Unlocking the ROI of AI workflows

Great to see everyone and great to be here to talk about a topic that I suspect we all care about, unlocking the ROI of AI workflows.

For those of you that I haven't had the pleasure of meeting, my name is Greg Yukley and I lead the Asana's business value consulting team. If you haven't worked with myself or one of my team members, what we do is we partner with Asana's top customers like my pal Jeff here on, on aligning on what value looks like in their Asana investment. So that could be, before you make a significant Asana deployment, aligning on what potential future ROI you should expect to see, or if you're an exist... Long time existing customer like Jeff, we align on what value you're realizing the ROI you're seeing from your existing Asana deployment. So instead of me talking theoretically today about ROI, what better way than having a real life customer talk about So we're delighted to have Jeff Webb with us today. And Jeff, would you like to say hi to everybody? Yeah.

Greg, hey, thank you so much for having me here, man.

Very excited to show all the stuff that we have going on right now.

Part of what I do is I work in the Medicaid segment, right, in healthcare and oversee strategic project intake. You know, we've been using Asana for a couple of years now, I gotta say it has made a tremendous difference into how we operate overall. And so, yeah, excited to really dig in and show what we've built.

You ready Sounds to dive good.

Yeah, absolutely. Before we hear Jeff and his team's amazing story, I thought we'd just briefly set the table a little bit about, I think, what we're all facing. So according to Asana's own research department, 89% of knowledge workers are using AI at work, about 75% of those on a weekly basis. The interesting thing is it's quite disjointed. There's a lot of different ways of using AIs, different systems of using it, What we're finding, when enterprises are trying to deploy AI across their workflows, 74% of those organizations report being stuck in what they call pilot mode or what I like to kind of call pilot purgatory. There's a lot of excitement, potential fear around AI.

As we're going talk a little bit about later on, AI requires a lot of human change management as well, so we think that in order to get off to a fast ROI for deploying AI, it's really important to have a strategy in place for that execution.

So that's a little bit of what Jeff and I will talk to you a little bit about here today. We're just gonna hit on briefly about a framework for AI prioritization that we've seen Jeff and his organization deploy over their past couple years as they deployed Asana AI. Then Jeff's gonna talk about how they first deployed Asana AI Studio in their PMO organization for their intake process. And now, then we'll we'll start a little talking about how he and his team are currently deploying AI teammates to really build on that success of the initial AI, use case. And then finally, we'll leave you with some benchmarks today of what all of our customer base is seeing as they deploy Asana AI to give you a little bit of a baseline to to use as you build your own business cases.

Okay. So we'll touch base now quickly on the on the on the framework for for how we think about AI. This is a relatively simple two by two matrix. I remember, yeah, probably a year...

About a year and a half ago, Jeff got his team together in Arizona, and, you know, they fielded a lot of great potential use cases as they started to think about deploying AI for the first time. And I think it's important to think about, you know, you can't do everything at once. This does require significant human organizational change management. So it's important to weigh the effort versus the benefits of the potential AI use cases that you're looking at.

And I think as you'll hear from from Jeff's story today, when you're when you're starting out, when you're building trust, when you're trying to build momentum around AI get to get some of those early ROI wins, it's important to kind of focus in on some quick wins, things where it's, you know, still some significant effort and and and, benefits that will be realized, but things that are attainable so that you get quick momentum. Then when you get that sort of validation and that momentum in your organization, you can really swing for the fences and go for the game changers. So I think Jeff's story today will really help bring that to life.

Okay. So, Jeff, you alluded to this a little bit, but before we kinda get into the AI piece of things, tell us a little bit about what's going on in your business and what some of the executive priorities are in your organization, as you looked to deploy AI for the first?

Yeah. Absolutely. So as I was saying before, right, I work in the Medicaid segment. And Medicaid at the most simple basic level, what it's here for is to support the sickest of the sick and the poorest of the poor.

Right? And, you know, that's one of the reasons why I'm always so excited and motivated to be able to work in this space. Right? Because ultimately, it comes down to being able to help people.

But, you know, coming out of the pandemic, we're in a a situation right now, right, where our enrollment numbers are going back down.

That's putting a tremendous amount of pressure on us. Right? You know, just because we are losing membership does not mean the complexity of being able to run this business has decreased overall. Right?

We have, you know, a large number of members with very specific health care needs. And, you know, on top of that, we also operate in a very, very highly regulated space. You know, we have to deal with CMS at the federal level, and then, of course, state regulators, right, in each and every single one of the markets that we operate in. And they each have their own fun and unique personalities too.

Right? So it's a lot to be able to deal with and manage.

Yes. I have the pleasure of working with a lot of different types of enterprises. And with all due respect to everybody, I think health care and financial services have have it the hardest on that last point. So, yeah, a lot of, significant challenges to to keep in mind as you were, beginning your AI journey.

And then, you know, I alluded to the fact that, you know, as you, were beginning to, your AI, deployments, you got everybody together in your team. You really sort of, fielded a lot of different potential ideas. And I think you caught your initial, use case for the strategic project intake fit nicely into the sort of quick win space. But I...

Before we kinda get into that use case, can you tell us a little bit about how you kinda considered some of the other use cases and landed on strategic project intake as a good landing point to begin your AI journey? Absolutely. Yeah.

So, I mean, like like you're saying, right, a lot of this comes down to, right, where can we get the most bang for our buck.

Right? Apply, like, a small amount of effort to get those big wins and really start to get momentum.

We really keyed in and focused on strategic intake for a few different reasons. Well, first one is that, well, I already oversee it, so that made it... That helps. Yeah. But there are also some real valid business reasons why we chose to focus on intake as part of that first step as well.

Because if you think about it, right, intake is the front door to our entire strategic project portfolio. Right? Everything has to come in here in order to get scoped out, funded, prioritized, and ultimately delivered. And so if there are like issues or gaps in any of our requests, you know, we're not properly articulating what the business value is.

Right? Like the timing, the cost, any of those things, not only can that create like delays and bottlenecks in the intake process itself, but it can hurt the overall confidence in the funding requests that we're looking to bring forward. So that was obviously a big reason. But on top of that too, you know, I have a a very smart talented team that supports me here at intake.

And they're having to do stuff like, you know, going in and updating custom fields. Right? Copying information back and forth between different documents. Is that the best, most efficient use of their time?

Right. Probably not.

And then another big advantage of Well, right, and where we've really started to bake in AI to a lot of our our intake workflows is that the benefit there is it allows us to really start taking advantage about what these large language models like do so well, which is to helping us like like synthesize and summarize a lot of just like really complex detailed information in a way that's digestible that we can understand it and then be able to turn and then explain what that is to our executive sponsors overall. So, know, combine that with the fact that it's a standard process. Right? There there's gonna be the same steps we go through, you know, most of the time.

You know, those are things that we can track. We can put metrics around. We can start to measure, like, the timing, the the velocity, all those types of things and show over time, right, how we're getting better, how we're getting smarter with those things. Not only from a pure intake process perspective, but then be able to show and display how that benefits our strategic portfolio as a whole.

Absolutely. Yeah. I think a lot of great things to hit on there. So one, you started it in a part of the business that you have direct control in, but, yeah, we often find with a lot of our customers deploying in an area where you have champions, folks that you have been more innovative thinking and willing to to take it on.

Yeah. I think you identified a lot of the, low hanging fruit sort of cost savings. You have really talented people that, yeah, are doing a lot of ad...

Admin work that, aren't necessarily getting them jumping out of bed in the morning and and what isn't quite their calling, by allowing them to sort of free up their time to do more high valued work. And then your sort of portfolio management point, you know, these are...

Having spent a good amount of time with you and your organization, these are highly valuable, you know, strategic initiatives in your organization. So, you know, getting that that first level of cost savings, but even being able to move the needle on a high...

An eyelash on the portfolio level would be massive.

So would you mind saying a little bit more about, you know, the intake process sort of end to end before you deployed AI Studio? What did that what did that kinda look like, and what what were some of the challenges?

Yeah. You know? So before we had fully started to leverage, like, AI Studio, you know, we had a couple of like a big challenges. I mean, first and foremost, right, oftentimes there could be inconsistent or incomplete like requests coming into us.

I mean, Greg, I can't tell you how many times like I looked at a request coming in and I can't make heads or tails about what they're actually asking for.

Right?

Yeah. And of course, that's assuming they even come to our front door in the first place. Right? Yes. Try to end around the process and try to find funding from some kind of different avenue.

I've I've I've I've never experienced that in any company I've ever worked with.

Yes.

But you know, on on top of that too, right, I think we were being inconsistent with how we were looking to prioritize this work overall. You know, I'm sure that you have never experienced this in your career ever. But, you know, a lot of times the way it worked is, like, the work that would get bumped to the top of the stack. Right?

It just came down to who would yell and scream the loudest.

Right?

And is that always going to be the best, most consistent way to do things? Right. Maybe not. And then I think the third big item I I I wanna call out is that, you know, we are really suffering from a lack of portfolio intelligence overall.

Right? Like, what is the impact of all this demand coming into us? Right? What does that mean for our capacity?

How does that impact the work that's already, like, in flight and going on downstream? Right. Right? We didn't have that intelligence or visibility before, but now we're we're starting to get there.

Maybe because I'm a business value business case person, but it's it's your portfolio intelligence examples that I've seen you develop over the years where... And going back to your executive challenges, you you you mentioned at the beginning of this. There are, you know, sort of headwinds at the moment, and whether it's reducing costs or being able to to get that value realization benefit earlier in these significant projects, that's, that's just massive. Now I I believe you very kindly brought some examples to to show us today of of how Asana AI is now helping you throughout the, you know, sort three parts of this process.

You you wanna walk us through that?

Let's dive in the workflow itself. Right? So folks, we're gonna start right at the beginning of our process. Right?

We have a very customer centric approach. Right? We wanna make it easy for people to be able to come to us. So what this means is we have a very simple intake form.

Right? It's just a couple of questions. Like, we're not expecting you to come and know everything about your project. You don't have have a whole business case ready for us.

We're gonna help you build that as we go through the process itself. But Greg, at a most simple level, all you need to do, just tell me, like, what is your project? Why is it important? And then we can take it from there.

So make it very easy. And then this is where now we're gonna get into like how this stuff works within Asana itself. So you submit the form, what's gonna happen is boom, that's gonna turn around and that's gonna become a triage task that we use. And this is where the AI rules and the workflow start to come into play.

Right? It's gonna take that first pass for us. It's gonna do the triage. It's gonna look to classify that based upon, you know, what portfolios we think could be impacted overall.

Right? Help us fill out like the 1,500 different custom fields that we use for all of our reporting purposes. Yeah. It will even like go in and help to be able to give us an initial prioritization scoring based upon the information that's been provided to us.

Right?

That way, you know, we can, you know, be able to to to rack and stack, you know, just the work that's coming that day or or that week so that we can focus on the most important things first. Right. But that's just like the first pass of it. Right?

I'm sure you guys hear this over and over again, but like a really important thing is like we gotta keep the human in the loop here as well. Right? So my teammates...

We have a daily huddle. Right? A stand up where we come in, we review the task, we look at what the AI did, we see if we agree with it or not. And then immediately after that, there's going to be a a fast follow for my intake leads.

Right? What I call a pre discovery call, where we're sit down with the business requester. Right? To truly make sure that we understand their request.

Right? And give the proper level of prioritization.

You know, the point I wanna be able to emphasize, right, is that the the the workflow and the AI is here to help support the administrative work, right, so that the people can can focus on the meeting.

Right. Yeah. Not not focusing on that admin work or the chasing work or the interpretation work that you were describing before, but really helping you help your stakeholders all that all that much faster.

Yeah.

So...

Oh, it'll continue too as we're kinda going through the process itself. Right? Yeah. So harping on this point about the meaning and keeping the human in the loop too.

Right? What it's like...

What are workflows helping us do? It's going in and it's helping us prep for our discovery calls. Right? Help us give like a list of questions that we can prepare to really make sure that we understand the work.

You know, like, lot of us, right, especially, like, in my role, you know, I'm not a subject matter expert by by any means and and neither are my team. Right? But we have to understand enough to be able to speak to it. Right?

So what Asana AI is doing is helping us to be able to prep for those calls. Right? Prepare. Help us find gaps in our project charter overall so that we can have efficient and effective meetings ultimately.

Right.

You know, and so the rest of the process, again, there's a lot of steps that we go through, right, to look to, you know, refine what the business case actually looks like. Right?

Get to the really tangible, like, cost and benefit estimations overall. But the key point that I wanna be able to emphasize is that, like, the cool thing is we can track all this work in Asana. Right? It's all in one place.

It's all part of one end to end workflow overall. That initial, like, form that gets submitted becomes that triage ticket. Right? That triage ticket turns around and we automatically convert it through our rules to a project record.

Right? We use that record. It has a template. It not only tracks what we have to do from an intake perspective, but that we're able to use all the way through, right, and turn around and give and hand over to our project managers when it comes time to delivery.

So it's like a seamless...

It's a closed loop system overall. All the work is there in one place so that nothing gets dropped or gets missed.

I mean, that's so huge, not just for for you and your industry, but we see it...

I see it among our customer base. We see it at Asana Yeah. Where, you know, have, you know, a presale team handing off to a post sales team or in the SaaS business. And anytime there's handoff, that creates friction, disruption, opportunity for things to get lost, and you haven't seen your process up close and personal. Yep. Yeah. It's a... It's it's really remarkable how you're able to make... Manage these strategic projects bay bay... Basically from cradle to... Not not grave, but success.

Success. Success. Yes.

I like that. Absolutely. And now I know, we've had the pleasure of of, kind of aligning with you all on, your your metrics, your ROI that you... That you've seen. We've we've kind of we've kind of anonymized it a little bit here, for for you today, but I think it kinda gets at the general thrust of of what you've seen. But this was about, six months after your deployment, and these were some of the results that that you saw. So wanna talk a little bit about... So you were able to handle three times the the output growth, essentially the three times the volume that...

Yeah. Yeah. I mean, it's really been growing exponentially with the amount of work that we've been able to intake and process. I mean, I think the first year before we took this over and had everything fully automated with AI Studio, maybe we're able to handle something like a 120 requests.

Right? That went up to 250, you know, last year. Right? And now I think this year we're probably on track to hit over, like, 320.

Right? So we're really...

Like, we're we're focusing on, like, working smarter. Right? Being more efficient and being more timely overall. Right?

In this business, right, we have to be able to move and react fast. Right? And and every second counts, and and that's what these workflow automations are able to help us do.

That's awesome. And I I really loved your story about backfill said you you guys, the team, continues to grow and develop and continue on their career path at your organization.

Yeah, that you were able as... Even with this volume growing as well, you were able to, still execute on the... Without filling the backfills.

Yeah. Yeah. I mean, so, like, last couple of years, like, we have grown a lot as a project management office. Right?

I like to think that's because we've been successful overall in delivering. You But, know, so we probably have at least right now maybe about 60 people or so on our team. And just what that means is, right, there's gonna be, like, churn and employee turnover. Right?

People are gonna find other, like, really good opportunities for them. That's fantastic. But, you know, one of the benefits is, like, through a lot of, like, the rules that we've put in place, not only through, like, intake, but how we manage our risk issues and change a lot of those things. Right?

We are operating more efficiently. We have better data overall, and we haven't had to backfill those wrecks. You know? And what that means is we've been able to give those wrecks back to, like, the other parts of the business too to help them.

Right? Like, again, we have very tight margins here. Right? We need every little gap that we can get.

And so to be able to help out, like, our business partners too, right, and be able to hand them...

Those in...

The...

Those those wrecks over, it it just helps to improve the business overall.

That's awesome. And then, of course, my my... Again, my personal favorite, your portfolio intelligence. And, yeah, I guess I would kinda spit the insights that you're able specifically to, raise up to your executive leadership about these important projects coming out of your PM organization?

Yeah. Yeah. That's a key point too. I mean, off top of my head, and this is approximate, we have 478 active records overall floating around in our our our portfolio between intake and delivery.

Right? Right. That's a lot to review. How do you stay on top of that? Right?

This is where, again, the the the workflows, the AI is so valuable to us. Right? We can really key in and find where are we having those trends? Right?

Looking for those commonalities. Right? Like risk and issue identification, helping us to identify you know, small problems before they could become big problems. Right?

That's what I think of when I think of portfolio intelligence.

Exactly. Be... Not just kind of executing, to the project to to completion, but also just highlighting quickly where there's problems or issues that that are coming up for sure. Now what what... What's a bit about three or four months since we started the AI team journey for you as well? So, yeah, just to recap, the

other... What we covered so far is when Jeff and his team started with AI Studio. Now they're pivoting, or enhancing into AI teammates as well.

And, you know, why don't you tell us a little bit now as as you're looking to continue your AI progress?

What what sort of metrics or, you know, targets are you setting for for AI teammates that you I mean, this is where we're seeing, like, a lot of, like, very promising and very exciting stuff overall.

Right? And we got a large list of, like, teammates and agents that we're looking to build right now. Like, thinking about, like, again, operational savings. Like, we're calling this, like, our stoplight agent overall.

Right? But red, yellow, and green, what work should we stop and continue? Again, I'm gonna keep emphasizing it, but Medicaid, it's highly regulated. It means that a lot of times we get these compliance requests coming in.

They have like a one week or two week turnaround time or it's, you know, God forbid, do that same day overall to be able right to the top. Exactly. These big complex changes. But like...

And and everybody else here is gonna experience the same things. Right? There's not unlimited time, budget, or resources. Right?

And we can't do everything at once. So in that situation, right, does work have to come off the plate and what work should that be?

Right.

And if we really wanna be smart about that, right, we have to factor in so much stuff, right, that the fight... The current financials, the spend, right, the value. We're getting a lot of these things. Right? We're gonna build teammates to be able to help with that, right, and help us make the call and the decision, review all those projects and say like, what is the most important work? Right? What is gonna truly drive the most value for the people that we serve?

Absolutely.

So a couple of other things on here too. Right? Like faster delivery time, even risk and compliance. Right?

Right. Again, it's a very large portfolio. Right? We're looking to be able to build agents to find, like, common trends and commonalities overall.

Right? You know, if we get a unique intake request coming in for one particular market, you know, there's a good chance that a few months from now, we might be able to see that from another one of our regulators. Right? Right.

Like, this level of portfolio intelligence and what teammates is helping us get to will help us to identify that faster, help us be more responsive, and then hopefully help us have a a better product offering too.

Yeah. Yeah. When you're when you're playing a 50 state chess chess game as well as with the federal government. Yeah.

Any any sort of commonalities or or themes can get emerge, I'm sure, quite helpful. And so...

And I remember you you...

I remember your initial list of AI teammates that you were partnered with our AI specialists with, and I wanna say it was probably about 30 or 40 that you all Yep. Kinda brought We got a backlog. Yeah. Absolutely.

But a lot of really innovative ones that, you know, I certain...

I certainly learned a lot about our power of teammates from working with you all. But, you know, I think when you were kinda describing your journey with AI with us, yeah, you were saying with AI Studio, what we covered before that was rethinking and improving existing work that you were...

You and your team were already doing. And and the thing that really struck me was when you're looking at teammates and after the success you've had so far, now you were thinking more broadly, more ambitiously, and thinking about new ways of working. Yeah. And how how can you kind of almost re...

Rethink how you're doing things. So, you know, that's...

Yeah. I'll let you elaborate on that more more eloquently, but that...

I think that's kinda how you landed on how you prioritize some of your teammates that you you bumped to the top of the the the list. Right?

Yeah. Yeah. Just have a prioritization framework. Right? And even like a good tip and a good technique is...

And I use this too. Sit down like with with an AI chatbot. Right? Have it like...

Explain like what your problems and issues are. Right? Have it help you come up with a prioritization framework. Be able to to rank it and rack it and stack it.

Right? Yep. And you probably heard the same thing if you were listening to Greg Miller talk on the Phoenix stage earlier. Right?

A lot of times, right, just be approximately right. It's about getting momentum and putting that that first foot forward. Right? And that's how you can start to drive and really make these changes.

Yeah. Well, any any Greg that is also a Philadelphia Eagles fan is a good person in my estimation.

So we're... We have built a a number of the teammates on on your list, and you wanna elaborate a little bit on your operational intake intelligence teammate and what... How would you sort of...

How would you assess the success of a teammate like that as you're rethinking your work a little bit?

I mean, ultimately, what it what comes down to is decision making. Right? Are we going after and funding, like, the right requests overall? There's so much work that that comes in at any given time.

Right? We have to be able to lock in. Have to be able to prioritize that. We have to be able to figure out, at the end of day, right, what is, like, like, discretionary versus nondiscretionary.

Right? We have, like true contractual compliance stuff.

But we have to be able to...

We also wanna focus on improving things and making things better overall. That portfolio intelligence is going to be able to help us come up with a framework. Right? So at least that way we have a a common like language and a common starting point for how we evaluate these records overall.

Right? And what you're seeing right here, it all comes down to, like, the speed and the timing. Right? We're not having to review each and every single request.

Right?

Our AI studio workflows. Right? What we're starting to with teammates is helping us to unpack these things more, make it more simple and streamlined so that we can articulate it and understand it. What you're seeing here on the board as well is really helping us to to prep for for those discovery meetings too and build those foundations so that we can get smarter each and every single time.

Right. And, again, we're talking, was it, 400 some odd Yes. Sort of projects across. So when we think about the ROI, you know, moving...

Just being a bit more intelligent on three, four, five with some of these really meaty projects, you know Yeah. There's a there's a a massive ROI there for sure. So wonderful. Well, Jeff, thank you so much.

This was a a wonderful overview of of your AI...

Multiyear AI journey, and thank you for being such a terrific partner. We're excited to see where you push teammates even further. And, yeah, I've I've been kinda lobbying Asana that we should take some of your teammates to make them prebuilt out of the box for all of our customers because they're so terrific. So we'll we'll just leave you with this.

So we we heard some great examples in in ROI from Jeff today. What does Asana's research lab tell us about some of the success that Asana customers like Jeff are seeing across the board as they deploy AI on their workflows? So we're seeing a significant movement early on in three buckets. So average task completion time, and that was dropping from about two point five days to about about a half a day, so 78% reduction in task completion time.

Probably some of the things like like Jeff was saying about in the intake process of, you know, having to kinda go back and forth with folks submitting requests, things of that nature. Across our customer base greater capacity, so in terms of the number of overall tasks completed per day, we're seeing about 27% greater capacity based on on our customer...

Top customer feedback. So not quite as good as what Jeff and and his team have seen, but, you know, still significant and and and a major ROI. And then, you know, again, I think similar to what Jeff was phrasing as portfolio intelligence, greater predictability. So when you're working on some of these in...

Significant projects, you know, really being able to move things through your your pipeline significantly, a significant reduction, about 28% reduction, and the number of times that due dates removed or pushed back. So really being able to kinda move the chains to ensure that your major initiatives aren't hitting those those bottlenecks.

So, hopefully, you know, keep these benchmarks in mind. Yeah. Reach out to your Sonic contact if you need some help putting a business case together yourself as you as you look at at AI, and and thank you very much for joining us today. Hopefully, we'll see you out in out in the demo lobby.

Thank you, folks. Appreciate it.