How @salesforce Tests Tableau First to Drive Internal Adoption

ahead and kick it off. Thank you all so much for joining us for this session on using Tableau Next as customer zero. Real quick obligatory forward-looking statement, y'all, we're going to be amazing, but don't buy stocks based on stuff we say. Buy stocks based on what's currently in the product. Um, and with that, let's go ahead and do some intros. So my name is Casey O'Donnell. I've been with Tableau/Salesforce for about 8 years now. And for the last year and change, I've been helping to run our Customer Zero Motion, which essentially means that it's my job to help make Salesforce as good of a Tableau customer as possible. And then to pass on what we learn to all of you. Sydney?

Hi everyone, my name's Sydney Dollar. I've been at Salesforce and Tableau for just over 3 years. I started as an analyst supporting sales strategy and over the past couple months have had the awesome opportunity to support Casey on the Customer Zero team.

All right, and with that, thank you all very much for joining us and let's hop into it. Now, like any good former English major, I never start something without defining my terms. So let's define what is Customer Zero. Some of you might have heard of drinking your own champagne. Maybe you've heard of eating your own dog food. Or maybe just really simply, you know about one of Tableau's values, which is that we use our own product. Whatever metaphor calls to you, what Customer Zero means is that we are early and enthusiastic adopters of all of our newest features and products. We find bugs so you don't have to, and we come up with really cool and creative ways to use our products so that we're always pushing it forward. Something that's really important to Sidney and I, we both started as Tableau analysts, is always bringing the voice of the analyst and the Tableau end user into all of our product conversations. So how do we think about what success looks like for our team? And Of course, the year is 2026. Maybe you've heard about AI. Maybe you've heard about agents. We can't just be successful analytics. We have to be successful as an agentic analytics enterprise. So first off, we want good dog food, right? We don't want to make like, I don't know, gross slop. I'm about to stretch this metaphor really far, guys, but we want our dog food to be good. Then we want a lot of dog food. Salesforce is an 83,000-person company.

We got—

Now open. Make your way to Hall F. The Data, Trust, and Humanity with Vera Tudej Thurston session is about to begin.

Okay. We have a lot of— we got to make a lot of food. And then finally, we want to make sure that people actually want to eat our dog food. I'm sure everybody here has probably at some point in time made a really, really cool dashboard, launched it, and then had approximately nobody use it. That's not a good experience, and that's not how we build good products. So we want to make sure we don't do that. Okay, enough with the dog food. What does that actually mean? So that means that we want to build good products that scale and get adopted. And for us, that means starting early so that we drive early use of and feedback on our products, building for scale and using the available telemetry to proactively identify and fix any potential bottlenecks, and driving awareness. And because part of that means agentic analytics, and that's new, and getting your data AI ready can be hard, that involves enablement as well and helping our analysts get up to speed on the newest capabilities. And so to tell us how we actually do all of this, I'm going to hand it over to Sydney.

So I can't— in a presentation about how we use our own products as customer zero go without mentioning our latest analytics innovation, which is Tableau Next. If you don't know what Tableau Next is, I'm gonna spend a couple minutes walking through you through this beautiful layer cake slide. But Tableau Next goes beyond dashboards to bring you insights directly with Agentic Analytics. So looking at this slide, you have the data layer. That's where all of your data orchestration takes place. Then you have the semantic layer. That's where AI really comes into play. This is where AI will add trusted business knowledge to your data. The visualization layer, that's providing you with all the visualizations that you know and love from Tableau. The action layer, that's gonna be where you can take action directly at the point of insight with things like Salesforce Flows, or MuleSoft. And then you have the cherry on top. That's Tableau Agent. And this is bringing conversational analytics directly to your data. We also have Tableau Marketplace, which allows you to share and reuse assets both internally and externally. And then Apps for Salesforce. So that's prebuilt insights for each of Salesforce Cloud. And all of this is built directly on the Salesforce platform. So with this slide with a bunch of different layers, you're probably thinking, how the heck did we build Agentic Analytics at Salesforce? This looks pretty complex. That's not necessarily the case. This is where the concept of analytics interoperability really comes into play. We didn't have to completely start over to use Agentic Analytics. We are able to build on a robust portfolio that we currently had using Tableau Cloud and CRM Analytics. We use Tableau Next to augment existing experiences. And you'll hear us talk about this a lot in our presentation. We're not just exclusively using Tableau Next for specific use cases or Tableau Cloud. Really, we want these products to harmonize together and not be in conflict with each other. So what we found is there's not necessarily a one-size-fits-all approach to analytics. So, back to how we drive success as a Customer Zero team. Casey talked about how it's so important to start early, but as you can see on this slide, sometimes it's a little bit easier said than done. So, I'm going to walk you through a couple of the friction points that we experienced as we got our Customer Zero program off the ground. The first one being, Salesforce is a huge org.

Good morning.

The keynote room is now open. Make your way to Hall F. The Data, Trust, and Humanity with Veritune Day Thurston session is about to begin.

Salesforce is a huge internal org. It can be hard for product managers to know exactly where to go and which users to target to get good feedback on their new features or usability. We solve for that by creating targeted feedback councils and pilot programs to be able to connect our PMs directly with the users to be able to give feedback. As customer zero, we're early adopters. We're the beta before the beta. Testing is time-consuming, and we're not always the only ones who are doing testing. We have volunteers who are helping to get our testing and features off the ground. So, we wanted to make this as easy as possible, and some of that I'm going to talk about here in a couple slides, but we streamlined our testing through Slackbot's skills. And then lastly, this one important piece is users won't give feedback if they think it's a black box.

If you—

users don't know where feedback is going, how it's being actioned on, or what even happens after it's submitted, they won't submit feedback. Feedback transparency is so key to establishing a good Customer Zero program. So, we established new feedback workflows directly in the flow of work in Slack. And use that tracking in Tableau Next, which I'll show you what that looks like here. So on this slide, you can see our centralized feedback view and dashboard for both internal users, product managers, and product leaders. In this dashboard, product managers can see all the feedback that's been submitted for their team, point of contacts, and who actually submitted the feedback. That allows product managers to quickly follow up with who submitted the feedback in a timely manner. Internal users can also use this dashboard to see where their feedback currently sits. Is it being prioritized? Where is it on— where is it at on the roadmap? And then lastly, product leaders. We have established SLAs for when and how long feedback should be sitting in a review process. Product leaders can hold their teams accountable through this dashboard and also get real-time metrics in Slack to see how many people on their team have feedback that's currently sitting in the review process. All of this to say, when we launched this process internally, we went from 5 feedback submissions internally for Tableau Next to over 80+ feedback submissions. I also talked about how we can make testing easier. You know, a lot of these people, this isn't their full-time day job. I could give everyone participation stickers and say, "Hey, thanks for helping out. Would love to have you come back." Maybe not the best tactic though. So that's how we use Slackbot skills to really automate agentic testing and accuracy results. Previously, our users had to go directly into Slack and test each agent question one by one, and then manually log results in a spreadsheet. We automated this process directly through Slackbot, through Slackbot skills. This was essentially a prompt we wrote that Slackbot can then run in the background, and we do not have to do any sort of manual testing on our end. So users can go in, quickly see what Slackbot spit out, check it for accuracy, and their testing's done for the day. So, you've implemented feedback, You think you have a great product. How do you go big and scale? Again, there's not usually a one-size-fits-all approach to scale. What we found is some of our friction points is you have to have streamlined processes. We didn't have a great single point of truth contact for how to submit bugs internally. We streamlined that bug submission process through Slack, where users could then go and track how many bugs they've submitted, and product managers could see how many bugs were submitted for their specific features. Agentic analytics is also a new skill. You'll hear a lot about how people think traditional analytics is going to be replaced by AI. We're very prescriptive about our messaging with our internal teams that agentic analytics is not a skill that's going to replace traditional BI. Instead, it's a mindset shift. So we really talk to our users about how can you think about how to build data that is great for making visualizations, but also great for driving agent success. It's just a slight mindset shift that we've tried to bridge the gap with our internal users through training and enablement. And then again, we really try to complement our existing dashboards and not push everybody to using one tool for analytics. And then this last point, there isn't a blueprint for scale for every single customer. Everyone's going to be different with their expertise and their usage. So how do we solve for that? Telemetry. Telemetry allows us to see how our users are using our products, what dashboards they're using, what data sources they're using, how often they're accessing resources. But we're also looking at things like performance load times. This allows us to make sure we're ensuring our dashboards are performant before we go bigger and scale to more users. Here's an example of what some of those internal dashboards look like. This purple monstrosity of a dashboard is our Tableau Cloud adoption dashboard that has key metrics like quarterly active users, monthly active users, and more. But again, we're not just tracking adoption for one product. We're also tracking adoption for Tableau Next, which we use metrics for. And then we're also tracking adoption for things like CRM Analytics. And coming soon, you can actually expect Admin Insights for Tableau Next. So similar experience to what you're used to in Tableau Cloud, but here you can see what users are using what dashboards, average load time, and more.. And one thing I will say that's been great about the admin, admin insights with Tableau Cloud is we've been able to take the data that exists and also layer in Salesforce internal user data so we can see exactly how many sales users are adopting Tableau and other user insights. All right, so you've scaled, you probably think you've done Great job. Customer Zero Program is a success. Boom.

Mic drop.

Not necessarily. That's just one piece of the puzzle. We also have to drive awareness. So as I like to say it, it's winning hearts, minds, and the I'll just do it in a G Suite crowd. Everyone knows that person that has, you know, 50 VLOOKUPs or IF statements in a G Sheet. We're not about to try and make everyone adopt a single analytics tool. But here's some of the points that we try to get across with driving awareness. You can have a great product, and that doesn't mean people are gonna shout to the rooftops, "Hey, this is a great product." So as Customer Zero, we really seek to champion our wins and prove value. So we're always, you know, tooting our own horn and sharing internal product success. You're also gonna have people that say, "We've always done it this way." And it's hard to break into people's legacy workflows. But our approach is, again, we don't necessarily want to replace what you're doing. We want to augment what you're doing with the power of Tableau and Tableau Next. Then analytics on an island. Can't expect to have a good Tableau community in a vacuum. If ideas are shared and nobody's listening, it's not going to be an effective program. We broke down those silos internally at Salesforce by establishing an internal forum where users can share wins, best practices, lesson learned, and just generally talk about our product. We also started with sending out monthly comms for Tableau Next. You know, our product is constantly evolving, but we want people to be using it internally. So through this monthly communication, we highlight what new features are upcoming. And when they could expect them, and then also pilot programs. So this is just a little bit about how we do Customer Zero at Salesforce, but I'm gonna let Casey take us home and tell you how you can be your own Customer Zero.

Thanks, Sydney. So this slide is all about how you can be like Customer Zero-ish, because of course we're Customer Zero, but we want that for you too. We want you to find value with Tableau Next and drive that value for your users. So let's start the way that Sidney did with starting early. First off, if you have not heard of the Tableau Pre-Release Program, take out your phone, Google it right now, look it up, look into it. Getting hands-on with beta versions of our product is such a powerful way to get ahead of what we're going to be releasing and be prepared for it. Already have your use cases and your data lined up. Centralized feedback and not just making it easy to give feedback, But making your feedback really visible, putting something like that dashboard where people can go and see, I gave this feedback, has it been reviewed yet? When is it being delivered in the product? If not, like, why is it not being? Do that with your analytics projects as well, because people want to know that their voices are being heard and that you're taking their feedback in. And of course, being accountable to responding to that feedback. Building for scale. Y'all, Data Cloud was like a brand new product for me when we started, and I did not know how to set up governance or security on Data Cloud, and I leaned really heavily on our IT team to help us figure that out. In a second, I'm gonna point you to a session where they talk about how they set that up. I highly recommend checking it out because building well from the beginning is going to be really critical to being able to scale over time. Also, Tableau gives you so much telemetry. Use it. Track your dashboard load time, track your extract performance, reach out to teams that are struggling with slower products because that's going to help you proactively give them a better experience. And make it easy for people to share when they're having problems. Finally, driving awareness. Really what this is, is about creating communities within your own organization. Highlighting your wins. Building communities so that people can get inspired. I get all of my best ideas from other people, right? From seeing the really great things that they do in Tableau and then following that and building it on my own. And finally, my sort of personal catchphrase for Agentic Analytics has been augment, don't replace, right? Nobody wants to hear like, go break the thing that's already working and build it somewhere new that you've never built it. So that's why when we think about our internal use cases, we've really focused on how do you take that same data but do more with it? How do you take that Tableau Cloud Publish data source and bring it to your users? All right, so this is like a lot of advice, but if you came here because you wanna talk about actual use cases, then I have good news for you. A, we're not going anywhere, so feel free to pop over and ask us if you wanna learn more details about like, hey, how did you actually build that dashboard? What are you actually trying to launch? Or if you don't want to talk to us anymore, no worries, not insulted. Check out the Data Village, right? I mean, we're in it, but the session right over there, we have an entire Salesforce on Tableau zone where you can learn more about how our finance, sales, marketing, engineering teams, et cetera, are using our products and see the actual dashboards that they're using. Also, I mentioned that our IT team is going to be giving a session in like an hour. I highly recommend checking that out if you have questions about how you actually get set up on Tableau Next in a scalable way. With that, feedback, we love it. It's a gift. And we want to make it easy for you to give us feedback because that's how we make sure that TC is full of content that's really valuable to you. So please scan the QR code, give us the feedback, give feedback on every session that you attend this week because it helps us continue to make TC a really valuable experience. And with that, thank you very much, everybody. Thank you for joining us.