Learn how AI can unlock productivity, according to AI founder Andrew Louder

In a business landscape racing toward AI-driven transformations, the question isn’t whether to adopt AI, but how.

The rise of AI has ushered in a new wave of possibilities for businesses seeking to optimize operations and win a competitive edge. Now, organizations are presented with a vast array of tools and solutions, each promising transformative outcomes, and the challenge lies not only in picking the right tools but also in ensuring a seamless integration and adoption.

And that’s exactly Andrew Louder‘s area of expertise. Andrew is the founder and CEO of Louder Co., a consultancy that helps businesses develop AI strategies and processes to optimize operations and drive competitive advantage. He is also the host of the podcast Dallas Based Innovators and a board member and partner of Social Venture Partners. Andrew and his colleagues have been talking about AI solutions for the past six or seven years, but with recent advancements in generative AI, especially the launch of ChatGPT, everything changed. Suddenly, everyone wants in.

The numbers speak for themselves – researchers from Stanford University and MIT conducted a year-long study that showed that generative AI increased productivity of support agents by 14% on average, with a 35% jump among the newest and lowest-performing reps. Similarly, in their latest report, The Economic Potential of Generative AI, McKinsey estimates that generative AI could increase productivity in customer care by up to 45%. And Louder Co.’s customers are already feeling the impact – one client went from $10 million to close to $20 million in annual revenues without adding a single person to their staff.

The potential is immense. More than a tool, AI is a transformation waiting to unfold. But where exactly should you start? Which areas or pain points should you address first? Should you build your own custom solution or buy an off-the-shelf one? And if you buy, how can you make sure you’re choosing the right tool?

In today’s episode, we caught up with Andrew Louder to chat about selecting and implementing the right AI tools for your business.

Here are some of the key takeaways:

  • Deploying AI used to require costly custom builds, but recent advances brought accessible off-the-shelf solutions. Still, the decision to build vs. buy hinges on one’s requirements and processes.
  • Change management is often overlooked, but it’s crucial for ensuring the successful adoption of new systems. This involves awareness, communication, and training company-wide.
  • An AI readiness assessment can help identify pain points and quick-win opportunities for adoption while determining the optimal degree of AI implementation.
  • Minimize risks of off-the-shelf AI adoption by assessing requirements and comparing systems, prioritizing configurable tools, investing in change management, and writing an AI policy.
  • To ensure AI effectiveness, implement a data management plan and build a robust knowledge base with accurate data, policies, and documentation.

If you enjoy our discussion, check out more episodes of our podcast. You can follow on Apple Podcasts, Spotify, YouTube or grab the RSS feed in your player of choice. What follows is a lightly edited transcript of the episode.


Breaking free from the legal grind

Liam Geraghty: Hello, and welcome to Inside Intercom. I’m Liam Geraghty. Customer service is at the forefront of the AI revolution, and everyone’s trying to keep up. Recently we published the State of AI in Customer Service 2023 Report, which found that this moment is rife with opportunities. The sooner you adopt AI for your customer service strategy, the greater the chance to win a competitive edge. And that’s why today we’re talking about AI and how to use it to set your productivity free. And who better to talk to than Andrew Louder, founder and CEO of Louder Co., a company that specializes in simplifying the integration of AI into businesses. Andrew, you’re very welcome to the show.

Andrew Louder: Liam, thank you so much for having me. It’s an honor, and I’m excited to be a part of it.

“The days of AI feeling unattainable and intimidating should be over”

Liam: You have an excellent presentation about the power of AI, which kicks off with a super interesting story about this criminal defense law firm. I was wondering if you could share that story with us.

Andrew: Yeah, thank you, Liam. For context, I’ve been a part of this peer group called Vistage, and we have a number of different types of groups. There are CEO groups and emerging leader groups, and I’ve been traveling across the United States, delivering this presentation to a number of those different groups. And the main message I like to portray is that the days of AI feeling unattainable and intimidating should be over. There are so many AI tools that are quick to access, relatively low cost, and relatively low time to implement.

“They went from close to 1,000 work hours per case down to about 40 to 50 work hours because the AI is able to pour through all that information relatively quickly”

I start things off with a story of this criminal defense law firm client of ours. For those who understand that whole process, there’s a major bottleneck in the e-discovery phase. They’ve got to sort through two, three, or more terabytes worth of electronic data to try and find evidence to support their clients. And what that amounts to is, usually, a team of six or seven people pouring in 60, 70 hour work weeks of high stress, and they only comb through about 25% of all that information.

And we said, “Look, there are AI tools out there that can help you.” We went through an implementation process of a tool called Nextpoint and rolled it out. And what we found was that they went from close to 1,000 work hours per case down to about 40 to 50 work hours because the AI is able to pour through all that information relatively quickly. The key part of the whole equation is AI plus human interaction. That’s where the team of administrative folks can pour through all that information themselves, validate it, rerun it, feel confident about it, and move on.

Those are some massive gains for this company. Now they can go out, grow more revenue without needing to add more costs and payroll, and maybe even consider switching over to a flat fee structure for some of these administrative tasks so they can still remain profitable through all that. It’s incredibly powerful.

Becoming AI-ready

Liam: For leaders thinking about AI, and I suppose for a lot of other people – you might say most people are kind of new to all this AI stuff –, what are the key differences between building and buying AI solutions? How could businesses decide which approach is more suitable for their own needs?

“Yes, it’s become relatively simpler to build AI, but it still takes a lot of time, money, heartache, effort, and sleepless nights”

Andrew: If you and I were having this conversation five or eight years ago, the answer would be completely different. Back then, in order to get AI into your business, you more than likely had to go and build something custom, and doing that took a lot of blood, sweat, tears, money, and time. But there’s been this movement of democratization of AI where the power of these tools and workflows can get into the hands of many to create powerful applications.

Today, the analysis around build versus buy is driven by a couple of things, but it starts with awareness. You might have a vision or a problem that needs solving, and if you don’t know what’s out there, you may find yourself gravitating toward a decision to build. And yes, it’s become relatively simpler to build AI, but it still takes a lot of time, money, heartache, effort, and sleepless nights. It’s possible, and the cost has come down quite a bit, but the reason why you would consider that is that maybe there’s not a solution out there that can solve your problem, so you’re looking for something very unique.

Maybe you’re trying to establish a major competitive differentiator, whether in the line of a new product offering or in the way you operate internally. But those are reasons you’d probably want to buy if something doesn’t exist already and you’re looking to gain some new revenue. And obviously, when you buy something off the shelf, that’s something more commonly used, accessible, relatively quick to implement, relatively inexpensive, and you’re usually targeting a major operational efficiency gain too.

“It’s not necessarily whether or not you should be utilizing AI because the blanket answer is absolutely – it’s to what degree, what depth, or even at what speed of change management you should be moving”

Liam: Are there any steps my business needs to be taking to be AI-ready?

Andrew: That’s a question we hear all the time. We have an AI readiness assessment that you can take through our website, and what it gauges is not necessarily whether or not you should be utilizing AI because the blanket answer is absolutely – it’s to what degree, what depth, or even at what speed of change management you should be moving or feel comfortable moving toward. We grade things on strategy operations, AI buy-in, and even the people who are part of the company to get a feel for how ready your folks might be. Are you thinking strategically with AI in mind? Do you have an AI strategy? A lot of companies end up scoring rather low on that strategy piece because they’re just now starting to think about AI.

Another opportunity I get asked quite a bit is to come and present AI to leadership teams, which we do in the form of a workshop. Then, we get into what we call our rapid assessment, which is a great starting point to identify the problems in the business that are major pain points and tedious manual tasks that need to be addressed to map those problems to potential AI solutions. We identify the quick win opportunities and get into a roadmap of selecting and implementing the right tools for them. A lot of times, it starts with the readiness assessment, building awareness in the company, finding those low-hanging fruit of AI opportunities, and implementing those to start gaining great outputs and outcomes and start getting some confidence in the business around AI.

“Change management is often overlooked, and it’s a shame because you spend all this time setting things up only to allow CEOs to ditch the effort and say, ‘Let’s just send a link out to our folks’”

Liam: So, if they buy the off-the-shelf AI solution, what do you think are the most crucial factors they should consider to ensure immediate benefits and a high return on investment?

Andrew: I think it starts with an inward look. If a buddy of yours asks you, “Hey, I’m car shopping. What car should I get?” I can suggest a car, but it may not suit you, right? Do you have a family? Do you want to go fast? Do you need to tow a lot of things? So you need to determine what’s going to be right for you and your business. You need to understand your processes and capture the requirements you need. You also need visibility into any security or compliance regulatory things your business or tool needs to abide by. And then, on top of that, what integration points do you need to be aware of? What systems and other applications would you want to integrate with? Once you have that list, it goes into what’s out there in the marketplace that can meet as many requirements as possible and choosing the right one – not just based on those needs, but also the cost analysis. Is the juice worth the squeeze? And then you get into implementation, which is all about configuring the tool, training it, and testing it up.

Change management is often overlooked, and it’s a shame because you spend all this time setting things up only to allow CEOs to ditch the effort and say, “Hey, let’s just send a link out to our folks. We believe they’ll do it.” They need awareness, communication, and training. Some are going to need some handholding. And that’s going to really drive adoption up. We help our clients work through that whole process. And one thing I love highlighting is that we help our clients do the piloting of the system so they get good feedback from those using it. And those people using it actually become positive influencers of the change in the company.

A productivity booster

Liam: Many businesses are going to be eager to tap into AI for increased profits. Just how attainable is that, and are there companies doing it already?

Andrew: There certainly are. In my talks, I actually highlight Fin, the product by Intercom. And somebody raised their hand and said, “Look, when we first started implementing Fin, we had a team of five people providing customer service support online. We rolled Fin out, and now, those five people are providing the coverage and support equivalent to 20 people.” That’s huge, right? That’s a huge ROI. So, for that particular business that had its sights on growing, now they don’t need to make that expense on the payroll. The tool is able to adapt and become that exponential factor to the employees using it.

“MIT put out a study that showed a 40% increase in productivity gains. McKinsey did a similar study that showed 30% that’s going to be attained, I believe, by 2030”

Liam: That’s great to hear. But time and cost are things I hear people mention a lot when it comes to this. They’re obviously essential metrics for businesses. What would you say to folks trying to develop strategies for evaluating the time and cost factors when making the decision to buy AI solutions?

Andrew: Look, a lot of the data’s been coming out, and I’ll give you three numbers to consider, okay? MIT put out a study that showed a 40% increase in productivity gains. McKinsey did a similar study that showed 30% that’s going to be attained, I believe, by 2030. Deloitte put one out even before all the generative AI craze that said that AI and machine learning are going to drive a 37% productivity gain amongst business users.

“We have a client that’s gone from $10 million to close to $20 million in annual revenues and didn’t need to add a single person to their staff”

You can certainly run your own numbers, go case by case, and try to estimate, “Okay, if I implement this tool, it might create 40% productivity gain, shave off 40% of hours; we’re paying people X amount of dollars per hour…” If you can get to a number around that and do a brief calculation of time saved versus the dollar spent, it’s a huge ROI. And I’m also seeing the other part of the equation – the growth you can now attain without needing to increase payroll. We have a client that’s gone from $10 million to close to $20 million in annual revenues and didn’t need to add a single person to their staff, which would’ve consisted of eight to 10 more people. That’s a massive profitability gain with that growth. It’s a mathematical equation you need to run for your business.

Liam: I am not sure if it’s the same study, but we spoke to a researcher from MIT, Lindsey Raymond, a few episodes ago, who was one of the authors of a study that talked about AI boosting customer service team productivity by this massive amount.

Andrew: Same one, yeah. Huge number in customer service.

Mitigating the risks

Liam: AI is this new thing, and people have all these different thoughts and concerns. De-risking AI implementation is a significant concern for a lot of people. How can businesses ensure they choose the right off-the-shelf AI tools and minimize the potential risks associated with that integration?

Andrew: First and foremost, I would recommend going through our four-phase approach. There are four phases: requirements gathering, selecting, implementation, and change management. And we’ve designed that to de-risk the whole process and ensure the right tool is selected. But even more importantly, make sure the business users are adopting it.

“The reason you go through those steps is to select the tool that will require as minimal customization as possible. Customization means time, money, and effort”

It starts with requirements gathering. There are some things you can do. Some folks enter into a demo situation on a product and come in with an open canvas, but we don’t recommend that. Send out a script of what exactly you want to see, maybe even share some of your data and information to have a real live business scenario put in play with the system. If it’s sensitive information, have a mutual NDA signed to continue that conversation. But once you see it in motion and you’ve done a good comparison across three, four, five, or so systems, you should have a high level of confidence in moving forward.

The reason you go through those steps is to select the tool that will require as minimal customization as possible. Customization means time, money, and effort. While there are certainly happy stories out there, it certainly increases that risk and likelihood of more costs and more time to implement. Finding tools that are highly configurable helps.

The other part of this is it doesn’t take long to sit down and write an AI policy for your business, and that AI policy can do so much for you. It can answer questions that your people might be having, it can ensure that sensitive information’s not going through certain channels. Because as you know, even ChatGPT has warnings on sensitive data. People need to be mindful of that. Sitting down and crafting a policy doesn’t need to be massive. It could take just 30 minutes. How should you use AI with customers? How should we drive accountability in AI? What are the use cases we should be using it for? Maybe even identifying approved tools. But those two things, taking the step-by-step approach of selecting the right tools and having an AI policy document can certainly help de-risk and drive success up in the business.

“Garbage in, garbage out is a huge component of AI. Ensure you’ve got the right documentation, the right policies and information that your customers can lean on”

Liam: That’s great advice. What insights do you have into the ways businesses can swiftly and affordably implement AI solutions without compromising on quality and effectiveness?

Andrew: Let’s pick on what I would assume to be the process there at Intercom in terms of what I’ve heard about Fin. It has a really great feature where you can turn on your access to your knowledge base, and then people can ask it questions just like they would with ChatGPT. My estimation would be to ensure that in that knowledge base resides the right data, your sources of truth of information, the right policy manuals, customer manuals, whatever that is. Garbage in, garbage out is a huge component of AI. Ensure you’ve got the right documentation, the right policies and information that your customers can lean on. And when they get that information, it’s correct. So, it’s going to be critical to take the time to do that.

Now, in the world of generative AI, that’s going to be the case across a number of different business use cases where you need to be sure, now more than ever, that you have a plan on how you’re managing your data. You want to have the opportunity to still live in a bit of a sandbox environment where you’re creating drafts and working toward a final document. If you’re anything like me, I sometimes have version 17 of something out there, but I don’t want generative AI going through that. So, having that data management plan where you have your sandbox and your production level environment, if you will, where you’re housing all those sources of truth data, is going to be a winning strategy to ensure that it’s working appropriately, you’re getting the right ROI, and it’s not creating any confusion amongst the users.

Liam: You’re obviously out there talking to many different businesses about this. What industries or business functions have particularly benefited from readily available AI solutions?

“Your competitors who are not looking at tools like that might sit back and say, ‘Gosh, how did we lose this? There’s no way they’re making any money.’ But yes, they are”

Andrew: I mean, it’s going to sound canned, but all of them, frankly. But I do think that some more than others. In a lot of blue-collar industries in bid situations to win business from others, whether it’s construction or something similar where you’ve got an RFP process, we’re driving toward the ability to leverage AI to increase profitability and margins. If you’re able to leverage tools that allow you to do that, especially in those hyper-competitive environments, you can be more competitive on your pricing to win more of those bids while still remaining highly profitable. Whereas your competitors who are not looking at tools like that might sit back and say, “Gosh, how did we lose this? There’s no way they’re making any money.” But yes, they are. They’re leveraging AI to drive efficiency gains.

I’ve seen, in the world of construction, estimating and takeoff AI software out there. I’ve seen project management software. There’s a tool called OpenSpace where you put a camera on your head, walk around, and it captures the progress of the construction. I do believe construction has a lot of low-hanging fruit. Law firms are also very much set in their old-school ways – lots of paper, lots of typing. Between generative AI and other AI management systems, I think they’re ripe for this as well.

Riding the next wave

Liam: Before we wrap up, and because we’re in the customer support space, I always like to ask, what was your best or worst customer service experience?

“Profits and processes are put ahead of customer experience”

Andrew: Oh, man. Are you asking from a personal standpoint, Liam?

Liam: Yeah, absolutely.

Andrew: I wear contacts. I love calling 1-800-CONTACTS and ordering from them because they’re so fast to answer. They have my information in front of them. They know how all that works. There’s also a really great company here in the Dallas-Fort Worth area called Moxie Pest Control. They provide outstanding customer service to their customers that, as a customer, I’m able to see firsthand.

Worst customer experiences, gosh. I don’t know if I want to put anybody on blast here, but it’s way too common, unfortunately. Profits and processes are put ahead of customer experience, which is another reason why I love Intercom. You’re able to merge both processes and experience. From the business side of it, it’s so easy to use and set up. From the customer side, it delivers such wonderful responses. When I see a company’s website and see Intercom down at the bottom right and need to use that thing, I’m feeling pretty good about what I’m going to experience.

Liam: And what’s next for you and the company? Are there any big plans or projects for the rest of the year?

“Right now, a lot of the conversations I’m having are at the AI 101 level. What I’m preparing for in the next six to nine months, maybe even sooner, is the next level”

Andrew: Oh, absolutely. We’re just trying to keep up with demand right now. We’ve been talking about AI solutions as part of our business offerings for the last six or seven years. It’s been a bit of an uphill climb because it’s been a more conceptual conversation. But with ChatGPT, what’s happening in generative AI, some of these tools coming out around Copilot, and what Google’s doing, it’s becoming a lot more concrete. There’s the wow factor behind it. The productivity numbers speak for themselves. And so, for us, it’s continuing to get the message out, drive the value out, and be ready to support our clients through it.

We’re also looking at the next wave of this. Right now, a lot of the conversations I’m having are at the AI 101 level. What I’m preparing for in the next six to nine months, maybe even sooner, is the next level. What are the big ones we can go after? What are the more impactful things as well? Those are the things we’re gearing up toward, and just trying to stay at the forefront so we can serve our clients in the best way we can.

Liam: Where can people keep in touch with you? There are probably going to be a lot of people who want to hear this presentation. Where can people find you?

Andrew: All right, our website’s louderco.com. You can also find us on Instagram and LinkedIn. And if you want to reach out to me via email, it’s Andrew.Louder@louderco.com. I’m happy to chat.

Liam: Andrew, thank you so much for joining me today.

Andrew: Hey, my pleasure. Thank you so much, Liam.

Fin AI Copilot CTA (Horizontal)