Connections 2025: Agentic Marketing Showdown

All right. Aloha, Agent Blazers. Every year we go, the correct response is aloha, Aloha, agent Marty. Okay. Aloha, agent Blazers. Aloha, agent Marty. One more time. Aloha, agent Blazers. Thank you for that warm Agentic welcome. I am Marty Kine and welcome to the first ever Agentic marketing showdown. As you probably already noticed, this is a little bit unusual. We haven't done anything quite like this before. Why? Because it's agentic? No, everything is agentic because it's about marketing. No, everything's about marketing because it's a showdown. What do I mean by that? Well, a couple months ago, we sent out a request to our valued partners. This would be systems integrators, consultants agencies, people who are adept at the Salesforce platform and also very excited about agents which. Who isn't? So that was pretty much all of them, but. And we asked them to build a custom agent for us to be entered into a contest. We had many, many dozens of submissions, some of which were excellent, some of which were almost excellent. And we've narrowed it down to three finalists. And these three finalists or representatives from those teams are here today. And you will see their agents demoed. We have a panel of judges, an esteemed panel, including one of whom is actually giving them the keynote tomorrow. Is that right, Steve? That's right. So we're very honored to have such an esteemed panel. They will be judging the submissions and you, the audience, will also get to vote. It's extremely exciting. What is the reward? The reward is a job well done. A feeling of satisfaction knowing that you've designed an agent that excites and delivers business value. And also $10,000. Not bad, huh? All right. I like the energy. I think I covered everything I'm supposed to cover so far. So these agents, these are custom agents. They were enterprise worthy, trusted, built on top of data that's sitting in the data cloud, which is, of course our, you know, leading data platform. And if any of you are interested in learning more about Agent Force, I can recommend a new book. There's an excellent new book, just came out two days ago. It's called Agentforce. Can anyone guess what it's about? I tried to make it very. It was written by me. So, you know, thank you. You know, you get a quality product in here, so I can highly recommend. Is available right now on Amazon. And it's okay to have your phones out and just make a place a few orders. Father's Day is coming. A lot of dads out there want to Know more about Asian force. All right, so let us now invite our esteemed panelists in this order. Robert Lopez, I like your. What is this? What is this?

Oh, my Lord. Leave the stage.

I do know. It's a rhetorical question. How did you. What did you get that for?

What did I get it for? You don't want to know. No, I'm joking.

Being a trailblazer.

It's being a trailblazer, supporting the community, the Salesforce community, and really giving back where I could.

Thank you so much for that. Next judge, Steve Hammond. Welcome, steve and bobby. Dania, Very good looking panel, wouldn't you agree? Very nice. Yes. More applause please. Thank you. All right, so let me just go in this order. Just take a few moments to introduce yourself and tell our audience who you are.

They always sit me next to people with great heads of hair. I don't know.

Bobby has.

Yeah.

Yes, Bobby. See, that's fantastic. So I'm Robert Lopez, Chief Marketing Innovation Officer at Norse Collective. We're a dynamic hospitality group in Sydney, Australia and we were the first APAC customer to go live with a customer facing agent earlier this year in January, which was exciting for us.

Thank you, Steve.

I'm the general manager for Marketing Cloud and I've been with Salesforce for two and a half years. I've worked in the industry for about 28 years and I couldn't be more excited about this event. We've been planning for Connections for about two months and. And this popped up on my radar about two weeks ago, three weeks ago. And this is the only session that I was really excited about even beyond my own keynote.

I share that sentiment, Steve. It's not to say anything about the other sessions, but there's something special about this one. I think I agree. Bobby. Who are you?

My name is Bobby Jainya, Chief Marketing Officer for Marketing Cloud. I will say that we pulled Steve into planning two months ago for the team out here. It's been working for at least six months, planning for connections. Been at Salesforce for 11 years this month and I've been on Marketing cloud the entire time. So I've really seen the transformation the whole way.

Wow, excellent. So in terms of the voting, I'll explain how that works. We have our three team now. I want to be clear that the teams came up with these agents, the custom agents and we have a representative from each team just for logistical reasons. We'll have one representative come up and introduce the agent and we'll have a demo, it's a recorded demo and then the Judges will comment on their feelings about their impressions. Their feelings. The judges will be in there in terms of their scoring, will be looking at three things. Number one is business value. So how much actual tangible business value do we think this agent would bring to marketing function or whatever? The second one is creativity. So is there a little sparkle in there? The third is the demo delivery itself. So do we like the demo? And again, the marketers. Then at the end, the three submissions will have comments. The judges will tally up their scores. They will confer, kind of like at the Olympics when you have the skating judges all together, and they will maybe not argue. They will agree. And then you guys will get to vote as well using your phones. And I'll explain how that works later. Okay, so I think we are now going to move on to our pitches. What we've all been waiting for. These are our three representatives and we are going to start with demo number one, our friend, Ateco, Elliot Harper. Elliot, come on up. Come on up, Elliot. You'll notice that Elliot also had a very long flight to get here and he is also.

Hello, my name is Elliot Harper and he's also an author at Presenting Team Ateco. Today I'm excited to introduce a groundbreaking, innovative approach to agentic AI, one that redefines what's possible with email. Now, AI has already revolutionized retail, reshaping everything from personalized shopping experiences to predictive supply chain management. Today, 84% of retailers are using AI in some form. And AI in retail is becoming increasingly agentic, capable of acting autonomously without human intervention. In fact, 43% of retailers say that they're already piloting agentic AI. But while the benefits are compelling, the challenge is clear. How do you make agentic AI feel seamless to the customer? For example, a simple task like updating an order. Typically, a customer might need to log into a retailer's website, assuming that they've actually created an account. And if they checked it as a guest, good luck. I mean, this is often a high fiction. Experience. Experience. But regardless of who they order from, there's one thing every customer receives. An order confirmation email. So here's the big idea. What if a customer could make changes to their order directly from that email? No logins, just real time interactive support right inside their inbox. Well, with marketing, cloud engagement and the Agent API, that's exactly what's now possible. Creating rich two way conversations that live inside the email itself. Let me show you how this works. Meet Karen, a software designer who spends her weekdays designing user experiences and her Weekends chasing real ones. Karen is an outdoor enthusiast through and through and a loyal customer of Northern Trail Outfitters. Karen recently placed an order and receives a confirmation email. When she opens that email, the order items are retrieved in real time. So if anything about her order has changed, then it's immediately updated at open time. And at the bottom of the email you'll notice there's a built in service agent. Now let's say that Karen wants to update her shipping address from 742 Meadowbrook Lane. So she asks please update my street address to 112 Summer street when she clicks, sends agent, force, launches a custom flow to go and update her order and confirms the change right in the email. You'll also see that her address has changed in the email in real time. Also, each message is timestamped showing when it was sent and responded to. Now let's try another change. Karen ordered a teal and turquoise Isabella backpack, but decided she'd prefer it in black after browsing other options on the NTO website. So Karen asks, could I change the Isabella backpack to the color black?

Black?

Again, the agent processes the prompt, identifies the correct action, launches a custom flow, confirms the change, and you will see that the item image and color have changed to a black backpack. Write an email in real time. Let me show you how this works. When the customer submits a prompt, a message is sent to the agent via the Agent API. The agent then identifies the topic that best matches the prompt and then based on that matched topic, the agent triggers one or more predefined actions like launching a salesforce flow. Then the flow goes and updates the customer order and returns a contextual response back to the email. Finally, the email is dynamically updated to reflect the latest changes. And there you have it. I've shown you how an autonomous agent can be embedded into an email, enabling frictionless agentic customer experiences without ever leaving the inbox. Thank you.

All right, Elliot Harper from Ateco. Excellent. Great way to start. So we'll turn to our panel just like a minute or so your general thoughts and we'll start with Bobby.

Yeah, I love the use case there because I think about how many times I've bought something online as a guest because I'm like, oh, it's one order. And then I realize I can't find the tracking information or anything else. And so the fact that there's that commonality that no matter whether you're a guest or logged in, you all receive that email. That was a great starting point there. I love too that it highlights how agenta capabilities can still come to products like MC Engagement. They're connected into the platform and they still work in that sense there. So it's also a great way to highlight that everything that our customers use today can be identified.

Yeah, it seemed very practical, Steve.

Yeah, I mean it's a use case, it's something we can all relate to. And it's also using a channel that's extremely common and it's everywhere, so it really makes it nice. Can I ask a question?

Of course.

Okay. You showed this incredible demo. Did you actually deliver it live? Did you have a functional version of this?

Yes, it actually works. Happy to show you. For those of you interested in how it works, come along to my session tomorrow. Shameless Plug. Look it up in Agenda Builder, but I'll be showing it live.

And the follow up question to that. Some email applications limit your interactivity within the email itself. How did you get around that? Or are there special conditions you have to work with?

Yeah, look Steve, a magician never reveals his secret, right? But I'll let you into mine if you promise not to tell anyone.

Okay, everybody close your ears.

So the answer is amp. AMP support. This isn't ampscript, but AMP for email specifically has been around in marketing cloud engagement for four years now. But to date no one's been able to have a two way conversation using this technology. So. And that's what we've actually been able to achieve and it works really well. But to answer your question, since in the past four years the number of email clients that support Amphor email has increased, but there is a fallback experience, two fallback experiences in fact. So one will fall back to using interactive email where when you complete and type the message in the prompt, then it will launch a web page, a cloud page, with the email and the agent already there. And then there's a secondary fallback experience for Outlook where the agent is replaced with a button and that again that launches the email in the webpage and you can interact with the agent there.

Brilliant. Okay, thank you. And just follow up my feedback on this. I really. You can sit down.

Sorry.

Thank you. I do. One of the other things I really liked about this was that in addition to having the innovation happening in email, I also liked that it used the current in market product. While we do have this new agentic capability as part of the next generation marking cloud, I love that you are able to use the APIs to connect an agent force to do as well. So congratulations and thank you, Robert.

Wow, wow, wow. Yeah, I think from Presentation perspective, excellent. The conviction in the presentation was astounding. Really, because you believe in it. You've lived and breathed it. You can tell. I, you know, I'll put my CMO of a hospitality group hat on and I can see so many use cases and following on from what Steve said, the fact that it's the now it's there, we could do it tomorrow really is fantastic. And God, the number of customers who at the bottom of an email can change a restaurant reservation or update their membership email address or anything, it's just, it's limitless. So this is just exceptional. It really is.

All right, great work, Elliot. Once again, the Teco team, some of whom are up there, some of them formerly Cloud Kettle. Right, Is that emerging? Okay. All right, our second contestant from Listengage we have Pratik Desai.

Prat.

Hello everyone. Pratik Desai, head of Applied AI at Listeningage. But today I'm going to serve as the fictional CEO of Ellie Optical. Now, I've been asked to tell you about my experience with AgentForce. Rather than tell you about it, let me show you. So we're going to meet Colton. He's a busy employee drowning in the day to day. He wears multiple hats and he is the linchpin. The thing is, sometimes, unfortunately, he's also the bottleneck between sales, service, loyalty and marketing. Now let's meet Leah, a busy consumer who desperately needs new glasses. Like us, she works from home, going from meeting to meeting. She doesn't have time for her errands. She's frustrated with these customer experiences. She's drowning in product pages, drowning in search results, tired of irrelevant offers. So Cole knows something has to give. As luck would have it, he's about to attend his first Agent Force workshop. Now, inspired by what he's learned, he's about to start something big. He's going to streamline his operations. He's going to drive business value. And he's going to do this by implementing Agent Force. So back at the office, Cole decides to use data libraries to give leoptical agent the ability to act like an employee for the last 20, 20 plus years. Now he's going to go ahead and create a custom toolkit. So he does this by giving Agent Force the ability to hit things like scheduling appointments, transferring to customer service reps and managing the entire loyalty management program with simple prompts acting as instructions. Agent Forest can now carry on conversations to do points, usage strategies, promotional activities and manage accounts. AgentForce is now ready for the first production rollout. And of course Leah, our busy consumer, starts to browse the website and she sees the little chat bubble. So she starts to interact with AgentForce asking about purchase options. And because Cole has used data libraries, data cloud and loyalty management tools, Leah's able to self service without any human interaction. She signs up for the loyalty program without any troubles. And as Leah, as AgentForce is receiving Leah's information, it's using loyalty tools to hit the right objects. All while Cole's enjoying his lunch break. The best part about how Cole rolled out AgentForce is that he put it behind marketing cloud personalization, which means that he's targeting the perfect prospective customers, the high value customers that leoptical has to gain. And he does this by using propensity scores from data cloud synced to marketing cloud personalizations, creating a campaign and then using AgentForce to target the customer. So Leah is delighted with the experience she's had so far. And thus she decides to purchase with AgentForce today. And AgentForce keenly recognizes that Leah's purchase today qualifies her for 10,000 loyalty points and offers a promotional coupon for future use. That's a happy customer right there. We don't need to see her to know she's happy. We have Salesforce, right? So in Salesforce we have a data integration between Data Cloud and AgentForce for a single unified view to see all the conversations she had with AgentForce and all the actions it took and why. That's the dashboard that Cole has access to now. Okay, but he doesn't stop there. He decides to take it one step further. He's going to streamline his marketing operations. What he's going to do is he's going to take Agentforce and connect it to the Marketing Cloud APIs. Then with simple prompts, he can talk to Agentforce who is now able to hit the APIs to do full scale rollouts of marketing cloud deployments, configuration and sends. And of course Leah, our happy customer, decides to download the mobile app at any moment when she has a question or wants to do another purchase. An agent is never more than a few fingertips away. And as Cole talked to his mobile developer, all he had to do was make sure they were hitting the Agent Force API. For Cole's final trick, he connected Slack to AgentForce. This allowed his cross collaboration team to collaborate in their collaboration platform. So for things like Leah's where's my status order? They could ask for a status update. They can fire an email off all through AgentForce with a job well done. Connecting Agent Force to sales service web loyalty Mobile and marketing. Cole enjoys a nice happy hour with his team. But don't take my word for any of this because it's live today. Scan this QR code and play with it yourself.

All right. Great job. Great job. Wow, that was fabulous and very good acting. Who was the star there? That gentleman. There he is. Oh, my God, he's here. Many of you want to meet him. Right there.

Well done, Cole.

Excellent work. Well, the competition level is very high today. I can see that. Let's start with you, Rob. What did you think?

Excellent work. Can I get your signature later? Fantastic. I do love humor, if you can't tell, and that was brilliant. Loved the multi cloud approach. Really, really good to see all those use cases. Love the personalization element. Big, big personalization fan. I think that was a really, really cool inclusion and I think very relatable. I think everyone in this room related to that situation on either side of the equation. So really, really good. Awesome work.

Well done. Steve.

Yeah, I don't know.

Cole, could you stand up for a second?

Yeah.

I don't believe that you eat potato chips. That was great. Thank you very much. Yeah, I loved the Brett of what you guys did. Really showing how you connected all these different systems. Being able to use the agentic web experience, including personalization in that as well. The purchasing through an agent. Can I ask you more about that one quick? Absolutely, yeah. I wasn't sure. Could you. Did the purchase actually happen through the agent as well?

Yes. So this is all built out. Truly. And we have the website live for a real fictional company. We have an OMS integration as an action instruction topic that it can then go into the OMS system and fire off the purchase.

Okay, and how were you managing, for example, the transaction of that? Was it through the oms?

Yes, that's correct.

Yep. Okay, got it. Awesome. And did you. Was that give us a scale of 1 to 10 on the complexity of integrating the OMS into that? Sounds pretty interesting.

Yeah. The reality is because of Agent Force's flexibility to hit topics through Apex and Flow, it was quite simple as long as the OMS is API capable.

Very cool.

All right, great.

Thank you very much. Appreciate that. Also, just quickly, I thought that the integration in Slack was fantastic. And then the dashboard, I was interested as well. Where was that dashboard showing up at?

So, yeah, you had built a dashboard.

Out where you can see the kind of the full view of the interactions and so forth.

Yeah. So that dashboard was built between data cloud, service cloud and loyalty management. And it was a custom dashboard that we did like lightning web components to bring it all together but it was accessible in all three of them so that we can depending on the user use case and permission set, they can see the information and.

All functional.

All functional. And we show it to you.

Yeah.

Very cool. That's awesome.

Congratulations.

Thank you.

And nice use of the full suite of capabilities. So very good breath. Bobby.

I do ask the very end was that coaches trying to get the right the bar at the end was it? All right, there we go. I spent a lot of time in Indianapolis being here for 11 years. No, I thought it was awesome. Just connecting all of the components together. Right. Seeing marketing cloud personalization, seeing data cloud loyalty service Slack and really seeing how AgentForce can be just a glue that connects all those through the platform, the APIs. So really cool.

All right, great work Pratik Desai. Let's engage. Phenomenal. Oh, I can hardly wait. Our final contestant number three from OSF Digital we have Rob Smith. All right, Rob, go get him.

So marketing retention and loyalty is really important. We've actually gone for a claims use case study today. I don't have a coal unfortunately in this presentation. So claims is something that you can't get humans to look at every claim. Right. It's far too expensive. Auto approved levels such as kind of so you can get things through quickly or open to abuse. And so this was previously an intractable problem, really difficult solved problem without AI. It's applicable across loads of industries. You can see them all on the screen there. So pharma, retail, all that kind of stuff. It's sometimes called guarantees, it's sometimes called warranties. Lots of different versions of this. Right. It's got massive business value. Right. So we increase revenue, sorry, increase margin because we have fewer payouts for abuse. We have very high customer satisfaction because their claims are dealt with quickly. It scales incredibly well. So we don't have to have huge contact centers and we have much better pattern recognition as a result. And let's not forget though that the human can still be escalated to. There's plenty of these claims that are very, very high touch and so we need that human still in the loop. We've got a headless agent on the website, so we're not using a standard experience cloud agent here. We're also backing off onto Heroku and we're doing an agent to agent thing so that we're backed by OpenAI to do the image analysis that you'll see later. And then we're using marketing cloud engagements send the triggered emails off the back of the claims process and also to do a headless agent that I'm going to describe a little bit later on. So a full kind of blooded use of marketing cloud engagement as well. And all this is backed up by the core platform and data cloud feeding the agent everything it needs to make sure we get this right. So let's actually go into the use case. I'm using a phone because everyone uses a phone these days and you're actually going to see me typing in real time here. Obviously not right this, but on the video. So I want to create a claim about my last delivery is what we're going to put in first. The first thing we need to do is classify the issue. Is this a damaged product? Is it a missing product? There's lots of different versions. So the agent's going to prompt me about that and I'm going to say in this case it's a damaged thing. Right. So I'm going to pop that in and it's going to come back to me and ask me who am I? I need to kind of identify myself as a customer. So I put my email address in. Of course, there's lots of different ways to do this in real life as we've found with our customers. And then I'm going to give it something, some of the proof of my claim. I'm going to upload an image and into the agent so that it can assess the product. This is where I just want to quickly describe our invoked agent. So every day we also have an invoked agent that runs over the caseload, finds any cases that have gone down a missing kind of rabbit hole, finds any unhappy claims using its kind of, you know, reasoning capability, emails the customer and escalates that case so no one gets left behind because it happens a lot in the claims process. Right. So back to the demo. We're going to hopefully see that the agent agrees that I've got a damaged good here. Not just that it agrees, but it's got a really in depth description about what's happened. You won't be able to read it very easily on the screen, but it's an incredible description of the image that I've just put up. And that's what that Heroku agent is providing us in this case. And it's raised the case for the customer service team with all of that together. Next we'll see the marketing cloud engagement email that gets sent and we use the image that was provided and the description generated all put together. So I'm a really happy customer because I'VE got something and I feel listened to. Not just a standard response but my response just for me, just for today. Right, so let's have a little look at the back end as well. This is obviously the case that's been raised. It's got the claim kind of analysis status. We could even have an agent take this even further and issue compensation right off the back of this. We just haven't built that bit out yet. In Agent Builder we can see the topic. So this is the identify visible damages. Plenty of instructions there to make sure it operates correctly and it's got the right guardrails and we've got the all important action that goes off to Heroku and that image analysis agent just there. And then finally we've got Marketing Cloud Engagement where you can see the trigger journey that's been sent this. We want to expand much further so that we can do follow up because we complete the customer360 by making sure Marketing Cloud Engagement sends these service emails. It's something our customers talk to about all the time. Okay, so what have you seen? You've seen a very, very high value use case. Huge application across loads of industries. It uses the generative and reasoning capabilities that we have in the platform. Great use of marketing cloud engagement to communicate with customers and as I say, complete the 360. We've seen invoked and headless agents at the same time which is a big new set of use cases and we've used the ecosystem beyond Salesforce because Salesforce is not an island and we do need to integrate with lots of other systems and OpenAI and AI in this case. We think it's a great use case. We hope you agree to. Thank you.

All right, thank you.

Thank you.

Thank you.

Great work.

Thanks.

Claim Force One. Is that what you called it?

That was it, Yep.

Just because.

Great job. OSF Digital Rob Smith. So we don't have much time. We'll keep it moving. Rob.

Yeah, really cool.

Yeah. Oh, there you go.

I've got to use this back into the old habit. Okay. Really cool. You know, I talk in marketing a lot. Our biggest currency is trust and from a customer perspective, I loved that, you know, single shot at the end where they can just see everything wrapped up nicely. It instills trust from, you know, in terms of the company brand back to the customer. You know, the trigger journeys are great and as you mentioned, being able to take it outside of just the island and integrating with the platforms that you need to be integrated with where the data lies in is really important.

So.

So no A really strong use case.

Great.

Steve.

Yeah, the use case was really nice because I felt like, how many times do you have a problem with a product or a service but you don't wanna call in because it's such a pain to try to reach someone? And from the company side, they're missing out on that feedback that that product is not really substantial or not working properly. So there's a big gap and most of it's because it's just a pain in the butt to do it. So you've really shown a solid use case here. Thank you. And I loved all the different ways that you used it. I do. One of the areas I was wondering about a little bit was when you're showing the claim and uploading the images, was there an opportunity, do you think, to be able to actually render the images in the Agent Force? Is that something we need to put into Agent Force to make sure that we can show images, not feedback for you, or was that an opportunity?

Yeah, it's a really good point. At the moment. Agent Force doesn't, as far as I know, someone might correct me in the room, don't get me wrong, doesn't have the ability to actually show an upload that you've given. It just has a little thumbnail. So that would be a really great thing because you'd be able to sign us.

We'll do that.

I'm sure it's in the roadmap.

Everything is, as far as I'm concerned, fantastic use case.

Great examples. Thank you very much. Great technology. Thanks, Rob.

I was literally watching this and thinking of, I had an insurance company mail me a check and I had moved addresses, so I never got the check. And I'd go online and file a claim and I get an email that was just like, we have your claim. Like, no number, nothing.

Right.

And so, like, just when you sent the email back with the description, the image, to recognize it, I'm like, oh, man, that can go so far with just like the customer loyalty and the experience of just like, we heard it.

Right.

I still don't have that check, by the way. So, you know, really?

Is that for your roof?

The Tahoe roof?

No, no, this was for something from the dentist.

Okay, okay.

But, yeah, it was really just. That really struck home with me.

All right, well, let's applaud for all three contestants. Excellent, excellent work. Very, very high level. All right. The judges will. And I don't envy the judges. Tough decision. They'll be tallying their votes and you all just scan the code up here, you will get to vote. You only have two minutes to do this, so do it quickly. Only vote once. There's no guardrail here. We're going to have to trust you. We will know who you are. The ballot stuffers. We're voting. Tension is building. We're thinking about business value. They all seem very practical to me. Creativity. How creative was it? And finally, the demo itself. Very slick demos. You can tell these guys are agents, agencies, they are seasoned marketers. And I will mention this music is original music written by someone in the room. Am I right? There he is, that guy right there, Kyle.

Nice. Does it come in book form?

Yes.

Okay.

He's a talented musician. You can hear. This is pretty cool.

Very nice.

Judge report. All right, the judges are. Oh, look at that.

All right.

Moment of truth awaits. Okay, now we're going to do a tallying moment. And I was told that I might have to kill like 90 seconds. What are we going to talk about? Don't DIY your agents. Does anyone know why? Look at this LLM. We all know what an LLM is. I can subscribe to it, I often do. But in order to enterprise ify your agency, you need a lot of other stuff, don't you? You need all of this. You need a reasoning engine. We have the Atlas reasoning engine built into AgentForce. That's the kind of a way to negotiate among agents and have a conversation back and forth to the LLM and make sure everything's working. You need to be able to implement rag retrieval, augmented generation, which is great. Grounding these agentic workflows in your own first party data. Data that's sitting in data cloud. It could be unstructured data. Unstructured data, very useful to agents and we make that available in data cloud. Then there's all these apps. We have the world of apps that you saw actually very nicely demonstrated, including Slack. Thank you for using Slack. And then unstructured data, metadata, semantics, all the goodness that's built into the Salesforce platform. Our mathematicians are now calculating the scores. It can't be that complicated.

But Marty, while we're getting the scores tally, can I just add. All of these were fantastic. Thank you so much. And I wanted to give them all a big hand for. I think that the thing that really stood out here is just incredible innovation across all the teams to go beyond just what we've put into manuals or into into our guidance from Salesforce. You really used your knowledge of the products and applied it through customer scenarios you work with. So congratulations to all of you. And I would love to work with all of you to figure out how do we make some of these things broader and more applicable into the product as well. So thank you for this great work. Appreciate it.

Yeah, it was excellent. And also the quality of the demos is very high level, sort of like professional level demos. So I was impressed by all the presentations. I appointed myself tiebreaker, by the way. If there's a tie, I decide. Looks like there might be. Oh, boy. I wouldn't be surprised. Three way tie. That can't happen in that basketball thing, right? With the Pacers and the. No, there's no ties. Okay. All right, moment of truth. So we will start with our third place contestants. And again, all excellent head. And they were chosen out of a field of many dozens. So they're all winners. Our third place contestant is in this envelope. We'll ask them to come up and receive their trophy. It is OSF Digital.

Thank you.

Thank you. Appreciate it. Thank you.

Great work, Rob.

Thank you.

Photo op. Absolutely. All right. Excellent. Thank you. Thank you.

Thank you.

All right.

May I ask the two remaining contestants to come up please, together.

Do they get to hold hands or anything like that?

You can hold hands. You're wearing the same jacket. In fact, you're dressed the same. This seems to have been planned. These guys are good. Very good. All right. Once I tell you who came in second, you'll by elimination know who came in first. That's why they're both up here, thought through this. All right. And in second place, doing an excellent job is Ateco. Congratulations. Excellent. Thank you.

Congratulations.

Thank you. And our grand prize winner. Listengage. Bravo. Bravo.

Larger.

Sophie.

Congratulations.

Congratulations.

Congratulations.

Yeah, of course.

Thank you.

Congratulations. And not least love the video. Arch tech. Thank you. Thank you all.

I think we should all get this picture.

Come on, listen, engage. Excellent work. And thank you audience for participating and thank you for your votes. And I want to thank our panel.

Again.

Excellent judging. I have a couple more announcements here. Quickly, very quickly.

Housekeeping.

Am I right? T shirt vouchers. Those of you who have T shirt vouchers, you can redeem them for T shirts. Redeem on level one. It's actually just down there somewhere. W183 between 1:30 and 2:30 today. And I think that's it. Marketing cloud keynote tomorrow, go see Steve.