Long，他说经销商行业的黄金法则就是按照经销商自己喜欢的方式与顾客沟通。没错，这就是一个超级版黄金法则。我觉得就是这个意思。你知道，经销商发明了这个黄金法则，就是顾客想打电话就打电话，想发短信就发短信。
They want to email, email if they want to do it through their AI agent. Let them do it through their AI agent, whatever it might be. Do it.
如果他们想要通过他们的AI代理发送电子邮件，就让他们发送电子邮件。不管是什么样的AI代理，都让他们通过它来做。就这样。
And so I think for us, like we're working now on what we think is the most important transformation for the auto industry, which is to take the CRM and the DMS, which are right now, the two legs of really the dealer text app.
所以我认为对于我们来说，我们正在努力进行我们认为是汽车行业最重要的转型，这就是将CRM和DMS结合起来，这两者目前都是经销商文本应用的核心。
Yeah. Website, but fine. That's, you know, their websites make common go may change, et cetera.
是的。网站，但还好。你看，他们的网站常常会发生变化之类的。
The CRM and the DMS, we believe the CDP will be like the next piece because they'll bring all the data into one, one hub and be able to push to the CRM, push to the DMS.
我们认为，客户关系管理系统（CRM）和经销商管理系统（DMS）中的下一个环节将是客户数据平台（CDP），因为它们将把所有的数据整合到一个中心，并能够将其推送到CRM和DMS。
Take from the DMS and then around that an AI operating layer. Right.
从DMS中获取数据，然后围绕它建立一个AI操作层。好的。
So what does that mean? Like the AI will know, okay, Sally's now looking to communicate. We're not paying attention to her. This was not answered. Well, someone's got a right now close that loop and it's urgent and pop that up and let everyone know and maybe even suggest talking voice and even write the email and then just have a person do the reach out.
那是什么意思呢？就像人工智能知道，好的，Sally现在想要交流。我们没有注意到她。这个问题没有得到回答。有人现在就要解决这个问题，并且很紧急，要弹出通知让大家知道，甚至可能建议用语音交谈，甚至帮助写邮件，然后让人员进行联络。
Or they're going to say, Hey, we need to generate the content right now to get into Sally's inbox or into her frankly paper mailbox, even in five years, whatever it might be. And it could generate a lot of that content and get that done. So you're going to have these like, like cyborg dealerships that are a mix of human and tech working to provide just the best possible experience in the way that the consumer wants.
或者他们会说，嘿，我们现在需要生成内容，进入Sally的收件箱，甚至五年后进入她的纸质邮箱，无论是什么方式。这样可以生成大量的内容并完成这个任务。因此，你将会看到这种混合人类和技术的类似“半机械”的汽车经销商，致力于按照消费者期望提供最好的体验。
You know, it makes me wonder, like, who is the GM of the future of the dealership? So like, how does a small dealership survive when, you know, the more sophisticated groups that have so much more resources have these tools, right? If I can send Sally a targeted message on a 2017 Prius that she wants and you're sending her flooding her with messages on 2022 GMC sierras, which she could care less about. I mean, how are you going to possibly compete like that in the future?
你知道嘛，这让我想一想，像是，未来车行的总经理会是谁呢？就是说，当更加精细的集团拥有更多资源和工具的时候，一个小型车行要怎样生存下去呢？如果我能向莎莉发送一条关于她想要的2017款普锐斯的定向信息，而你却在给她发送大量有关2022款GMC希拉斯的信息，她对此毫不关心。我的意思是，你将如何在未来与此竞争呢？
No, but this is the opposite. This is power to the people here. Like, you're basically democratizing things that were very, very difficult to do.
不，但这正好相反。这是人民的力量。就像，你基本上是在实现原本非常困难的事情的民主化。
So as an example, like you remember what it was like to create a website and I don't know, whatever 2005. I don't forget. It's a mess. It's some point was a mess. You had to know code or you had to hire a wizard, a K a developer to go write the website for you.
所以举个例子，就像你还记得2005年创建一个网站是什么样子，而我不知道，不管怎么说，那时候是一团糟。有些地方很混乱。你要懂得编码，或者要雇佣一个巫师，一个开发者来为你编写网站。
If I want to set up a website today for something basic, I use Shop Fives, Wix. It's, it's, it's, it's everyone's ability to go and do something.
如果我想要今天建立一个基础的网站，我会使用Shopify或Wix。这些平台的使用非常简单，任何人都可以去做。
And by the way, that's like, I don't know if you remember based on your age, but I see we do the web 2.0 phenomenon, right? A lot of that was predicated on this idea that the cost of technologies is just just dropped. You can launch a web app, not for a million dollars to server cost, but for $1,000 a server cost. So it really, it really opens things up.
顺便说一下，你可能记不记得了，根据你的年龄。我看到我们正在经历Web 2.0现象，对吧？其中很多都是基于这样的理念，即技术成本大幅下降。你可以以1000美元的服务器成本，而不是以百万美元的成本来推出一个网络应用程序。因此，它真正地打开了一些新的可能性。
And I think the same thing is happening in automotive where dealers who are interested in using these types of tools and who are curious, you ask about the jam of the future, they've got to be curious over the technology, have good instincts. Of course, they need the same skills they have now. They just have to be open to technology. And I think they already are many of them. Those folks are going to have such power as individual dealers.
我认为在汽车行业也正在发生同样的情况，对于那些对使用这些工具感兴趣并且好奇未来趋势的经销商，他们必须对技术保持好奇心并具备良好的直觉。当然，他们需要和现在一样的技能，只是需要对技术持开放态度。我认为很多经销商已经具备了这种态度。这些人将作为个体经销商拥有巨大的力量。
Again, not to use like too much military analogies, but like they talked today about how a little squad commander with a little computer has the ability to do things that you would have required like a, you know, a Brigadier General to do beforehand. It's just, it's really putting power out to the edge and it's giving every dealer the ability to like aggregate and clean their data up.
再说一遍，不要过多地使用军事比喻，但是他们今天谈论了一个小小的排长，有一台小电脑的能力能够做以前需要一个准将才能做到的事情。这真的把权力下放，给予每一个经销商能够整合和清理他们的数据的能力。
So use AI to like get functions done to build sophisticated marketing that they never put a dream. And that's not happening. I think that's what's going to happen.
所以使用人工智能可以完成工作，构建他们从未梦想过的精细化营销。但这并没有实现。我认为这就是即将发生的事情。
So I think you bring up a good point, which I wasn't, I didn't mention, right, which is you don't need to be sophisticated to use these tools. You meant use the word democratizing, right? In my head, I was initially assuming probably need to be pretty sophisticated, but based on what you're saying is no, right?
所以我认为你提出了一个很好的观点，我没有提到这一点，对吧？你的意思是使用这些工具不需要很复杂的经验，对吗？在我心里起初我可能认为需要相当高级的技巧，但根据你的说法是不需要的，对吗？
Like that's essentially the hard work you're doing to end user, in this case, the dealer, the business, whoever gets just to reap the benefits without, you know, needing to go through all that sophistication.
就像这是为最终用户（在这种情况下是经销商、企业或任何人）所做的艰苦工作，他们只需享受好处，而无需经历所有这些复杂性。
Yeah, I think that's the way that we see it. Like, it's the same thing by the way with a, with a vehicle, right? Like if I have a, if I have a car in 2000 and 24. Think about how much easier it is to sort of handle that car than it was in 1974 or I don't know easier. I mean, I don't. Cars today are computers. The steering works for you. You, you don't need to get underneath it now and like monkey around with stuff. If you do, you'll break it because it's a computer. So essentially, like, it actually becomes, you, you need less subject matter.
是的，我认为这就是我们的看法。比如说，同样的道理也适用于车辆对吧？如果我拥有一辆2000年以后的车，想想看它相比于1974年时要容易操控多少。或者说不仅仅是容易，我是说我不需要在车底下忙来忙去了。现在的汽车都是电脑控制的，方向盘可以自动运作，你不再需要亲自调整了。如果你这样做了，很可能会损坏它，因为它是电脑控制的。所以基本上，实际上，你需要的主体知识会更少些。
Expertise and you could just use it. Another example I read, Oh man, it was, it wasn't a wire or ankle a long time ago. Fast company, maybe not even years ago. I think it was Nicholas Nutterponti who used to work at MITs or one of the Nutter Pontis. Anyways, he pointed out like music. Like when, when the digitization of music came out, every producer was using it and like using synth sounds and they, like you had to know that it was digital. Because it was like the cool new thing. Now I always external. It's all digital, but you don't have to feel the digital. It's just there. The users just use it. And it's not a thing anymore. So similarly using like technology, like what we develop or GPTs in the future in certain ways, that will just be there at their fingertips. It won't be a big deal anymore. It won't be exciting. It'll just be the way they do business, but they will have a measurably more power and capabilities than what they currently have.
专业知识，你只需要使用它。我读到的另一个例子是，哦，哥们，它不再是很久以前的电线或脚踝了。快速公司，也许甚至不是几年前的事了。我想应该是尼古拉斯·纳特庞蒂曾经在麻省理工学院或纳特庞蒂家族的一个人。总之，他指出了音乐的情况。就像音乐数字化出现时，每个制作人都在使用它，使用合成音频，你必须因为它是数字化的，所以你可以知道它的来源。因为它是那时的潮流。现在一切都外在。一切都是数字化的，但你不必感受到它的存在。它就在那里。用户只需使用它。而且这已经不是什么了不起的事了。所以类似地，使用像我们将来开发的技术或GPT（生成式预训练模型）等，这些将只会在用户的指尖之间存在。这将不再是一件大事。也不会是令人激动的事。它将仅仅成为他们开展业务的方式，但他们将拥有比目前更强大和更多功能。
Well said. Dude, I'm scared for this next conversation we're going to have at some point later this year because every time, I mean, it feels like every conversation we have, there's so much progress done so quickly. Yeah, who knows who knows. We'll all do our best to make it a make it work for everyone. You know, that's. I love it. Yeah. Well, super insightful and well said. So our own. Thanks so much for coming on. It's been great. It's been a great meeting. My pleasure. Anytime.
说得好。伙计，我对我们即将在今年晚些时候进行的这次谈话感到害怕，因为每一次，我是说，每一次我们进行的谈话，进展都那么快而且那么多。是的，谁知道呢，谁知道呢。我们都会尽力让它对每个人都起作用。你知道，我喜欢这个。是的。嗯，见解非常独到，说得好。所以，我们自己。非常感谢你的加入。真是太棒了。这是一个很棒的会议。不客气。随时都可以。