Name: Garbage In, Garbage Out: Why Clean Data is the Secret to Salesforce Success
Uploaded: 2025-05-22T17:30:56.468Z
Duration: 1 h 28 s
Description: Garbage In, Garbage Out: Why Clean Data is the Secret to Salesforce Success

Transcript for "Garbage In, Garbage Out: Why Clean Data is the Secret to Salesforce Success": Okay. Well Hello, everyone. Welcome or welcome back to another Salesforce banner webinar. Today, we'll be diving into the importance of clean data and why it's a secret to Salesforce success. We've got a big session today, and there's lots to cover, so I won't keep you for long. Just some general housekeeping. If you do have any questions, please do submit them via the q and a box, and we'll get around to as many as possible. This session is being recorded and will be shared in our follow-up email after the events. So, just to cover a little bit about what we do here at SalesforceBen. So we are building towards a future where there is a clear path to success for anyone wanting to embark on or accelerate a career within the Salesforce ecosystem. As a part of this, we offer practical, relevant, and accurate information through many different mediums such as articles, newsletters, events, YouTube, and podcasts as well. So there really is something for everyone. If you haven't already had a look through any of what you see on screen, don't forget to check them out later. But just to introduce your speakers for today's session, we do have Dixon Blue, senior account solution engineer at Salesforce, Peter Bernardi, chief executive officer at ElevateActual, and Kayla Condello, knowledge and training manager at FormAssembly. So I will hand over to you again. Thank you, Saima. Thank you, presenters. Dixon Blue, Peter, and Kayla, good to see you all. And audience, great to see you as well. My name is Josh. I'm gonna be a bit of a moderator today. I'm gonna take things over from Saima on the screen share and chat you through our agenda very quickly. You know, today, we're gonna be talking about kind of a wide ranging topic, you know, data cleanliness. We're gonna be talking about why clean and effective data is so important, for your processes. And to start off, you know, if you're familiar with FormAssembly, we're a data collection platform. We care about data. We care about what you do with that data. We wanna empower data stewardship. And because we're interested in that topic, we tend to go out and survey audiences about that. So to start our conversation off today, Kayla's gonna walk through some, some brief survey data from a recent report that we conducted. But before we do that, I wanted to open a poll of, today's audience. We wanted to hear from you as well and get a feel for what challenges you and your organizations have experienced when it comes to data collection. So I'm gonna open a poll. You should see that pop up, on the right side of your screen here. If you don't mind giving us a little information, you know, just what you're seeing in your industry, in your organization, we'd love to know what challenges you're seeing. And then you're gonna hear from, our friends or pardon me, not our friends, but our survey audience respondents, in just a moment. Thank you. I'm seeing the votes roll in. Looks like a lot of people, are starting well for an early, an early winner is outdated systems and lack of integrations and, oh, manual and time consuming corrections are taking a lead. It's a little bit of a horse race with this poll, and it looks like it's staying in the lead. Kayla, I I'm gonna hand things over to you. Let me go ahead and pull our agenda up. After we hear from Kayla, we're gonna hear from Dixon on data formatting for AI and data cloud, something that I'm sure you're all very interested in, especially after all of the AI news has been pouring out this week. We're gonna have a brief discussion with Dixon and Peter about, applications for Agentforce that they've seen, in their respective industries, the industries they serve. And then last but not least, Kayla's going to show you a little bit of form assembly and how, we can help you take control of the quality of your data. Just another tool to help you, in this battle with dirty data. But speaking of dirty data, that's enough for me. Kayla, I wanna turn things over to you. Thank you so much. Hi, everyone. I'm Kayla Condello. I'm the knowledge and training manager at FormAssembly. And, today, we're gonna dig into something that's become mission critical in 2025, what Josh just talked about, which is clean, high quality data. Because whether you're trying to, like, personalize your journeys in Salesforce or get results from your AI, garbage in will always still mean garbage out. And you can have the world's best CRM setup and, like, the flashiest AI tools, but without clean data, like, let's be honest here, it's just expensive guesswork at that point. So to dig into that question, we ran a survey late last year. We focused on the state of data collection and data use, and we asked, data professionals what was getting in the way of clean, usable data. Where do they see the most room for improvement? And we pulled together the results into a full report, which you will be getting after today's session. But for now, we're going to take a quick look at a few of the insightful excerpts from that. Not the full picture, of course. We don't have all that time, but, some thought starters. So if you're like me, once you see these stats, your brain's gonna already be jumping to the implications for your own organization. So let's take a look at what the survey showed us. Almost everyone, 99% of those respondents said automation would save them time. I'm sure we knew this. So here's the breakdown on that one. 70 percent said it would save a lot of time, and 29% said it would save some time. And 1% said it wouldn't help at all. But that's not just a trend. That's a signal. And people want automation. They're confident that it's gonna make a difference for them. And so we asked a little more about manual processes, and that's where the cracks started to show. Teams were reporting that a significant chunk of their data collection work is still partially manual, and with that comes this familiar set of issues. Right? Time consuming workflows, integration problems, inconsistent and error prone data. So even when we are, you know, throwing automation in the mix, inefficiencies are still deeply embedded, and the downstream impact on that is not subtle. So 80% of those respondents said they're facing data quality issues, and when they that's when they're trying to scale, automate, analyze their data. That's gonna include incomplete records, mismatched formats, fragmented data across their systems, all of which undermine everything from lead follow-up to reporting accuracy. So, of course, automation is a powerful goal, but clean connected data is that foundation that you have to build on because at the end of the day, automation can't fix bad data. It will just move it around faster. So let's go to the next slide and talk for a minute about the human cost of all of those manual processes. So half of the survey respondents said manual work was eating up team bandwidth, introducing some errors, and creating integration headaches. And 44% of them said it slowed down scalability, which made perfect sense actually because it's really hard to grow a smarter system when your team is still buried in the maintenance of the old systems. So here's where the impact really kinda came into focus with that time. On average, respondents said that they were spending more than fifteen hours every week on manual data related work. That's time spent entering, checking, moving, correcting information. That's two full workdays a week every week. And for more than one in 10 of those people, that number jumped to twenty six hours or more. So while that time is being spent, it's pretty fair to ask how much of this could be avoided with better processes and cleaner data. And we are seeing that some teams are already starting to reclaim that time. Among those who've automated parts of their data processes, 70% said they're saving up to ten hours a week. That's time that could be you know, it was used to you they used to spend on those repetitive tasks. Let's be honest here. It's time that could be spent more efficiently, more effectively. And when asked where they see the most opportunity to go further, the answers were pretty consistent. Correcting data was a huge one. Processing submissions, manual data entry, you know, the areas where smarter processes and cleaner data from the start can make a huge difference. And so, honestly, that's obviously the part that we get really excited about. Not because it's flashy, but because it's clean, and clean reliable data has this ripple effect where it simplifies your workflows, it improves your reporting, and it strengthens your integrations, clearing the way for all the other stuff you wanna build. So what does this add up to? We are gonna move to the next slide and talk about the downstream impacts. So these are gonna show up everywhere from back end systems to the customer experience. The chart on the side here is gonna show the top data related challenges that organizations are facing overall as per a survey. Leading the list, you can see accuracy and consistency. That's followed by things like data security, integration, reporting. And all of those might sound like some system level issues, but they obviously have real operational consequences. Broken processes, frustrating customer experiences, struggling with accurate reporting. And as Dixon's gonna show you shortly, these issues become even more critical when you bring, like, AI or Salesforce data cloud in where that clean, well structured data isn't just helpful, kind of a requirement. So it's not surprising that accuracy and consistency top that list because without those, nothing downstream is gonna work the way it should. So that brings us to the next big question, which is technically our first big question. What does clean quality data actually mean in 2025 and how do we get there? It's gonna come down to a few core principles. The data needs to be consistent, look the same no matter where you're accessing it. It's complete, correctly formatted, and accessible where it matters most, whether that Salesforce reporting dashboard or the form someone's filling out right now. And getting there doesn't have to be complicated, but it does have to be intentional. You have to require the right fields, validate input as as collected, and monitor how data is flowing through your systems. All of those small steps are gonna have a huge impact because at the end of the day, good data doesn't just happen. You really have to build for it. So, obviously, a little bit later, I'll be walking you through a high level look at what that actually looks like in practice. I'm gonna be using a job application workflow that you'll actually probably recognize. It's not gonna be a step by step demo. It's gonna be kind of a bird's eye view of how clean validated data can move through each stage and update Salesforce in real time to keep your systems aligned from start to finish. But first, Dixon's gonna show us why this kind of foundation becomes absolutely critical when AI enters the picture. So, Dixon, over to you. Absolutely. Thank you so much, Kayla. It's nice to see all of you on our webinar today. My name is Dixon Blue. I am a senior solution engineer at Salesforce, more specifically on our nonprofit team based out of Denver, Colorado, And really excited to to be here today to talk about data and in this case, more specifically, how it feeds and relates to AI, the hot topic of the year. So if we wanna go to the next slide, we'll get it rolling. So as you all know, AI is top of mind for organizations and enterprises across pretty much every industry. I know it's top of mind in the nonprofit one where I focus all my time. And we know that AI becomes tremendously valuable when it works together with people to deliver results. But like Kayla alluded to, data is the lifeblood that fuels AI. Right? So getting clean and integrated and contextual data is what deepens AI's understanding of our business, our constituents, customers. And for it to deliver accurate insights and support solid informed decision making, it needs to get fed high quality and reliable data to guide those actions, and get to specific outcomes. So, Josh, if we go to the next one, yep. Oh, ahead of me. Incredible. I wanna talk a little bit about where we are in the AI wave, the AI revolution. Right? We're now in the third wave of AI with assistive and autonomous agents. So as we keep going, right, up and up and up, data becomes more and more critical for success. Right? We all can laugh and giggle if we, you know, chat with, with an AI assistant and it tells us that one plus one equals three. Right? I used to play I play chess with chat GPT occasionally just to pulse check how well it's doing. And the very first time I did it, I said, I make this move, and it said, I make this move, and then indicated that it had in fact made two moves, neither of which was the one that it said that it had made. So, you know, it's all well and good to have a chuckle out of that. But when we are a business or a nonprofit or an organization, and we're going to rely on, an AI assistant or an agent for things like growth or for numbers or for, you know, next best action recommendations, they better be correct. Right? We don't have we don't have leeway to sort of laugh off, things that are wrong. And when we do things like deploy agents that are customer constituent facing, the stakes get even higher. Right? Our reputation as a brand is now at stake. So and data is at the heart of all of this. Right? Not just sort of structured data, but also unstructured data. So a lot of, you know, data that lives in PDFs or Word documents or knowledge articles or images or whatever. Right? And a human agent might be able to look up multiple screens to answer all of those questions. But if you want an AI agent to be able to respond to those accurately, it has to have access to all of that data, and the data has to be good. Easy. Right? Unfortunately, Keith, as, you know, I think all of you are probably aware based on the poll, enterprises and organizations today struggle a lot with data silos and lack of context for AI and poor actionability for their data. And, again, this concern gets magnified even more in this agentic era in which we find ourselves. Right? AI agents need the right data and the right context to produce actionable insights. For the most part, this, like, isn't really possible with other agents or copilots. They can't get access to all of that data that is governed or secure or unified in order to deliver, on its agentic mission. Right? So, if we go to the next one. Gosh, thank you. Enter, Salesforce data cloud. Right? And data cloud, which, you know, may or may not have heard of at this point, is the trusted foundation that fuels your customer three sixty and also AgentForce, Salesforce's agentic AI platform. Data cloud is a hyperscale data engine that's built directly into Salesforce and generally comes with a couple of major benefits for organizations from a data perspective. The first is going to be customer three sixty, full blown. Right? With deep platform integration. Two is you get trusted and contextual data for agent force, super important. And three is you get autonomous actions inside the flow of work. So data cloud provides all of the unstructured and structured data and metadata that, AI needs to produce actionable insights that are grounded in your organization's customer data and records, which is really, really helpful. Right? Like, if we could hire an intern that knew literally every piece of information possible about our organization and could recall it like this, what might we have them do? And what sorts of maybe manual processes or inefficiencies could we have them cover while the rest of our team focuses on high value, things instead? Right? Literally very exciting to get to to dream around that. Right? And at Salesforce, at least data cloud is the engine that powers all of this. It takes all of the structured and unstructured data that lives in data lakes or data warehouses or CRM data, brings it all together and uses it to, synthesize full 360 degree views of customers and constituents, and also makes that data available to be actioned in agent force as well. So if we go to the next one, and I wanna talk a little bit about the power of data cloud, why people are turning to it. Right? So I've talked about the integration of the structured and unstructured data from everywhere. Right? And this has been something that people have talked about for as long as enterprise software, I think, has been around. Right? Which is it would make a lot of sense to develop a full 360 degree view of who your constituents, your customers, your people are by bringing all the data together. But historically, that has been a really challenging problem to solve. And Dataglub tends to make it much easier. It can pull in and connect to data from literally any source, and then harmonize and synthesize that data together to generate this 360 degree view of people. Right? All of that is obviously deeply integrated into the Salesforce platform. And so by bringing bringing all of this data together in data cloud, one of the major benefits is that we've eliminated, for the most part, these data silos that we're talking about a minute ago and created this comprehensive view, of people. Right? So it's really, really simple but very powerful idea that sets Data Cloud apart and makes it sort of a much more powerful solution to this data problem and gives us a phenomenal platform from which we can feed AI to get the biggest impact that we can, right, to make the AI as powerful as possible. I should go to the next one. And I wanna talk a little bit more about that. So data cloud maximizes the potential of agent force by grounding the agents in an organization's customer specific data and then using things like retrieval augmented generation and hybrid search capability to to bring that data to life. Right? So from a data cloud perspective, you have these built in rack capabilities, real time insights, context from unstructured data or structured data to emails, product photos, voicemails, whatever sources might be. Right? Are all context and all power for Agentforce to leverage to do whatever its job happens to be as effectively as possible and resolve issues as possible. Right? So when we leverage data cloud as the the data platform foundation, teams that are using AI can build trust with their customers, with their constituents, through this governed and secure platform that natively manages the data access and the security and compliance. And lastly, we're talking about activating data. That's is that's really like that sort of, like, last 10% of the data journey becomes really, really important. Right? You can have the cleanest and the most well organized and the most effective data in the world, but unless you can then do something with it, it's not actually all that useful. Right? So with data cloud, we can do things like like activate data for AI, for platform automation, for analytics. Right? And drive action seamlessly, as we say, inside the flow of work. Right? So we can leverage data cloud to improve how data that you how the data that you have gets used across your entire organization. Right? So in nonprofit, for example, we talk about this a lot from a fundraising standpoint. Right? Every nonprofit needs to raise funds, and that's all well and good. But if we also can surface and bring to life data from other parts of the nonprofit, like programs, for example. If I'm a fundraiser, it is much more valuable for me to be able to take program and impact data and weave that into how I communicate with my potential donors. Right? To help them fully understand the impact of the mission that they're supporting. That's a much more powerful story to be able to tell, and is enabled by things like data cloud and that full 360 degree view. Right? And the actionability for the data doesn't just stop with people. Right? It carries on through to AI and to agent force so that agents can also act on insights or predictive models, surface answers that would take humans forever to find, retrieve knowledge, do routine tasks, automate workflow. We can do all of this right in the flow of work that allows for sort of seamless handoffs to people as necessary. Right? And, data cloud as part of the Salesforce platform makes all of this possible without needing to do a lot of investing in really, really expensive, like custom data integrations and that generate a lot of technical debt and, you know, eat up a whole lot of time from a maintenance standpoint. So I'll talk just really quickly about how data cloud works. So it starts by we're gonna start on the left and go all the way through. It starts by connecting, like we talked about, all of those disparate data sources across the organization. So whether it's data that's in Salesforce orgs here or there, external applications like SAP or NetSuite or Workday or whatever it might be, or data lakes and warehouses, Snowflake, Google Cloud, AWS. Right? Whatever the data source might be, we wanna connect to that first. And then we'll take all the data that we have and harmonize it and create one unified view of our customers that every application across the the Salesforce three sixty and Agentforce can access. And then after all that data is unified, we're gonna make sure that it is secure with governance and security. Right? We have extensive tools that are built into data cloud to help people confidently manage data access and security while remaining compliant which is hugely important especially in this wave of of AI. Right? And then lastly, like we talked about a little bit, we can activate all of that data, those unified profiles across any application or workflow or AI agent. Right? So, how does that actually work? Right? We've talked about data cloud powering Agentforce. And Agentforce runs on this tool called the Atlas Reasoning Engine, which is basically the the fancy name that we gave Agentforce brain, and lets it that lets it think for itself and it enables it to retrieve relevant data to reason and to act. Right? Which is really, really exciting, really, really powerful. So the the Atlas reasoning engine handles this really wide variety of interactions including, things like multiple layers and deeper thought. Right? So, like a question that's really straightforward like, hey. What's the status of my giving portfolio? My employee like, really basic reasoning and get rapid responses. And then a deeper question, like, I don't know, what might be the right investment for my kids' college fund based on current income and risk preference, would use, like, enhanced reasoning and advanced data retrieval to improve the answer, by making sure that it has as much context specific data, as possible. Right? So, essentially, we've given agent force a brain that can allows it to think for and make decisions for itself inside of, the guardrails that we give the agent, which is really, really cool and really powerful. And this is what separates that sort of agentic AI from the maybe like chat bots or copilots or whatever of the past. Right? Those are really good for, getting information and surfacing it up and all that. But when we need AI and when we want AI, quite frankly, to make decisions and do different things, take different actions depending on situational context, it being able to think for itself and being grounded in all of this really solid contextual data is massively important. And so all of this is really good in theory. I wanna spend a little bit of time talking about it practically, and I think that, I believe that Peter and I are gonna spend some time talking about potential applications for Agentforce as well. But just, you know, the nonprofit space being near and dear to my heart, I wanted to bring some practical examples up of Agentforce in. Like I said, I talked a little bit about fundraising, marketing as well, program management, right, IT, literally anything you can think of, as far as areas that your organization operates in can all be hugely benefited by leveraging Agentforce or Agentic AI. And as we all know, it needs solid and clean and good data to feed it for it to be as powerful as possible. Great stuff, Dixon. Thank you so much. And I love that you brought a slide with applications because that's exactly what we're gonna talk about next. Peter, I'd love to have you join us, on stage. You know, the reason I brought, Dixon and Peter into today into today's session is I know both of them in their respective industry or folk industries of focus, you already started, you know, like building out Agentforce applications for their various clients and obviously, you know, you know, Dixon, working at Salesforce. So, of course, you get to see this a lot sooner than a lot of people, but I know a lot of organizations, and I'm I'm I'm betting a large, portion of our audience are still trying to wrap their heads around where can I bring agents into my business or organization? Where can I leverage the power of AI? And the pressure's on. You know, like, at our own organization, we feel the pressure to use AI to, you know, unstopper efficiency. As a marketer, I love using it, you know, to, like, essentially hop over writer's block. It's a great way to just power pass that first sentence in an email when you're really struggling to write some copy. But I'd love to hear, you know and and, Peter, let's start with you. We'll we'll give Dixon a break for a second. I'd love to hear, you know, the with your clients, what kind of applications are you seeing people interested in? What are they building with your help? What are you seeing out there? Yeah. Definitely. We're in industry, just for those not familiar, is mostly higher education. So kind of the other side of Dixon nonprofit coin in a sense. Really, the biggest driver of that minimum interest is all top of funnel recruiting, so drive revenue. How can we drive more conversions, more intelligent conversions, the right conversions for degree seeking students, non degree, some college adult returners? How can we target those types of of prospects with more intelligent, either automated marketing or as soon as they get to a page, interacting with an Agentforce agent to direct them to an application for a full degree program. Or maybe in that conversation, they understand based on data that the organizations already have, this person might be a better fit for a credential or a certificate, a non degree program. So we can divert the appropriate prospects to the appropriate, really, swim lanes that they're gonna hit so that they're applying for the right things and thus more likely to convert. So really top of funnel is always number one. And then once they're actually applied in an applicant experience cloud site, we have an embedded agent to help get them to actually confirm their place actually enroll. And then further down that learner life cycle, once they're there, that front level, first tier advising support, okay, you need to chat with an agent. Let's help schedule the right meeting with the right person or get more information about financial aid. So really taking a long look at it to have it integrated in each part of that learner life cycle. But first and foremost, we gotta drive top line revenue. Okay. Interesting. And and, you know, that makes sense. Really, probably, in any industry, I think, you know, proving out ROI on agents and Agentforce is probably gonna be a a a high pressure for organizations. Dixon, anything, you know, nonprofits also have funnels. Anything, you know, from your the slide of use cases that you pulled that you've been really excited about that you've seen, NPOs, interested in? Yeah. Always always things to be excited about, I think, in in the AI space. I wanna just and I'm just really quickly before I do that, I wanna just say from a, like, a how we think about this perspective. I I think that a lot of organizations that I talk to are like, oh my gosh. AI is here, and it it's going crazy, and we need an AI strategy. We need to put AI in, like, all of the places of the business, and it's gonna fix all our problems and make everything sunshine and rainbows and every oh, it's gonna be amazing. And that's, like, not really a great way to approach it, actually. Right? You know, AI, when we think about it, is a tool that we have in our belt that we can leverage to help us achieve more, like, specific outcomes. Right? So when I think about AI and where we can leverage it, I like to think about, okay, where are we trying to go and why will using AI in this particular situation help us get there better than something else might? Right? So and the way that I usually like to start off people thinking about that is what are the most painful processes or areas of your organization. Right? Like, what hurts the most from a people perspective or, you know, from a process perspective? Okay. Great. That's where we wanna zero in and start to be like, okay. Let's use AI here maybe to lift some of that burden. So, like, I work with a lot of YMCAs, for example. And YMCAs have tons of programming that they deliver right across a huge variety of areas. Right? Fitness, childcare, summer camp, aquatic, all this stuff. And so navigating those program options and making sure that you're finding the right one and getting registered for the right one is really challenging for constituents. And so we've got a few different, agent use cases right now around guided program registration where folks can come to, like, an experience cloud site and say, hey. You know, I'm trying to get my five year old signed up for swim lessons. They haven't taken any before. I need something that's, like, on Tuesday and Thursday. And instead of having to go through the entire program catalog and make sure that I'm finding the right one with the right prereqs, the agent can go through all of that, surface it for the person, confirm that that's the right thing, and then complete the registration for them. So from an experiential standpoint, it's astronomically better, than it was before. And then one other thing that I that I like to do from a from a where might we leverage agentic AI standpoint is think about, like, what might be a thing that would be completely ridiculous to do if you were doing it with people? But now that you have AI, this would be cool. Like, for example, it would be absurd for a nonprofit to write a completely personalized, acknowledgement email for every single donor in portfolio. Right? From a model perspective, that's not very feasible. But if we put an agent on that particular job, it's really easy to have it synthesize all of the data that we know about this person who's given us money and write the comp and write a completely personalized acknowledging email for them and have that go out. Right? So those are those are sort of the two ways that I like to think about Agenctic AI. Right? Like, what hurts the most or what might we do that would be ridiculous to accomplish with just people? Oh, I I love that. You know, I'm I'm thinking of my local it's not the why, but it's a a local rec center. And I'm thinking of their program guide, and all the programs have, you know, like, several different day opportunities. And as a, you know, sign up for those, it's always like, oh, man. Just Tuesday and Thursday. You know? Like, it's so it's so difficult to get into those. And that's not that's not what you want. Right? You want people's hobbies and interests to guide them into these programs, and, that's a that's a fantastic use case. And I think in, you know, one of those painful things that you mentioned, and this might be, especially, applicable to higher ed is the AI doesn't need to go to sleep. So if somebody is, you know, trying to apply to something, register for classes, classes, you know, at 3AM as students tend to do, you can have someone there to guide them in the process, whereas normally the bursar's office or the registrar's office is gonna be, you know, closed at five. You know, Dixon, you mentioned something that is on my list. It was at the end. Let's jump around a little bit. It's the misconceptions that organizations are having. Right? It's the do I put AI everywhere? Is that what I need to do? I wanna talk toss to Peter here. You know, when you're talking to organizations, what questions they're asking, you know, like around AI or or agent force in particular or even data cloud? And it's like, oh, that's not what it does, but I need, you know, like but here's what it is. What what are people coming in? Yeah. I think probably the number one conception misconception we've seen in getting to, like, what Dixon's talking about of building an adoption strategy is that it is not a 100% rip and replace for an admissions counselor. I think there's a lot of fear, especially in higher ed right now, that, oh, we're gonna turn on agent force and have to fire all of our staff. It it's quite the opposite. It should be and the the the blend is, like, 45, 50 percent whether it's content, like Dixon was saying, finding the right programs for the right people. So when a learner comes to your landing page, they're not gonna only chat with Agentforce forever. Wanna close that gap, exactly what Dixon said about program and and application enrollment with make it so that a learner could get to your landing page and apply for college in a chat. Like, that that is reality. But once they've applied, if they have that appropriate data, which can be hydrated with a FormAssembly form, we have really good applicant data so that sends them to the right admissions counselor. So you're not having to do this weird round robin with the flow based on who's on from 9AM to 10AM, like you said, Josh, where the agent's always on. So it's tagging them in the appropriate way, and we are not replacing staff. We're allowing them to do their job better, easier. Get rid of all that painful stuff that people have to do, like Blue mentioned, that now the agent is gonna take that lift off of them, allow those admissions counselors to do what they do best, which is build connections with those learners and get them enrolled. So I think the biggest misconception and really a big fear point, especially in higher ed, is that it's it's gonna take over and we're not gonna have any staff left. Quite the opposite. We wanna empower them to use AI in a responsible way so they can serve those learners better. Yes. And, you know, I think I'm so glad you keyed in on that fear. It's something that I think every organization needs to reiterate to their employees because, you know, when we talk about agents, you know, there's this idea of this autonomous, you know, non person person coming in and and fulfilling key roles. And what you know, that will prevent people from adopting AI. You know, that fear is something where they won't they won't feed into the processes correctly. And, you know, like, really, what what we should the viewpoint that, you know, like, I've adopted, so I can say this personally. I don't wanna, you know, like, put this as a should. But, you know, like, people need to recognize that, this is an accelerator for your work. This makes your work easier. To Dixon's point, it takes away the most painful work that you have to do, and makes things easier for your constituents, your customers, you know, your volunteers. There's some really exciting stuff ahead. Quickly, because I do wanna show you know, we've got Kayla still on the line to, to show us where FormAssembly can can help play a role and specifically the accuracy and cleanliness of that data, you know, where where data intake. I wanted to ask the two of you. You've already kinda talked about what exciting successful implementations that you've seen. Is there any specific ones that you wanna call out that, you know, like, weren't on your slide, for instance, like you're seeing because my favorite thing is when people are using tools to do things you'd never imagined. Is there any really interesting kind of, like, out of the box use cases that you've seen, some early success with? I've been just just to give one always give a shout out to Unity Environmental University, presented at Education Summit of Briuna, their agent live, first higher ed institution, at least in The US, to do so, and having that just just getting live with Agentforce, data cloud backing it up, using Education Cloud as a student information system Like that, having all of that work together where, you know, you see the presentations and you think, oh, yeah. That's a great like, they just showed that whole integration landscape. It's like, yeah. That sounds nice. Maybe not for my institution, but small growing institution that's taking distance education as a priority. They've done it. It's live right now. Like, it it is it is possible. So I think getting over that fear hump, a little bit. Personally, I am a very self trained, poorly practiced JavaScript developer. So I use it to refactor my code to make it, you know, efficient at all instead of whatever I'm doing. So I think it enables a lot of skill skill gap posing with that as well. Whereas strong form assembly background myself, great. We can make wonderful forms super connected to Salesforce. Okay. Let's bring a little bit of extra UI layer on top of that with an AI assist, something that, you know, someone in my role wasn't doing before. Having more of a managerial role, it's okay. You too can be technical with a little bit of of know how with AI. Nice. Dixon, anything to call out? Yeah. Absolutely. So, as has been stated, I I work on Salesforce's nonprofit team. One of the things that we do is we offer a a power of us program where for any five zero one c three nonprofit organization, their first ten licenses of Salesforce, are free, and they get to use the the tool for free. It's part of the way that we give back. And there are a lot, like, tens of thousands of nonprofits that do this. As you can imagine, that generates a lot of questions, that are Salesforce related. Right? And we don't necessarily have a team that's big enough to handle or field all of those. And so we were like, what if we spun up an agent for this, grounded it in all of the, like, product documentation and, like, the Salesforce help documentation around nonprofit cloud and everything like that, and a little bit of sort of, like, sales ops kind of things as well, some internal stuff. What if we grounded the agent and all of that and had it take the first pass at, like, all a lot of these questions that are coming in, around how the the p 10 program gets used and how people, leverage it effectively. And, you know, they they released this recently, and it's already handling, like, 80% of the volume with, like, over 90% accuracy. So it's pretty wild, and it's one of, my new favorite stories of, like, a, how the tool gets used effectively in general and also how, like, we use our own tools at Salesforce to, to do things a lot better. Yeah. So Right. Yeah. Well, you guys are the experts. I would expect you to be. But but 90% accuracy, 80% of the queries, that's incredible. Yeah. That's amazing. You know, thank you so much for sharing your experiences with this and and and and Dixon for the presentation earlier. I'll ask you to stick around, because we might get some questions for you if you don't mind, but I wanted to bring Kayla onto the stage. And before I, before I let her, start showing off how we see FormAssembly kind of assisting in this data quality conversation, I wanted to make, you know, just a quick announcement here to the audience. Some of you may be familiar with FormAssembly. You know, we've been around almost twenty years. We've got a very strong Salesforce integration. Just this year, we released a redesigned version of our Salesforce integration that is giving even greater control to data professionals like yourself over when data enters Salesforce, what data enters Salesforce and your other systems. You know, we support HubSpot, and we've got a webhook connector as well. But the control that we're giving people, we really expect to help with this data quality issue that so many organizations are, are are experiencing. That's why I've invited Kayla on today, specifically to show you how that works in practice, and she's prepared a an excellent demo and I think a very, a very interesting use case here. So, that's enough for me. Kayla, I'll turn things over to you and, let you get on with the show. Awesome. Yeah. I'll go ahead and share my screen then. Let's go ahead and get this. There we go. So, yeah, I'm here to talk to you about a workflow that I built in here. It's a really familiar process. It's a job application. So, you know, most of us have either filled one of those out or reviewed at least one. So it's a perfect backdrop for the idea that I'm gonna illustrate here, which is how do you build a process that collects clean, structured data from the beginning and then still allows that data to grow and improve as it moves through your system. Because let's be honest here. Applicants are almost always, they're putting their best foot forward. It's not uncommon to see them, like, check advanced for Salesforce skills when what they really mean is I ran reports one time. And that does not make them dishonest. It makes them human. But it also means that the data can easily end up in your system looking a little more complete than it really is. So how do we collect data that's not just clean, but realistic and updatable and trustworthy over time? So I'm gonna show you what that looks like here in the FormAssembly workflow builder. But just to be clear, this concept is not just for our workflows. This is gonna apply no matter what tools or platforms you're using because what we often see is kind of one of either two extremes. Some teams are kinda gonna overcorrect, and they're gonna lock in that applicant data like it's carved in stone, even if it turns out to be outdated or incomplete. And others are gonna kinda patch things later on once the inconsistencies have already spread. But the real opportunity lies in the middle there. Validating what you can upfront and then giving your process the space to refine and update that data intentionally as that context comes in. So that's the mindset behind what I'm about to show you. I know that was a lot of talking. I'll show you the thing now. So this is going to start at the top with submit your job application. Our workflow has steps in it. This is the form step that's going to collect the data, and I will actually show you what that looks like here. This is the form. So in this form, the applicant's going to see a set of structured validated fields. They can enter their name, email, phone number, all these are required. I've got, you know, validation checks in place to prevent any poor formatting from slipping through. They can put in their desired salary, years of relevant experience. Both of these are validated as numeric fields, keep things clean, like I said. We're also gonna choose them we're gonna ask them to choose their top three skills here from this curated checklist. That's gonna help narrow their focus. It's gonna make the data more useful later on. It's also gonna tell you a lot about how they see themselves. I once saw a candidate, apply for a support engineering role, and their, like, top three skills were Figma, video editing, and creativity, which is, like, super valid, but maybe not the right fit for that particular job. And I was thinking about that when I was hearing what what Peter was talking about earlier about how, you know, that ability to work with, AI and data cloud to really help them find the correct role would have been awesome there. I just wanted to call that out because I remembered it as we were listening in. But those kinds of things are hard to catch in a resume, but a little more obvious when you start to structure the input or talk to, you know, an agent. But instead of asking those open ended questions like tell us about your strengths, I'm actually going to be capturing some self ratings on things like working independently, meeting technical requirements of the position, understanding the industry. That's gonna give us a little bit more of a grounded place to start, a little easier time spotting any mismatches later on. Of course, you know, how did they find out about the role? One of the most important fields in a form. If you know, you know. So after all of this is submitted, that's going to land right here in what we call an approval step. And I'm gonna name this approval step. This is our recruiter, Hannah. She's going to open this application, and she's going to see those, like, clean structured inputs. She can obviously check out the resume as well. But with this, she doesn't have to waste time double checking, like, any formatting or decoding any of the vague answers. She can review it quickly, you know, make a confident decision to approve or reject, and that will allow it to either send a rejection email or move into the approval flow. But in this version, it's a really simple review step, obviously, but this could easily be a chance to refine that data further. You could sync it to an ATS, kick off a different flow entirely. We've got webhook connectors because clean data unlocks all this flexibility. I think that's something that you've seen with a lot of what we're talking about. The cleaner and better your data, the more you can do with it. Once you trust what you've collected, you can really hand it off and enrich it, you know, route it with confidence. So even though we've just got Hannah reviewing it here, you can picture the step as like a launchpad for whatever is gonna happen next. So once the applicant has been approved, the first thing that this workflow here is going to do is actually it's gonna, like, quietly create a little snapshot of the original submission. We've got an advanced document generation step. That's a PDF that I formatted with with some nice little fields to collect the data. That way, if I update or refine things later on, I'll have a clean record of what was submitted originally, but it's not going to affect the fields in the actual Salesforce record. That's that will come into play right here when we get into these three connectors here. So this is a set of connector logic. I could have definitely packed all these into one connector step, but I broke them out here to talk about three really different ways that you can interact with your data depending on what you're trying to achieve. So step one is actually this one right here, confirming whether or not they're a previous applicant. If they have, we're going to be updating their existing record. And like I said, this isn't a step by step walk through. If you want one of those, you can definitely reach out to us. This is more about the concept here. So I've got a parameter here indicating that, you know, we're gonna update the applicant record, and it's going to do a lookup query. It's gonna look in Salesforce for that record. And if one is located, it's actually going to check a little checkbox I have on my record indicating that they've applied previously. Now why I care about this is because I've got, you know, the small checkbox on my Salesforce record, but it's so helpful. It's you know, if they wanna see patterns in repeat interest, track how many people are applying more than once before getting hired, or be able to ask my ask my agent what kinds of, you know, people are applying multiple times. I need to know who has applied more than once, and duplicates are gonna be a little harder to deal with than a single record with a checkbox, at least for me in my org. But obviously, everyone's work is different. So let's talk about this one, which is if no record is found, we're going to create a brand new applicant record, which is probably a little more familiar. For everybody, creating a record, Everyone does it. It's fantastic. I did add this precondition, so it was that if the previous rec if, you know, if the record already existed, it would not create a new one. Again, we're gonna avoid those duplicates. But when I map all this information into my Salesforce record, I'm just sending it on over and I'm collecting it with intention. And my intention is to, to make sure this data is clean and my intention really comes into play here with step three. So this connector is a prefill connector. It's going to prefill the following hiring manager review form with the applicant's data from Salesforce. When I say intention, this is the intentional choice. I actually could have pulled the information from up here when they submitted the application and just brought it down here using our form to form prefill. I didn't do that. This is because I made an intentional choice to think about this as something very real world, which is that Hannah might have updated that record after a recruiter call. Another form or integration could have added more detail. That record could have changed in the meantime for when they first submitted the form to when this got into the hands of the hiring manager. And so if Salesforce is my one source of truth in this scenario, I wanna pull the information from Salesforce again and make sure that this manager, when they're reviewing the data, they deserve to have that most current version of it. So, you know, if they're gonna use it to make a decision, I want that in front of them. So that's why I've made that very intentional decision to prefill the information from Salesforce instead of pulling it from the top. So let's talk about this for another second, the hiring manager review form. So we've actually technically seen a similar form a moment ago. It may look very familiar. This is, this is a form that I imagined would be opened, say, during an interview by the manager themselves, the hiring manager. This is the exact same form that the applicant filled out with one key change. And this is actually, like I said, prefilled from Salesforce. So the manager is seeing real time prefilled view of their record, and this is where we're gonna shift from reviewing to refining. So during the conversation, they may realize that the years of relevant experience should maybe be adjusted higher or lower based on how they explain past roles. Maybe industry knowledge needs, you know, to be a little more adjusted. So the manager will actually be able to change these details here and click to update the Salesforce. And in doing so, we've given them the power to effectively change the Salesforce to be more accurate, because that zoomed out there. Because if during the interview, they notice these things, that's important. If it's just in their own individual notes but not in the record, that can cause a lot of problems. This course correction means that, you know, your resume, your applicant record is now, if it changes, it's not able to be used to the fullest degree. In in fact and I noticed that I don't have a ton of time, so I'll mention this really quick, which is let's say the recruiter, Hannah, has seen something in this person and so she's moved them forward despite the fact that maybe, you know, their experience level doesn't quite match what they had put in their resume. That original record, if she didn't change it, stayed in Salesforce. And that could actually lead to something interesting where perhaps that slightly overconfident self assessment would be routed by automatic workflows in maybe in Salesforce, handoffs between teams, system filters, and suddenly the the candidate could be routed to a senior level role that they're not ready for, and that the conversation's off, expectations are misaligned, the hiring manager is frustrated, and the opportunity for this person, that right role, got lost in the miscommunication. Plus, if you're tracking hiring metrics, that's just not a not a great idea. So in this moment, the manager isn't just deciding who to hire. They're kind of shaping the accuracy of the data. And I think that data validation by the person with the most context is just that important. So, of course, once the interview is complete, they have another approval where they get approved or denied. It sends out emails. Choosing that path, again, is very straightforward. But in real hiring, the story obviously doesn't stop there. This could lead to second round interviews, follow-up questions, deeper candidate assessments, but that's because workflows like this aren't just for decisions. They're for conversations. This view is just kind of a snapshot. It's a moment in time, and what you do with the data after this is where it really gets interesting. Because in my personal experience, like, a good process makes the data clean and a really great one makes the data honest. So that's why I wanted to show you this flow. It's simple on the surface, but it's built around that kind of thoughtful data stewardship that we really care about at FormAssembly. And, you know, that's been the through line of this whole session. Whether you're collecting it or powering decisions with your AI, clean data is what makes any of this actually work. So despite our three different perspectives, I think we're all circling the same truth of garbage in, garbage out. And, with that title drop, I will pass it over for q and a. She said the name of the session, so you know that means it's coming almost to an end. Thank you, Kayla. I I'd ask for you to stay on, stage for just a moment because we did get a question during your demo. Actually, two. Josh asks, I have a question building off of Kayla's point around having the workflow leverage Salesforce data instead of passing data from one form to another. If we have a use case where the form is sending data to both Salesforce and another system, and the other system needs to send data back to the form in order for the user to complete the form process, could FormAssembly maintain or store data behind the scenes? Oh, that's a fascinating one. So if it's within the same workflow, it does continue on, and there's some really incredible tools there. So if you have your form in workflow and it's set data to Salesforce, another system, etcetera, and, you know, let's say we're gonna hand this to another person and you use an assignment rule. In the middle there, you know, you may have sent the data off to another tool, but it still exists. The workflow is still able to use it. You've got all kinds of, like, workflow, workflow formatting tools and everything that really while it looks so clean, there's so much going on in the back, especially with formulas. And I could talk about this for literal hours. So instead, I will say that while they are still within the workflow and completing that one workflow response, the data can be reused in some pretty incredible ways. If you want to go beyond that, I would love to hear more about your use case. And if you could please reach out to us, I actually would love to talk more about your, like, overall use case there. But within the workflow response, absolutely, it it keeps hanging out there. Nice. Thank you. And and Diane actually had a question after your demo as well. And and audience, if you have questions for either Dixon or Peter, feel free to ask those as well. I'll bring them on the stage if you have questions for them. Diane asked, in this example, in your example, Kayla, would the hiring manager make the changes to the resume or to that application in FormAssembly or in Salesforce? It could be either. So okay. One of my favorite use cases hi, Kayla. I'm a nerd. So one of my favorite use cases is actually that in Salesforce, on your page layout, you can, set up a button and you can use merge fields to make it so the button is actually opening a prefilled form or workflow. So that means that every single record, whichever one you click that button on, it, like, opens the prefilled one. And I phrase it that way because we actually use that for, for feedback collection and stuff. It's really cool on support cases. But I bring that up because whether it's inside FormAssembly or in Salesforce, it can it can be done. You can continue the workflow. You can bring the information in and out. There's so many possibilities. So I'd say, definitely, again, I keep saying reach out to us, but I just love hearing people's use cases. So we can talk more specific some other time. Yeah. And thank you for that. You know, to your point, you know, with, especially with this new integration, FormAssembly is so tightly connected to Salesforce that, you know, if you're making these changes on the FormAssembly form, we're sending those over to Salesforce. It's almost the same thing. It doesn't look like the audience has more questions. So if that's the case and you've got about thirty more seconds, if you'd like to squeeze one in, I'd love to ask you a question. You know, I know some of you have more questions about how FormAssembly works. If you would like, I am happy to have one of our, colleagues reach out to you, offer a personalized demo if you'd like to see our new integration. Or as, Kayla mentioned, we also have a new document generation feature we call advanced document generation that makes it incredibly simple to generate personalized documents at scale. If you'd like to see a demo of that, go ahead and answer the new poll that I just launched. We'll have someone reach out to you in the next couple of days. If you're North America based, you know, Memorial Day may make scheduling difficult, but we will get in front of you, eventually. Fantastic. Kayla, thank you so much. I'm not seeing any, specific q and a come in. So before I let this audience go and, audience, thank you for coming today. Thank you for showing interest in this. We know that Agentforce AI are heavy topics, you know, like, that every organization is debating. We wanted to bring a little bit of content. So, again, thank you to all of our speakers, Dixon, Peter, Kayla. Thank you so much for your expertise, for the content that you brought. The more that we can learn, the more that we can be inspired by the use cases that you've seen for AI, I think the better our own organizations will are going to be when we, you know, bring AI to our own processes. If you're interested in learning more about FormAssembly, I wanted to tell you a bit more about our platform. You know, we've got one of the leading Salesforce Ben integrations in the market. We offer unlimited forms and submissions. You can prefill. You can live look up all of that with Salesforce. We've got some really powerful tools that we'd love to show you. And speaking and but but beyond the platform, we've also got some resources. So, if you click the link that Saima shared, you can access a report that we produced earlier this year. And as Kayla alluded to, the survey that we conducted at the end of the year, the data she presented at the beginning of this call, we just published a report with this. So we're able to send that out to you. After this session, look for an email from FormAssembly. We'll get that data into your hands so you can learn from it as well. And if you don't wanna wait for our email and you wanna get in touch with us, as Kayla mentioned, please go to formassembly.com. You can reach out to our team. You can book a personalized demo. You can get in front of us, in any way you'd like. I've got plenty of buttons for you to click on that website so you know where to find us. To my team members, thank you again. I really appreciate all of your work and your presentations. And audience, I wanna thank you again. If you'd like more information from FormAssembly, go over to formassembly.com, and we host our own webinars on there as well. You can find more from us there. Have a wonderful day, a wonderful weekend, and we'll talk to you later.