What Salesforce learned after saving 50,000 hours of work using AI

Salesforce recently announced that it has introduced more than 50 AI-powered tools among its workforce and reported that these tools have saved its employees in excess of 50,000 hours—or 24 years’ worth—of working time in just three months. 



As a company, Salesforce serves as an especially compelling case study for the impact of AI on work—not only because the company tests tools on their own workforce, but because so many others rely on Salesforce’s products to do their jobs each day. Simply put: Salesforce is in the business of work. 



Salesforce has more than 70,000 employees worldwide—a 30% increase since 2020. And the software giant builds the products that are used by employees at some 150,000 workplaces, from small businesses to Fortune 500 companies; from sales and customer service teams to marketing and tech teams. 



Salesforce has implemented its “Einstein” AI capabilities into many internal tools. For instance, Salesforce embedded Einstein into BaseCamp, a platform that Salesforce employees can use to ask company-related questions and get personalized answers. Salesforce also built Einstein into its employee onboarding processes so that new hires can ask questions about their benefits information and orientation checklists. Salesforce also added the AI-powered application into Slack, which can complete tasks like scheduling and summarizing meetings. 



Slack (which is owned by Salesforce) recently estimated that globally, general AI usage is up 23% since the beginning of the year and that 47% of office workers worldwide are enthusiastic about using AI in their daily work. 



However, most workers are also worried that they are going to lose their jobs to AI , according to research. And while these anxieties may prove to be valid in time, AI has not yet completely replaced large swaths of the workforce. Instead it has changed or replaced workers’ specific tasks. So it’s up for debate whether AI will chip away at workers’ tasks until it replaces people completely—or whether workers will find new tasks to complete.



Fast Company spoke with Salesforce’s chief people officer, Nathalie Scardino, about what happens when AI takes over tasks that workers used to do. (This interview has been edited for length and clarity.) 



I understand this is your first interview since taking on the role of chief people officer in January. But you’ve been at Salesforce for a long time. Could you walk me through your tenure at Salesforce and your career before? 



This is my 13th year at Salesforce. 



A long time ago, I started as a business major. And after that, what does a typical business major do? Obviously, they go to university to study film and theater arts—a natural connection. But the arts have always been a big passion of mine. And then after that, I had a long background in sales, selling HR software. Then, I moved into executive search. I got to do a lot of cool work around hiring C-suite executives into private equity-backed tech companies. That was my main focus before joining Salesforce as a recruiter, building out our Chicago office at the time. And now our footprint is so much bigger. I think there were 5,000 people when I joined. And now we’re at 73,000.



I focused on every part of the recruiting organization [at Salesforce]. Then a couple of years ago, I also took over learning and development and onboarding. Learning and development and continuous learning has always been a big area of focus for me. How people like to learn in this world that continues to change is a fascinating topic. So I took on all those components, and then I became the CPO officially in January. 



So I’m a mix of business and creative arts coming together, which, I think when we look at the skills for the next decade, those are skills that will be even more in demand. Skills like empathy, active listening, self-awareness—they become the hard skills versus the soft skills.



What was selling HR software like when that was your job—and how have things changed? 



A lot has changed, and also not a lot. I was essentially a sales rep selling into the technology, media, and telcom space. It was a software called Expert HR, and it was for smaller, midsize companies. It was about how do you set up your payroll? Where do you find good employee benefits, all in one place? That was the pitch: You don’t have to go to all these different sources and websites.  



We were selling the fact that instead of going to all these disparate systems and places and resources, your HR needs as a software service could be in one place. And we would connect you with HR experts, whether that was employment law or benefits protocol by country. 



And then here we are today, seeing how the benefits of AI for employees and businesses are massive. And a lot of those benefits involve bringing consolidated data sources and sets into the flow of work. I think that’s the through line. Now, you’re selling expert HR software as a single source in one place as you work. And I think that’s what AI has accelerated. 



What have been your biggest challenges since becoming CPO in January? 



We’re in a transformational time. It’s incredibly exciting, the pace of innovation and change, not just for our customers, but for our employees. And if you think about the nature of work, even over the last four years: coming out of COVID, into digital transformation, into generative and predictive AI. 



When you think about AI, a lot of the questions are around the trust and the ethical use of AI. And our employees require exactly the same level of standards. What I’m excited about is working very closely with our CIO, to deploy the ethical use of AI into the hands of our 73,000 employees. Those are the biggest changes.



Plus the rapid pace of how work continues to change in a hybrid world. The hybrid nature of work requires you to know the engagement of your organization. And I think we see higher engagement when employees are working with AI to finesse their work, to augment manual tasks so that they can focus on higher-value work activities. And we’re seeing that now with several use cases.



When it comes to the ethical use of AI, what do you think the stakes are? 



Trust is our number one value and without it, you don’t have a customer or employee base. So when we’re building products, at the very beginning of the product life cycle, we’re implementing ethical-use policies into how products get translated to the market. And I think that is a fundamental shift that we’re seeing across industries, rather than [AI ethics] being an afterthought. 



We’ve been in the AI space for over a decade. And so we actually have an office of ethical and humane use that sits directly with our product team to think about bias, how bias could show up, and how to eliminate it.



Salesforce says that team AI-powered tools have saved employees 50,000 hours of work. What specifically is that number capturing? 



We know that research has shown that 41% of workers are spending time on tasks that they equate as low value or repetitive. That is not new news. To help alleviate some of the mundane tasks, we introduced our AI-powered Slack-application Einstein. Now it assists employees with day-to-day tasks like scheduling and summarizing meetings in the flow of work. . . . The other thing that it does is it answers general questions. In one quarter, the application processed 370,000 queries. And that’s how we calculated the employee time savings of about 50,000 hours. So the appetite from our employees to leverage Einstein in their flow of work, which for us is Slack, has been significant. 



Another use case is our sales organization. They use Einstein to create summaries from conversations to help sales identify key insights for our customers, takeaways, and follow-up actions. And instead of going to disparate systems, you’re doing it in the flow of work. You’re not spending hours trying to find information anymore. It is consolidating those systems into one, from an intelligence and insights perspective, that we’ve never seen at this scale before. 



Why is measuring these kinds of efficiencies important? Would you say the main incentive for investing in AI is increased worker efficiency?  



I think it is about efficiency, but for us, we really listen to the voice of our employees. We are constantly listening to employee sentiment. And for us, it was addressing the question: How do I do less work for work’s sake, and [do fewer] administrational tasks? And how do I focus on more value-driven work? So a lot of [our investment] was in direct response to our employees’ feedback. 



Also, what I have experienced is our engagement increases when you have time to focus on less mundane tasks. Because now you’re engaged with creative value-driving for our customers, for our stakeholders—versus having to spend sometimes eight hours in a day, so much time, figuring out where to find a knowledge article, how to get your work done, who to talk to.



You mention employee sentiment, but across the labor market, employee sentiment is currently being measured at record-low rates. What have you seen in terms of employee sentiment at Salesforce? And what’s kind of moving the needle in a positive direction for you?



I’m thrilled to say we have a highly engaged workforce. We are globally distributed across the world. And when we look at the stats around engagement, it is really high. Then we unpack sentiment and asked: What would make your workload quicker, easier, and bring you more joy? And it came down to how do we deploy AI at scale in the flow of work? That has been the appetite from our employees. 



As a tech company, we’re focused on being the world’s number-one AI CRM and data company. Our employees leverage data every day. And we’ve deployed over 50 AI applications in partnership with our CIO. So sentiment and engagement is super high and it’s because, by design, we have focused on that since the beginning of Salesforce.



Typically, I hear that AI can replace more administrative work. But some of the use cases you bring up sound more like the work of middle managers. Do you have any thoughts about what tasks these new technologies might be replacing, and how it might impact the landscape of those jobs?



We’re AI- and human-centered. We’re seeing that [AI] is augmenting work.



I’ll give you an example. We have a team of recruitment coordinators. They are sort of like brand ambassadors for talent because it’s often the first person you might connect with. And we launched a tool where now the candidate can schedule an interview directly with the hiring manager—a task that often coordinators took. Now what we’re finding is coordinators are spending more time with the candidate meeting and greeting them, helping navigate them through the process, helping create a different level of experience. Because now the task of scheduling has been augmented through technology. But the value proposition of the coordinator is still incredibly important from an employee value proposition and an experience.  



And we’re in the business of experience. We’re in a highly competitive market still for the skills that we are looking for. And so, what I have witnessed is more of an augmentation of work. 



And actually, different roles are being created that were not around a year ago. Prompt engineering is one, ethical AI architects is another; and I’m sure that will be a trend we will continue to see. When the internet launched, social media marketing managers and influencers weren’t a thing. They weren’t jobs. I don’t think we’ve fully unlocked the new [AI-related] jobs yet. 



But what AI will do is change work. And I think that is a big unlock. When you think about the number-one skill for the next decade that’s required for highly engaged employees, it’s creative thinking. And I’m really excited about the level of creative thinking we’re about to enter even more so because of removing the manual administration tasks.