Insight ON How to Evaluate AI Partners When Everyone Claims They Can Deliver

Audio transcript:

How to Evaluate AI Partners When Everyone Claims They Can Deliver

Ferhan Zaki (00:02):

Every single one of their customers is deploying AI and production currently right now, every single one of your customers. And if we're not having a seat at that table, we're missing an opportunity to be relevant to those customers for years to come.

Jillian Viner (00:17):

That's a pretty direct challenge and honestly, it's one we're sitting with for a minute. Welcome back to InsightOn. I'm Jillian Viner and my guest today is Faron Zaki. He's the senior vice president of sales for the Google Cloud solution line at Insight. Ferran is on the road every single week or on virtual calls talking to clients across North America, which means he has kind of a real-time market intelligence that's hard to get anywhere else. He knows what organizations are actually asking, where they're getting stuck, and what separates the companies from making real progress with AI from the ones that are still stuck in that strategy mode. So if you're wondering whether your peers are as behind as you feel or maybe further ahead than you'd like, Ron has a pretty clear answer. All right, let's go. Well, Ferhan, thank you for joining me on the podcast.

(01:07):

It's lovely to talk to you. I understand that you are on the road quite a bit, yes?

Ferhan (01:11):

Yes, yes, absolutely. Every week on the road over the last months.

Jillian (01:14):

What's your role at Insight?

Ferhan (01:16):

My role at Insight is managing our sales teams on the Google Cloud solution line, which was the acquisition, the Side acquisition from about two years ago.

Jillian (01:25):

And our clients should care about this conversation today because I think a lot of leaders kind of want to know what's going on next door. They want to know what their peers are doing. Are their struggles the same as their competitor struggles? You talk to a lot of different clients. Absolutely. So you've got some really good intel on what's going on. Yeah. So when you're on the road talking to clients, give me some of that intel. What are some of the common themes that you're hearing from clients today?

Ferhan (01:54):

Yeah, absolutely. And I think it's really interesting to see because there's organizations that are born in the cloud and they're kind of used to this new Agentic world that we're living in. But organizations that have been around here for 30 or 50 or hundreds of years are needing the most help. They have the most amount of technical debt. They've got multiple platforms if you have multiple acquisitions. And every day they wake up, they open up the newspaper and they're seeing their competitors deploy AI into real production workloads and they're trying to figure out how to do the same thing.

Jillian (02:30):

So the conversation is still very much AI, no matter where you are in terms of your infrastructure, your devices. It is about how do we get ready for AI? How do we enable our teams and our company with AI?

Ferhan (02:41):

Absolutely. The conversation is, how do I take AI and apply it to my most complex problem that the business is trying to solve? And the problem has to sit in something that's going to help me drive revenue for the business and it's going to be a clear differentiator in the market because that's what the stock shareholders care about.

Jillian (03:01):

Yeah. They're feeling behind?

Ferhan (03:04):

They're feeling behind. And every time you ask them, "How quickly do you want to deploy this? " They always the same conversation months ago.

Jillian (03:12):

Months ago.

Ferhan (03:13):

Months ago. When I deployed this months ago, we're ready. How quickly can we get started? And where I see most organizations struggle with is they're talking to customers and they're looking at presentations and they're trying to make a strategy decision. That strategy does not have to be fully vetted out. It has to be how quickly we can build a proof of concept and can we take that proof of concept and roll it into production? We'd rather fail fast than to be behind.

Jillian (03:42):

That's interesting. We had a conversation with one of our insight leaders not too long ago, Reen Gideon, who gave kind of similar advice where the idea that you can plan your business strategy three months, six months, a year down the road is almost impossible these days because things are changing so fast. So if you are saying that you're recognizing even client conversations that they have to adjust to how they're thinking about timeframes, but they want things fast, what does that actually look like in practice then? How are they calculating fast wins?

Ferhan (04:18):

The way they're calculating time right now is they're looking at their competitors and they're seeing how quickly they are improving revenue, bringing products to market faster and creating that holistic experience that their customers are asking for. That's how they're measuring time. And so for them, it's an opportunity to take this capability, this latest Agentic capability and get rolling with that and not be left behind.

Jillian (04:44):

Let's illustrate that a little bit more because I think that's a goal of every company, obviously. Absolutely. Listen to the customers, deliver on the customers. So Agentic AI in general, I think is really introducing and opening people's eyes to brand new ways of working, brand new ways of serving customers. And probably for a lot of organizations, this is wildly different and new and it's hard to even know where to get started. So let's start with an organization. Let's start with financial services, right?That's pretty traditional, right?

Ferhan (05:17):

Absolutely.

Jillian (05:18):

Where are they seeing AI breakthroughs or innovation? How is this industry being impacted by AI?

Ferhan (05:24):

Yeah, that's a really good example. And financial services, again, it's an industry that's been around here for a very, very long time, has gone through acquisition of, you could say, multiple different types of banks, could be regionally, it could be nationally and they're dealing with challenges of driving engagement from net new customers. Their goal is how do I bring the next generation of wealth management or portfolio of products that is going to help me retain my customers versus these Banborn and the cloud customers or companies that are really app driven experiences or interfaces that the current generation is dealing with. And so what we're seeing traditional banking do is take agentic capabilities and use that as a way in to self-service customers as well as small businesses. And their goal is to offer up a suite of products or a value of differentiated offerings they could then use to lower in those customers and provide a better experience and use that to differentiate against some of their competition almost like our Robin Hill type experience at their fingertips.

Jillian (06:31):

Oh, but this is more traditional banks you' talking about

Ferhan (06:34):

Though, competing

Jillian (06:35):

With kind of the new kid on the block that's cloud origin story.

Ferhan (06:39):

But again, you're dealing with a regulated industry. Yes.

Jillian (06:43):

It's

Ferhan (06:43):

Complicated. When you're dealing with financial services, you can't just let these agents run wild. In a world where anybody has the ability to build agents, could be autonomous agents, they then have the ability to have access to a lot of confidential data. That's exactly where customers are needing help. They want to be able to have a strategy in place to govern, secure those agents and deployment in a manner where IT has control over that and has visibility over that.

Jillian (07:13):

So how are they navigating that today? When they come to you with these questions, are they asking how are other organizations in financial services doing it? Just specifically where are they looking for help? They want to know what people are doing, but when you meet with them to have these conversations, what are the questions that are top of mind for them? Where are they getting stuck?

Ferhan (07:34):

Yeah. The questions that are top of mind for them is, number one, how can I build this safely and securely where I have visibility and control over this deployment? How quickly and how fast can we go? And how quickly can we have measurable KPIs along the way that we can measure the success of this project? Those are probably the top three questions that are coming up.

Jillian (07:56):

Okay. So that's the financial services industry. What about, I don't know, manufacturing space? What does that look like? Well,

Ferhan (08:05):

What we're seeing with companies, especially manufacturing companies, is they want real-time visibility, real-time insights into products that they're releasing out into the market. So if you went back just even a few months ago or a few years ago, an organization would launch a new product and they would release this product out to the consumer base, but in order for them to get any sort of real-time sentiment analysis around that product, it would take them months, if not even years in some cases to fix a product. They're not getting that real-time visibility. They're not getting that real-time feedback directly from the consumer. And now with the capabilities that we're applying via AI to these businesses, we're able to get real-time sentiment analysis about that product regardless of where it exists on the internet. So we're scraping a lot of these different data sites to provide real-time visibility to these businesses.

(08:59):

Now they have a product that they've launched out into the market and we can make those changes right on the factory line in days versus months or years and that is something you can quantify. Those are hard light savings that business can say we pivoted a product and look at the uptake in revenue in those products. So we have customers that are doing that exactly that today. It's more around consistency, it's more around speed, it's more around governance, giving them almost like an AI sandbox to play with. And then what we don't want to do is prohibit growth or stock growth. We want everybody to go out there and develop agents and have those capabilities, but we want to ground them, right? Yeah. We want to ground them in the data that they have access to. Want them to be able to govern them but not limit them is essentially the goal for that.

(09:49):

But yeah, because again, organizations are going to run wild with all the different developer platforms and capabilities, LLMs that exist out there. And so that creates sort of this agentic sprawl that we're starting to see already.

Jillian (10:03):

Let me ask you a question and tee that up. When we talk about bringing AI into an organization, there's a couple different layers. So you actually present that and start to see the outcomes. You see it at your individual level, your back office workers, your finance team, marketing team, individuals creating agents or doing prompts for their own business purposes.

(10:24):

Then you have it at the larger scale use case, which is what you just described, where you can actually get real-time customer feedback, make changes to the factory floor. I mean, that's really the dream I think a lot of organizations are looking for. That's the transformation that's promised. Then you kind of have this other third layer, which is more not necessarily in a product, but it's in the product development. So this is your engineers, your technical teammates. So what does that look like for organizations today? What questions do they have there? Because obviously they have some technical talent, like why do they need a partner to help them?

Ferhan (10:58):

Yeah. And I think that look, developers know how to build. They're traditional folks that traditional code writers, they road code, they know how to do that. Now getting them to use these agentic platforms where they're just giving the prompts and telling them what to do is almost a completely different mindset. I always use the example of sitting in a Waymo for the first time and letting an autonomous vehicle drive yourself and you're like, "Oh my gosh, am I really trusting this with my life?"

Jillian (11:25):

Yes, I've done that. It is a little scary the first time and then you get used to it very quickly. It's strange.

Ferhan (11:30):

Exactly. And so it's almost driving that comfort for the developer to let the AI help with coding and then trust that output that it generates almost like with the human in the loop to make sure that that code makes sense. And what organizations are struggling with right now is a sprawl on the developer sides in terms of the tools, the productivity, the capabilities that they have to offer and they're looking for standardization in those works and they're looking for governance in those environments. So once you do that, once you kind of lay out that foundational work, which is exactly what we can help them weigh, they slow down and they tend to go much faster because now we have essentially the sandbox for them to play with.

Jillian (12:12):

Yeah. So again, you talked to customers that are kind of like all over the map in terms of their AI journey. What have you observed that really sets apart the ones who are making great headway on this and others that are feeling a bit stuck?

Ferhan (12:26):

Yeah, that's a great question. And I had to see this in action to believe it because everybody always almost like talk the talk, but walk the walk. And what we're seeing is organizations where IT and the business are working together towards one outcome. IT needs business to help remove unblock challenges in real time. That could be in the support in the manner from like a financial standpoint or a resource standpoint and the technology needs that commitment from the business in order to be able to focus on this capability as their main project. So the number one thing is to ensure that the business is aligned. And then organizations that are really moving ahead are organizations that are looking at the capabilities and picking a partner and picking a platform that they want to build on such as Google Cloud and a partner such as Insight that has the capabilities.

(13:23):

Can I do an insight plugin here?

(13:27):

Okay.This is exactly what we're primed to because we have an organization in Inspire 11 that helps tie the business challenges to the technology capabilities and really wear that consultative hat on and help customers identify those problems and quantify them. So that's number one. And then you have the technology teams with insight being able to actually deliver this into production in real time for our customers and customers are not interested in pilot purgatory. They want to see these demos in real time. So if you're walking into a customer meeting and you're meeting with your client, know that your competitor has already been there with a demo and they're already making a decision in terms of what they want to deploy next in production. And so we're walking in and we're building demos in days, not weeks, hours, not days in some cases and we are showcasing that to them on that next call and then taking that strategy and rolling that into production.

(14:22):

That is how fast organizations doing and that's how fast we're moving and Insight's positioned well to do this.

Jillian (14:28):

As a client, I'm evaluating my partners because I'm tired of hearing the hype.What should I expect from my partner at this point?

Ferhan (14:37):

Yeah. What you want to be doing in your evaluation with partners is can they walk the walk and talk the talk? So you want to ensure that whatever they're showing you, they're able to demo and build that capability live for you. That trust has to be shown in real time because you might not get that second meeting with that Kusky. And from a customer standpoint, I want to be able to see what you're telling me that you can do. Secondly, if you're making a decision on a partner, you want to ensure that whichever platform you are building your business on top of, there's a trusted OEM relationship there and that OEM will always help you make a recommendation and help validate that and ensuring that they have the capabilities to deliver that. Third thing I would ensure is that if you're an organization that does work in financial services, manufacturing, retail, you want to ensure that your partners are bringing examples forward that are focused on your vertical, focused on your business and have an understanding of that.

(15:39):

And I think those three things, bring in all those things together will ensure that you have a good experience.

Jillian (15:45):

That makes sense because when it comes to AI, AI is not new, but generative AI is new. And so you're not going to have people that have 15 years of generative experience, but you could have a partner with 15 years of retail experience,

Ferhan (15:57):

Manufacturing

Jillian (15:58):

Experience. So the foundation is there. And then the other part that you're saying is really test that they're not just explaining, talking to you what they can do, but they're proving it out in the moment.

Ferhan (16:09):

Absolutely.

Jillian (16:09):

Show me the evidence.

Ferhan (16:11):

Yeah, absolutely. You want to be able to see what they're talking about and what they're delivering. You want to be able to see that in the form of a pilot. And there's no longer is it acceptable to just see it on a slide because you're going to be essentially investing in this platform and capability for the foreseeable future. And then from a vertical perspective, you're absolutely correct. Organizations have been around for a very long time operating these verticals for 10, 15, 20 years in some cases. So that expertise and that knowledge, that wealth of knowledge, those problems don't change. Those problems don't go away. They're still valid in this agentic world that we're living in today.

Jillian (16:49):

Yeah. What's a question that you love to get from clients?

Ferhan (16:57):

The number one question that I got actually this week at Google Next, the number one question was we have agents sprawl, we have agents all over the place. How do I govern these agents?

Jillian (17:09):

Very common challenge right now.

Ferhan (17:10):

Common challenge. I have no idea who's building them and we don't want to stop. We're telling everybody what to do, but nobody's paying attention and we need help governing these agents. And so there's a lot of conversation around security. There's a lot of conversation around governance of the agents and data governance as just in general because all of these datas, all the different data is sitting across disparate systems across the organization and now you're giving an agent access to that. That is scary for an organization.

Jillian (17:40):

It's kind of ironic, isn't it? When you think back to like a year ago, I think the biggest problem was just getting people to use the tools and now we've opened Pandora's box.

Ferhan (17:49):

That's

Jillian (17:49):

Right. We have to claw it back a little bit.

Ferhan (17:52):

Yeah. It's going so fast and it's going to be incredible to see where it goes over the next few months and over the next few years, but we're essentially it started with AI then gen AI and then agentic capabilities. And now we're looking at autonomous agents, now we're looking at super agents that are going to be coming here pretty soon. And so it's exciting. It's a really exciting time to be part of this business. All

Jillian (18:17):

Rig. So what's a question that you wish more clients would ask?

Ferhan (18:21):

Yeah. I think that the questions that I want organization asking me more is honestly around the data side of this because everybody wants to run so fast on the agentic front. They're talking about governance, but they still have their data sitting in multiple platforms and that data is still not in a state for the agent to consume as fast as it can and as efficiently you can. And that is essentially going to give you the best output that you want for your business versus an agent hallucinating.

Jillian (18:56):

Thinking again then to organizations that want to move quickly, they're feeling behind, what would you say are the top two, maybe three reasons that organizations are not getting the AI outcomes that they want or at the speed that they want?

Ferhan (19:16):

Yeah. I think I will say first thing is everybody feels like they're consistently behind, but I would say some of the things that I see is cultural. The culture of the organization, which is we see this a lot in more traditional organizations where folks may be ... Gosh, I don't know if it sound cruel from saying this, but they're maybe a few years out from retirement or they've kind of been doing the same thing for years and years. They don't want to learn anything new. They don't want to do anything different and they definitely don't want to adopt a monumental shift for the organization, which is go heavy in AI, go heavy in data, which means that they have to invest in these platforms and capabilities. I think that's something that I consistently see. I think what organizations need to do instead is say, "Hey, I don't need to disrupt everything that I have.

(20:09):

I can just start fresh, create a completely different environment, dump a subset of my data, create some agents on top of that, start to challenge and optimize real world workflows that can impact your business and start there. You don't have to start by moving every single thing around. Let's just start small and get the organizational comfortable with it. That's going to drive change organically within the business.

Jillian (20:38):

I know that it's a hard thing to express, but I do feel that's a valuable thing to recognize. I mean, the change management is hard and knowing where you're going to face that resistance can help you overcome those things. Are there other areas of culture or business operations that leaders are maybe undervaluating or not investing enough in order to make this AI thing work? Because AI is a big investment. It's kind of a risk. And you've already committed to this, but is there an area that they're like, " I don't want to do anymore, but if you did, you would 10X the results of your AI investment?

Ferhan (21:20):

Absolutely. Most of the AI work we see, again, it starts from the technology side. We're seeing the developers kind of pushed the envelope and developing capabilities, building applications, using agentic workflows. But where AI is having the biggest differentiators for the business is really in cases like marketing, HR, legal, finance. These are the businesses, parts of an organization that have been doing business for years the same way. And if we just start to automate some of those workflows using AI, that would be way more than 10x returns for any sorts of organizations. So my advice to always to an organization is let's start from a top-down initiative that should be from a CEO on down and work right to left in your organization and sit down with every part of your business function and figure out where we can operationalize AI and help optimize those workflows to have that 10X impact on their business.

Jillian (22:21):

Are clients giving you a sense of where their biggest AI champions or successes are starting to come? Does it come from you for different business units or user groups?

Ferhan (22:32):

Yeah. So I think in organizations, what we're starting to see now more is business users speak up about AI capabilities and are starting to use it started to use it. Previously you would have just the technology teams and developers get access to these capabilities. Now everybody's almost building agents. So your next champion that is looking to bring forward this capability is coming from the business side of the house. And you have consistently even all the way up to the CEO trying to figure out these capabilities because they're opening up the newspaper, they're seeing their competitors dab in this new capability.

Jillian (23:14):

Yeah. All right. So I'm a business leader. I'm going to go ahead and invest in an AI solution. Maybe it's Google Gemini or something else. I give that to my team. Is that enough?

Ferhan (23:26):

It can't just be an organization buying licensees. We all know from past that that becomes shelfware. What you need to do is ensure that you have a proper strategy for adoption. And I think that often organizations overlook investments in key areas like change management and the goal of that is an organization that invests time and money, and in a lot of cases, their businesses on these agentic AI capabilities and they don't have a proper process for adoption. That's exactly where an organization that invests in change management really helps folks in the organization create agents, automate workflows and help them figure out how to use it and their job on a day-to-day basis versus an organization that invests in that capability and we see this time after time, six months or a year down the road, that becomes showfare. And at that point you have organizations that have spent time and sun costs and capabilities that they haven't received the value out of.

(24:31):

And so I would recommend every organization that is investing in this capability to really be incredibly thoughtful about how you deploy this capability to ensure you're driving the proper adoption. And those could be in the form of training, ensuring that your folks that are adopting this capability first, we bring those to the forefront and really make them champions in your organization to ensure that this capability is highlighted on a regular basis.

Jillian (24:57):

If you had to summarize what customers should be doing right now

Ferhan (25:01):

I think for every organization right now is you need to identify a real world problem impacting your organization, but start to apply AI to it. Let's figure out a way to build out that proof of concept and put some KPIs around it, measurable impact, and figure out a way to solve that business problem and roll that into production. And then you need to also look at a data governance strategy or in a platform that you want to build your business around and hit go. Let's just go. Just hey, go.

Jillian (25:31):

Veron, thank you so much for talking to me today. It was nice to hear what's going on in the market with clients and hopefully everyone's feeling a litle bit calmer now and not too far behind.

Ferhan (25:39):

Absolutely. Thanks for having me, Jillian.

Speaker 3 (25:42):

Thanks for listening to this episode of Insight On. If today's conversation sparked an idea or raised a challenge you're facing, head to insight.com. You'll find the resources, case studies, and real world solutions to help you lead with clarity. If you've found this episode to be helpful, be sure to follow InsightOn, leave a review and share it with a colleague. It's how we grow the conversation and help more leaders make better tech decisions. Discover more at insight.com. The views and opinions expressed in this podcast are of those of the host and the guests and do not necessarily reflect on the official policy or position of Insight or its affiliates. This content is for informational purposes only, should not be considered as professional or legal advice.