Would employees use AI differently if a practical project could earn them a 2 to 4 percent salary increase?
In this episode of Tech Talks Daily, I welcome back Rytis Lauris, CEO and co-founder of Omnisend. We last spoke in December 2022, before generative AI became part of almost every technology and workplace conversation. This time, we examine why so many company AI projects attract attention during a demonstration but never become part of the work people do each day.
Rytis calls this the "beautiful junk" trap. A prototype can look impressive, yet employees return to the old process when the agent makes mistakes, lacks context, or requires extra effort. He believes prompting is partly a delegation skill. People must define the result they want, supply enough context, and review the output. Managers face an unusual tension because employees often perform best with room to use their judgment, while AI agents require much tighter instructions.
Another problem is the way organizations treat implementation. A traditional CRM project begins with mature software, installation, training, and a defined handover. An AI agent may begin with inconsistent results and improve only through continued use, evaluation, and correction. Rytis argues that businesses must treat agents as products that require ongoing ownership rather than projects that end after launch.
Omnisend's response began with broad access to AI tools. The company then created recurring AI days when most employees canceled meetings and spent time experimenting. Accountants, lawyers, and other teams began building their own tools rather than waiting for developers.
Rytis shares an accounting agent that checks whether employees have supplied reimbursement documents, sends reminders, and asks a person to intervene when repeated requests fail. He also describes a legal-review agent that examines new AI tools and assigns a green, yellow, or red status. Green tools can be used without further review, yellow decisions go to legal counsel, and red tools are rejected.
Omnisend is now formalizing this approach by offering salary increases of between 2 and 4 percent. Rytis says individual contributors must demonstrate that AI is saving time on repetitive, low-value work. Employees can qualify by building a useful tool, helping colleagues create one, or becoming an effective adopter. Managers are also assessed on whether their teams are using AI to reduce time spent on routine tasks.
The policy creates a genuine debate. Financial rewards can give employees permission and motivation to change established habits, but they could also encourage people to automate work simply because a reward is available. Rytis says Omnisend has not identified cheating or harmful behavior so far. Decisions are decentralized to managers, which gives teams flexibility but also places considerable responsibility on management judgment. He notes that Omnisend has approximately 260 employees, a scale that may make this approach easier to oversee than it would be inside a much larger enterprise.
The conversation includes an example of a recurring agent designed to identify and recover failed customer payments. It assesses risk signals, detects failures, contacts customers through several channels, brings account managers into the process when needed, and reports results through a dashboard. The goal is to remove friction for Omnisend and customers rather than adding an AI layer with no clear result.
Rytis also offers a candid account of customer-support automation. He says AI now handles around 40 percent of Omnisend's support tickets. When the system launched two years earlier, customer satisfaction was almost three times lower than the human team's result. Two employees worked continuously on training the agent, and Rytis says human and AI customer-satisfaction levels are now comparable. The figures are Omnisend results shared by Rytis during the interview and should remain attributed to him.
A separately recorded section also covers AI inside the Omnisend product. Rytis describes recommendations that identify possible improvements in marketing automations, natural-language segment creation, and MCP connections that let customers work through ChatGPT or Claude before sending campaigns through Omnisend. He says these capabilities have received the strongest customer usage among the company's AI work.
This is an honest discussion about incentives, experimentation, uncomfortable tradeoffs, and the patience required to turn an unreliable agent into a dependable colleague. Would a salary increase encourage meaningful AI adoption inside your organization, and who should decide whether the result deserves the reward? Listen to the conversation and share your thoughts with me.
[00:00:00] AI agents are only as strong as the data that they're given. When provided with an outdated data set, your agents could end up doing more harm than good. But not with Denodo. With an AI data layer built within your platform, your agents are provided with real-time data changes. So with Denodo, your agents can finally make the right business decisions. Simply visit denodo.com to learn more.
[00:00:27] Would paying your employees a little extra turn AI experimentation into useful everyday work? Well, today I'm thrilled to be welcoming back Patrick Mostbeck, CEO and co-founder of Omnisend and friend of the show.
[00:00:51] Our last conversation takes us back to December 2022 when the world was a very different place. We weren't all excited about AI, but of course today that's all that is in our newsfeeds. And at Omnisend, they're offering their employees salary increases of between 2% and 4% when their AI work saves time, improves operations, supports revenue, or helps their colleagues work differently.
[00:01:20] And he will explain today why tools and training alone fail to change habits, and how flashy prototypes become what he calls beautiful junk. But why AI agents need to be treated like junior colleagues who require direction and patience. And we'll also discuss employee built tools for accounting, legal review, payment recovery, and customer support,
[00:01:47] along with the harder questions around pressure, fairness, management, judgment, and whether financial incentives could encourage people to automate the wrong work. But I don't think he has seen any of that. But I want to bring that up anyway. So, enough from me. We've got a lot to get through today. And I think we might set off a few light bulb moments. So, let me bring my guest back onto the show now. Well, it is a huge honor to have you back on the podcast to join me.
[00:02:16] For everyone that's missed our previous conversations, could you tell them a little about who you are and what you do? Yeah, thanks for inviting. Yeah, apparently it's around five years when we talked the last time. It's really great to be back. So, I'm co-founder and CEO of Omnisend. And Omnisend is a marketing automation platform. So, we basically help everyone who sell online to communicate with their existing customer base.
[00:02:40] So, invite them to purchase again and again by sending email messages, SMS messages, web push notifications, WhatsApp is closing soon. So, basically it's all opt-in communication as the first purchase usually is a loss for anyone who sell online because you pay more for Meta than Google. And it's very important to communicate with existing customer base in order to invite them to purchase again.
[00:03:05] And yeah, we help to automate a lot of that communication and do manual campaigns as well. But it's really great to see that a lot of automated, very targeted communication works the best. So, basically a bit fewer advertisements, very, very, very targeted and very well received by customers. So, of course, we will talk about AI.
[00:03:29] So, we bring a lot of AI capabilities into our product as well as we do a lot of changes, a lot of many changes inside the organization to adopt AI. And I felt compelled listening to you there just to very quickly look up the last time we spoke. And it was December 2022. So, not quite five years. But the world was a very different place in December 2022. Nobody was talking about AI.
[00:03:56] And, of course, fast forward to present day, AI is everywhere. I've seen that you've since described a lot of corporate AI projects as falling into the beautiful junk trap. So, what does that look like inside a business? And why do so many impressive prototypes never actually become useful everyday tools that we expect them to be? What's happening here?
[00:04:17] So, I would say kind of there are two fundamental shifts that are happening in our society and like in corporate life and in professional life in general. So, first shift, I would say, is like habitual for all of us. So, those who are individual contributors or were individual contributors in the age of prior AI, we have no skills to delegate. And basically, what does like prompting mean?
[00:04:47] It's a delegation skills. You have to ask for some specific outcome and then you will get good outcome. If you don't ask for that and you just say, okay, write me copy, write me a text, write me a book, yeah? So, AI will do something, but it'll do it at the level of someone junior is doing. And if you don't prompt well, so basically, if you don't formulate the task well, you will get very mediocre or very bad outcome. So, that's the first thing.
[00:05:16] And for those who are managers, the big challenge is that for many years, we all, I mean, we managers, those who manage people, we thought that, okay, you should not micromanage. You have to give a direction, give a big vision, hire good people and then basically free them up to do their things. But with AI, managers now have to combine two of those.
[00:05:41] They still have to manage humans in the same way, like a lot of freedom, contextual management and freedom to find their ways to accomplish your goals. But with AI agents, you have to really be way more specific and micromanage at many points. You have to be very specific, what outcome do you expect? Because if you just give too much freedom to AI, generative AI or AI agents, they will hallucinate.
[00:06:09] Or maybe we'll do something, but not exactly what you expect. Because they don't have the full context as human to have. So, this is kind of, I think, this is the first thing that this fundamental shift is happening in our heads. And it's not that easy for this to happen. So, that's the first reason why a lot of great initiatives and a lot of great tries end up in junk box, as you say.
[00:06:34] The second reason, though, and this is the learning from Omnisend experience as well that we face, that deploying AI capabilities, and here we talk about agentic AI, not just Google replacement. Yeah, so everyone is adopting this. So, like, grammarly, spelling replacement, it works really well. Everybody's using it. There are no challenges with adoption at all.
[00:06:57] But, like, agentic AI, basically what happens is that you have to shift your mindset from project to product. Because whenever you start employing AI agents, usually at the very beginning it's quite clunky. It hallucinates. It does a lot of mistakes, but you already have to start using it. And we in organizations, we usually kind of think in the project way.
[00:07:24] So, okay, we made a decision to have a new CRM. So, we find the CRM, which is polished, established software. We hire someone to install it into our organization, to train people, to do the things. And basically, the project is over. From now on, everybody is using new CRM. And it works really well because, like, there is a company behind the CRM that was building the CRM for 10 years or maybe 20 years.
[00:07:53] It's really well polished. It works really well. With AI agents, it's completely different. We build some new agents. We try to use it. Oh, nice. It does a great job, etc. But then we see that every third time it makes mistakes. And then we default back to our behavior and say, okay, I mean, it's easier for me to do it in the old way, in the manual way.
[00:08:17] So, I think there are those two main reasons that are very, very deeply entrenched in our habits, in our mindset. And those are the two main reasons why a lot of AI projects and initiatives fail in companies. And I would say over the last 18 months, two of the biggest topics have been around getting the return on investment and delivering measurable value and increasing adoption. And before you joined me, I was doing a little research on all things OmniSend.
[00:08:45] And I was reading that you're now offering employees salary increases of between 2% and 4%, I think, for AI projects that have actually gone on to improve efficiency, revenue, or operations. So, tell me more about that. What made you decide financial incentives could work better than simply saying, I need you to use AI more? What was the story behind that? Yeah. So, yeah, you're absolutely right. So, we started at the very beginning of this year, 2026.
[00:09:15] So, and prior to that, there is a bad history of that. So, basically, we started with just giving all the freedom to try all AI tools anyone likes to. So, basically, it's unlimited. You can buy any license, et cetera. Initially, that was like total freedom. Now, we started measuring a little cost of that. So, basically, giving all the licenses, try everything you want.
[00:09:40] And second step was we had AI days for maybe 18 months or so. So, once per month, maybe less, maybe 12 months. So, once per month, we basically forced everyone in the company to cancel all your meetings, internal meetings, external meetings, et cetera. Have nothing else planned except, of course, tier one support agents. And play all things AI. So, initially, we really played with it.
[00:10:09] So, we built some little games. We built some videos. We built some small tools. And from there, we start basically familiarizing ourselves. What are the capabilities of AI that I can code? I mean, I'm an accountant and I can write a code. I can write a program. I never thought that I can write a code and build a program. And just to familiarize everyone with what the possibilities of AI.
[00:10:37] Then, from there, we start building inside the team some small projects. And some of them really worked well. And they stick. Just to give you an example. For example, accountants and our bookkeeping and accountancy team. Now, accountants is being basically not by AI. I travel a lot. And our colleagues do. Sometimes we forget to provide documents for reimbursement.
[00:11:07] It tracks. And prior to AI agent, there was human who was checking, writing as reminders, asking to provide it, etc., etc. Now, it is being done by AI agents. Every month, repeatedly, okay, checks what invoices did readers provide, which invoices did readers forget. Send me an email. Runs the same cycle after a few days. Checks if I provided something additional. If not, sends me a reminder.
[00:11:35] If I don't do it for a few times, then it informs the person. And then she or he reached out to me personally. But basically, it was written by our accountants. I have a good example about our legal team, let's say. Then we opened up that you can use any AI tool you want. But at the same time, there is a lot of junk in the market as well. So we need to vet each tool a little bit.
[00:12:04] So legal team was doing that and it was pretty, pretty time-consuming work. Not difficult, but time-consuming. So now we do have an AI agent that tracks. That not tracks, but basically you put the link to the tool you want to try, which is new in the market. So basically, there's an AI agent that scans web, checks the reputation in general, and gives a color coding. If it's green, we can just go and use it. No human confirmation is needed.
[00:12:34] If it's yellow, then the human, the legal counselors are in the loop. They have to confirm. And if it's red, so you just don't use it. It doesn't look good. So, and those two tools that I mentioned were being built by our accountants team and legal team accordingly, not by developers, not by engineers. So, kind of that was a goal to train everyone that actually you can build tools yourself. So that was a second iteration.
[00:13:02] And from there, we appointed a few people that now we have AI operations people inside the organization. And we said, okay, so now we have to formalize it a bit more. And yeah, so this salary increase is one of the tools that we're formalizing. And you know, on top of that, at some point in our all-hands meeting, I said to the entire organization, you, all of you will lose your job.
[00:13:31] The job that you are doing today, currently. But I have no doubts that majority, vast majority of you, or maybe all of you will be able to adopt a new reality. And I hope all of you will stay with Omnisend. And you will earn way more than you earn now, just because you will be way more efficient. But you have to start doing things in a new way. Yeah. So, salary increase is very kind of related to that.
[00:13:59] And for people listening, maybe a business leader listening in the US or somewhere around the world that are thinking of following in your footsteps, I've got to ask, what does an employee actually have to demonstrate before their AI project qualifies for the award? And how do you separate something that is genuinely useful from just a clever kind of cool experiment? How do you separate that? So, in this case, kind of if those are individual contributors, so they basically have to prove that they really save time,
[00:14:29] especially kind of in tasks that are really time consuming, but low value added. So, that's the main goal, because all of us, all of us, doesn't matter at all, we have some routine tasks that are really time consuming, but they don't create much value. It doesn't require critical thinking. It doesn't require creativity. It doesn't bring the organization forward. But we spend time. So, basically, it's pretty visible. I mean, experimenting is nice. You can do it. And it's both.
[00:14:59] I mean, you can build those tools or capabilities yourself, or you can just be very first follower and first learner from whatever your colleagues do, or maybe someone in the market. That's the key to copy, basically. We don't expect everyone in the company to build their own agents. It's either you can help others and build agents, or you can just adopt them efficiently and start using.
[00:15:25] And if you are a manager, so basically, it's not only you are using, but your team using AI and basically saving hours for low value tasks. So, that's the main criteria. Ah, okay. Makes perfect sense. And I believe on the back of this, you've had employees building and using AI tools inside Omnis then already. But do you have any useful examples that would bring this to life?
[00:15:52] And what was it that made those projects in particular succeed where others might have stalled? What have you seen here? Yeah. So, probably kind of like to emphasize here that we as Omnisend, as organization. So, the example I'm providing currently, those are us as organization. And, of course, Omnisend as a marketing automation tool. So, we do have AI capabilities both inside like core of company and as well as in tools.
[00:16:20] So, currently, maybe kind of the inside the company, inside the organization is very important that we're focusing now mainly on like, as we call it, closed loops initiatives. So, the example is I made here that, you know, like a vendor's vetting tool is not a proactive agent. Yeah. There is someone I have a need. I go there. I check. I use it.
[00:16:49] Which is, I mean, that's really great. It saves a lot of time for our legal team. But what we're focusing currently mostly on are those who are like perpetually running basic agents. Yeah. So, one of the examples like failed payment recovery, let's say. Yeah. So, we serve hundreds of thousands of customers. And each month, we have quite a bunch of failed payments. So, we do, we used historically some tools to recover those failed payments. But what currently we build, the agent now.
[00:17:19] So, the agent like detects, okay, first, it goes further. So, the risk detection that there are multiple signals that, okay, this payment can be failing in the future. So, that's first step. Second thing is really kind of detects the payments that have failed already. Then it reaches out automatically via email, via AI agents' rent calls, via text messages, etc.
[00:17:49] And it's the idea of AI agents to connect with a human account manager, basically. Our account manager and AI agent to connect with a customer who has those failed payment issues. Because sometimes those are just maybe lack of funds in credit cards. Sometimes there are some issues with processing credit cards. And then we can find solutions to it to prevent it in the future. Because it's a friction for us.
[00:18:14] It's a friction for our customer, usually, because sometimes maybe they set up the campaign, schedule the campaign. And it's not being sent just because their credit card charge has not been processed. And, yeah, and then it's really, really persistent. It's ongoing, like, loop machine. And tracking, finally, it shows all in dashboards. So, how many failed payments did we have? How many of our customers did we manage to reach out automatically? No human in the loop yet.
[00:18:43] How many of them did connect our human account managers, as we call it, account techs per person in this case? And how many of them did we recover? We always have a dashboard. And each time, each time the new, like payment approaches, our AI agents analyze risks. If it identifies further prior to failing, it's even better if it fails. So, it takes action, action, action, action, and each time.
[00:19:11] So, basically, this is a loop that it always happens. And the agent does, like a new colleague here in this case, not only on demand, but it's really proactive and acts and includes human colleagues in the loop whenever it's needed. It's incredibly cool what you're doing here. And I suspect there'll be people listening around the world incredibly envious of what you've accomplished.
[00:19:33] And there is a lot of pressure on every organization to almost show that they're adopting AI and obtaining value from doing so. So, how can a leader listening maybe recognize the difference between those genuine operational progress that can be made and also what you and many others call AI theater? And I think we've all seen plenty of examples of that. Oh, that's a very good question.
[00:20:00] So, I think it's kind of like, I generally believe that leadership has to show the lead here. If like CEOs of companies and like top-level leadership does not encourage and does not show the example, it does not like in our case, as I mentioned, we had those AI days. It's a massive cost for organization. I mean, we dedicate time for everyone, for every employee to play with AI, to experiment with AI.
[00:20:30] And we have like a lot of unsuccessful attempts. And then we've been in multiple times in traps. That's okay. At a glance, it looks, oh, I will automate this. Boom. But it doesn't work. Yeah. It does too many mistakes. It's just apparently it's like too complex, et cetera. So, but that's okay. I mean, through those mistakes and through those successful attempts, you really make a progress.
[00:20:54] So, and I believe it's kind of like the main question is really how much of your work, low value work, do you really offload to AI agents? So, and yeah, getting back to this search replacement. If you would ask people to use AI, yes, they do. Yeah, of course. But you just replace Google with Gemini, with ChagieBT, Claude, et cetera.
[00:21:20] But how much of your really work have you offloaded to AI agents? Usually this is kind of repetitive, routine, low quality, I mean, low value generating work, but inevitable. So, I think that is the real litmus test. Yeah, completely agree with you. And I think many organizations say they provide employees with AI tools and training, but usage still remains very inconsistent.
[00:21:49] And I'm curious from the incentives you've offered here, what have you learned about the human side of adoption, particularly around motivation, confidence, and giving people actually, just giving them permission to change how work gets done and question that old mantra of, hey, we've always done it this way. What have you learned there? So, probably the main learning permission to change is not enough. You have to push people to change.
[00:22:15] And I think that Tom Nesend in general is we are really innovative organization. We built a tech organization. So, the change, the constant change, the constant adoption to changing various technologies, it's in our DNA. But even for us, it's not that obvious to fully adopt AI and fully change old habits.
[00:22:39] And especially to what, I'm sorry, to what I emphasize is really that I try new AI agent, it is clunky, and I stop there. I default back to my old habits instead of like going through this initial phase. It's something similar like training your new junior or colleague, you know. You have to put some effort because in the first day, someone junior in your company will not perform like as a great specialist.
[00:23:06] But we, so that's why it's kind of AI agents are more like, more like colleagues, very efficient, very fast in many cases. But still, it's more like, you know, like colleagues, less like a software and our expectations has to be like that. So, I think it's, for people, it is a combination of both and it has to come from leadership.
[00:23:33] It's a combination of freedom as well as a push. So, that's why we have those even salary incentives. Salaries incentives that it's a tool to be pushing. And yeah, we have and we had some colleagues who are not willing to really adopt AI at the pace and scale we as an organization are willing to.
[00:23:57] And although they were really great specialists in their age prior to AI, but they left the company because they did not feel like comfortable with demand that we currently are expressing to everyone to change and to adopt.
[00:24:21] So, I think it's, I mean, the change, adopting to change for us humans is not that obvious and it's not that easy. So, and I think the transformation we are going through, we as a humankind is fundamental. It's one of the major disruptions, probably equal to steam machine or electricity and bigger than internet, I believe. So, of course, the change is not easy, but that's inevitable.
[00:24:49] The question is how fast you will go through that change or will you get stuck in the previous age as a dinosaur, you know? Yeah. And on the flip side of all the positives we're talking about today, I suspect there will be people listening, a few that will say financial rewards are great.
[00:25:09] They can encourage greater experimentation, but they could also create maybe unintended behaviors such as people chasing bonuses or automating work simply because they can. So, how are you managing that risk and making sure that the incentive produces that real business value without unlocking anything else that you don't want to? So, probably, currently we do not.
[00:25:36] And at least we have not spotted any, any, any, like ill intents or misbehavior in that or like cheating attempts, etc. Yeah. Yeah. Yeah. Yeah, there are sometimes like discussions with managers and we left it kind of, it's not HR, it's decentralized.
[00:25:57] So, basically managers are evaluating their direct people who report directly to them and they make decisions if they are like, not even like eligible, but eligible is more kind of, but if they are using AI in the way that it like fits like what we expect. Yeah.
[00:26:25] So, we have a list of criteria, but, but it's left to manage just to make the final decision. So, so we, we didn't spot any, any kind of bad behavior or cheating attempts, etc. Potentials can be what we are not like that big of organization where 260 people or so. So, so I believe that's, that's okay.
[00:26:46] You should not, in general life, I think that you should not overthink unless you really have any challenges or troubles or issues. So, we don't spot any issues here, so we are not fixing it. So, when it comes to your customers, how is AI transforming email marketing and how is Omnisend internal AI practices? How's this helping meet their needs too? Sure. Yeah, absolutely.
[00:27:13] So, it's not only us employing AI within our company organization, but, but we're bringing a lot of capabilities for our customers as well. And probably kind of the most of the, the most of the best received capabilities that we currently do. So, first one is like AI driven recommendations. So, our customers have like hundreds of thousands of automation set up.
[00:27:40] And in a lot of them, they kind of miss something, maybe small details, etc. But they still miss. So, so now we do have AI recommendations in line. So, basically in protocol, whenever someone is browsing Omnisend, we do put recommendations. Look, we saw this kind of inconsistency in the automation. We saw this, maybe not mistake, but the way to improve it, etc.
[00:28:04] And people can just review it and then we can accept it or discard, of course. But if they just accept it, we fix it for them. So, this is kind of by far the best received AI capability. Second is like building segments, usually for marketers, especially those who are more creative people, was always quite the challenge. Because this is like all those logical junctions, you know, if and or if blah, blah, all those things.
[00:28:34] So, now you can prompt it in the human language. Please build me a segment of those customers who are the best buyers in the past three months or have not bought in the past six months. And we just build it for customers. So, those are really, really well received. And probably the biggest is like MCP. So, connections directly to ChatGBT and Cloud. So, basically, we have already customers who just barely come to Omnisend this platform.
[00:29:01] We still use heavy users, but we do everything on the interfaces we prefer doing. So, they basically write a copy, analyze everything, design everything in Cloud.ChatGBT. And then they just push everything to Omnisend. And we do send. We collect all the data. And we can pull any analytical data that they want to Cloud and ChatGBT.
[00:29:25] So, probably kind of those are like three examples how we are really making or AI is making life way easier for our customers. And, of course, making them way more efficient in the daily work they do. So, yeah. I would imagine little things like being able to ask who has not purchased from us in six months. A huge opportunity there. What kind of feedback have you had from… Yeah. So, that's the best feedback we receive.
[00:29:53] Of course, and we tried a lot of other capabilities that are not… I mean, they're not kind of negatively received, but maybe not used that much. And those that I just mentioned. So, our customers love them. And they're really using them very, very actively. Great to hear.
[00:30:13] And for anybody listening inside an organization that may be looking at its own AI pilots today and wondering why so many are making it or why so few are making it into their real workflows, what would you advise those leaders to change about the way they identify, fund, measure, and scale employee-led AI ideas? From everything that you've been on your journey here, what would you advise people listening? Yeah. Yeah.
[00:30:40] So, probably kind of like I would emphasize to what I already mentioned. So, it's more like product thinking. So, basically, you have to start using AI agents. You can do it in lower scale, but initially they will not be ideal. It will be far away from them being ideal, but you have to start using it. And from there, you have to keep training it.
[00:31:04] And maybe just a good example here would be our currently around 40% of our support tickets is being handled by AI agents. Yeah. But what is very important, once we launched the AI agent two years ago, the satisfaction of quality of support service was way lower than our human support got. Like, almost three times lower.
[00:31:28] So, for two years, we have two full-time human employees who are always, every day, training, training, training, training, training our support agents for two years. Now, after two years of quality and customer satisfaction of a support service, AI agents and humans are on par. But, again, we should have launched this way less quality support agents and constantly train, train, train.
[00:31:56] And after two years of constant training, constant improvement, now it's really doing a very efficient and very good quality job. So, I think it's a good example of that I would advise. Yeah. So, it's inevitable that we will have to employ AI agents in the future. So, you better start now and paying the price. And for the time being, highly likely the efficiency will drop, the cost potential will increase.
[00:32:23] But it's, again, it's something similar to hiring the very smart but still junior colleague. It takes time to train them, to fully onboard them, for them to learn everything about your organization, everything about work, etc. But after some time, it will be a massive help for you. To prove that, that's the main advice. Wow. Well, I absolutely love chatting with you again today.
[00:32:51] As I said, the last time we spoke, the world was a different place. And hearing you today talk about what employees are building and using with AI inside companies. And some of those great examples of what's happening at OmniSend there. I'm sure it has unlocked a lot of light bulb moments and people listening around the world. So, people listening want to follow in your footsteps, want to find out more information about OmniSend, connect with you or your team. Where should they go? Yeah.
[00:33:20] So, OmniSend.com, our product. You can find me on LinkedIn. That's my preferred social media channel. So, Reedus Loris on LinkedIn. Well, excellent. I mean, just talking with you today about that growing divide between AI theatre and operational AI. And also why companies might need to consider financial incentives rather than mandates if they really want to drive adoption. I'll include links to everything. I encourage people listening to check you out.
[00:33:49] But more than anything, thank you for joining me today. We've got to promise each other we're not going to leave it four years until we speak again. And we still need to grab that drink as well. But we'll make that happen. But thanks as always. Thank you. Great to be here with you. Wow. I think my guest offered a useful reminder that operational AI, it rarely arrives polished and ready to go on day one.
[00:34:12] An OmniSend support agent reportedly needed two years of daily training from two employees before customer satisfaction was able to match the human team. And this is a very different story from installing a finished software package and declaring the project complete. I think the practical lesson here is to choose lower value repetitive work. Begin on a controlled scale. Measure the result and keep improving.
[00:34:41] Keep improving that agent while human judgment remains available. And yes, salary increases could provide a reason to start. But leadership behavior, protected learning time, honest failure and a product style ownership. All these things can together determine whether an idea will truly survive in the workplace. Remember, you can learn more about OmniSend at OmniSend.com. Connect with my guest on LinkedIn. You will find all the links in the show notes.
[00:35:11] But over to you. Would financial incentives improve AI adoption in your workplace? Or is it? Or could it create the wrong motivation? I realize there'll be a lot of opposing views on this. I want to hear from you all. TechTalksNetwork.com. That's where you will find me. But right now, it's time for me to go. Speak with you all again tomorrow. Bye for now.
[00:35:35] Bye for now.