This year we've worked with a lot of companies to integrate AI solutions whether they werememes or mid-market firms or enterprise and even governmental agencies and we've done it across several industries as well. And in this video I want to share with you the seven top insights we've gotten and what differentiated those that got an ROI from AI and those who did not get an ROI from AI because we've seen the reports by McKenzie. We've seen a couple of posts or articles on the web saying that AI adoption is low and nobody's getting the ROI. It's all hype and so on.
And in this video, I'm going to show you what is the moving factors or the seven principles that make that company get an ROI from AI. So without further ado, let's jump into this. So the first question you may be asking is what is the urgency behind this? Why should I care even to think about this right now?
Because everything is going downhill from here, right? No, because if you look at the companies here in the Middle East, we are all driven or we have a certain momentum. They set a vision far into the future that we do not see but they have this big plan in the direction we want to move and from that plan what happens is people start building the infrastructure and under that the talent pool and under that the companies are born and under that individuals which are benefiting from AI right now. So it's kind of a hierarchy over here.
And if I give you some examples, look into D33 by Dubai or look at vision 2030 in Saudi Arabia. These are all the momentum pushing forces and where AI is a very high leverage factor within these initiatives or visions that have been set for the future which are approximately seven years let's say plus minus 7 years from now. But this is where it becomes important and why I'm sharing this in this video is that reality on ground with the businesses we've been talking to right now is that many people are looking at AI in a way that is not pushing them forward but more of stagnant and staying their place dabbling with AI and not getting the ROI they want. For example, they're buying certain tools that they do not need and they do not need it because they haven't gone through the approach of why do I need AI?
How do I use this tool? where do I fit it in within my business? So many problems occur under that. So let me give you seven principles right now which guides you in order to get that ROI from AI and use AI as a tool and leverage it to what you want to achieve within your business. So let's start with point number one which is point number one being diagnosis before prescription.
Meaning that the biggest mistake I've seen businesses do here they start with the solution instead of the problem or where they need AI. Meaning that a vendor would come from outside or let's say any company I don't want to name companies but company X comes from outside and says we have this amazing AI solution you just put a prompt and you get the app out of it or we have this amazing AI solution where you can agentically code whatever and then they see that solution and say okay maybe we need this within our business that's the opposite approach of actually using AI so they go through the prescription approach and not the diagnosis approach and of course I understand branding is important important and everything is important. For example, if we have in our company, we got associated with someone very well known in terms of we branded, we signed an MO and so on, it looks good for us. We may get more business and so on.
But that doesn't mean that we're getting an ROI out of AI itself. And the majority of businesses since the vision is set, we need to contribute as business owners. We need to set good, let's say, objectives or problem statements so we can achieve with AI. And this is where principle one starts.
Every project we enter, we start with a diagnosis. We don't go and say this is an AI solution. You need a chatbot. You need a voice agent or whatever it is, but more of let's look at what your top 10, 15, 20 employees are doing on a daily basis.
You want, you said you want to get, for example, this example over here. There was a S Saudi logistic company that spent $200,000 on an AI chatbot for customer service and it didn't give them the ROI they want. But if you have done the audit initially, you'd see, okay, maybe you don't need a chatbot in this case. Maybe you need something else.
Maybe you need this chatbot for somewhere else. The point I'm trying to make is before touching any AI solution, first look where do you need to implement AI. The point number two over here is that asking these three important questions when you start. Question number one is where am I losing the most time right now within my business.
So if you think right now like on spot if you think about it where are you losing the most time? Where what processes take a lot of time to get done? who's what like tic tac kills for example the back and forth between emails or researching a particular thing and giving a report or if you make a request to your company and it comes back to you after 3 months or you've had situations this is a very nice example or you've had situations where you see someone do something in one hour when you have employees that take six months three months I'm over exaggerating here to do it so think about those small tasks and those repetitive work So think about that and you're going to start focusing on your pain. I'm going to show you a framework in a bit. So that's point number one.
Then ask yourself where do I lose money the most? Like even think about the expenses that you cannot see. For example, a salary. A salary you're paying someone to do a certain part particular task.
Maybe they're doing low leverage task where that salary is being spent or let's say let's not take it an expenditure but you're giving that salary and it's it's it's not being utilized a proper way and giving you the ROI you want because we're handing that person the wrong task. So input output kind of system uh design thinking and the third point over here where am I losing talent where again coming to this point which is where am I giving the wrong work to the right person or the person that could do something proper and this is a framework usually we go through so to find an opportunity within our business first we start with what what do I want to do or let's start with an audit and see the workflows and point number two we ask these questions of where am I losing time the most where am I losing money the post and where am I using my talent or I'm not using my talent and resources I have properly and usually this comes into a nice intersection let's say it becomes the sales department in the SDRs or if the marketing department in the communication between the sales like it comes down to a very nice it's company specific as well and you have find this AI gold mine over here which is your highest ROI opportunity and then and only then you start looking for solutions for this exact object over here and the right solutions. We don't start with the vendors but given the position of the Middle East and this is why I'm bringing it up over here is that many people come here because they think okay Middle East they have money okay let me go and pitch my idea get funding this is very common and this is very understand it's very transparent there's nothing to hide over here but if that is the thought process we need to understand and we understand that okay we're not going to take any project or any anything that we have if we don't have a need for it unless it's for branding purposes of course. And then point number three over here, implementation over strategy.
So behind closed doors by June, I've been hearing this exact code. People saying that we're tired of the presentations. They haven't gotten me the ROI and so on. These are executives saying it behind closed doors and saying we want execution right now because we have real KPIs and these presentations don't get the KPIs.
And again if we come that back to the point of branding and marketing yes it's nice to associate our names with these big players but is it doing us good in any way if we associated with these big players and we're not getting an ROI because people would see our business that we have right now for example we're a laundry business or whatever we're associated with this big huge name but we're as a business our AI chatbot is insulting people on our customers so it does not look good as a reputation so We need either to match each other or they're going to pull us up to them and if that doesn't happen it has a backfire effect of course. So that's why we always start as I said with an audit then we build then we measure then you scale and as you build we go through the process of we start with an assistant we don't go high automation AI agent agentic mesh kind of thing where is highly integrated within every system without testing it or piloting it or having any metrics behind it but more of let's have an assistant and see how employees react let's have an assistant and try this tiny use case over here and see if there is potential or signs of life that this could be profitable or bring us the ROI we want. So a slow approach is the best approach. And point number two for implementation and insight I would share over here is that AI itself or if you look at the nature of artificial intelligence is a very probabilistic kind of technology.
So one solution, let's say an image generator that works for one company or industry or and in one department does not necessarily mean it's going to work for your industry and department. Even if you're in the same industry, but you have you're at different let's say income levels or different maturity of a company, it's not necessarily it's going to work because again this is probability and that's why we start with a pilot every single time when we come to the implementation. So strategy is good. Yes, amazing. that exists.
It's there but it does not achieve unless we have proper implementation and right systematic way. So point number four and this is where the majority of people miss is that what works outside does not necessarily mean it's going to work here and for example Arabic is very important. For example in Saudi Arabia SAMA regulations are very important by the central bank. The DIC standards is very important if you're in Dubai. the Arabic language processing is very important that a big complaint I've been hearing in this region is that these models are optimized on English.
What if I want to use it for Arabic? It doesn't make any sense. I need to fine-tune it or I need to fix it to make it better for Arabic speaking language or Arabic applications, let's say. Or think about the cultural thing.
This is something many people miss. How do our decisions made in the Middle East? How can I make AI understand that? because we can't I mean use um anything and make it fit within this region as a whole and then how aligned is it with the visions this country or the countries in the Middle East have set. So this is very very very important and I've seen this being a make or break kind of scenario for many big companies and without naming names there was this big company that came to set up in the UAE and they were talking to this particular entity and the reason that deed did not happen and I was talking to people and they were saying because they didn't understand this region this is not how we operate this is not what how we do business and so on so this is super super important point number five ROI with real numbers so as I mentioned In June when I had meetings the first question used to be asked is what is that one metric ad if you provide me this or we do this together what are we measuring to measure our success and I can tell you for a fact in January this was not the case in January was like I want AI just add AI let's let's test it out and see but in June it was more of give me a metric I need to measure something and for from someone to go from there to there as a decision maker something happened something wrong happened or something some pressure came from somewhere right so setting metrics and the right metric as well I can say I'm going to set for example we're let's build let's this is just an example and [laughter] it's not real example as well so let's assume there's a company that wants to build capacity for AI AI native or AI first employees a metric can be that okay we're going to put the metric being how many people we trained or let's let's say how many heads attended in their words in this workshop that we run this is a metric I have 100 people attended that means I have 100 AI native or AI first employees because they were taught about XY Z but that does not necessarily a right metric to track for this case even though it checks a box within an organization or a company the right metric is how many of these people actually got the outcome which is being more productive or whatever the point of that workshop or education or consulting experience was and how many are actually implementing it and making business better because at the end of the day any business cares about the bloodline of the business which is cash flow and the reason they do all of this is to get more cash flow whether it is to save time save money or get more money and that that is the final question.
So the closer the metric is towards the money, the better it is in terms it makes more sense towards a business. And you can see this a lot in pees and VCs like private equity. They they really focus on these particular things. So this is very very important as well.
Then systems not tools. So every single day I you know I know a new tool is coming up. Yesterday our team just created a new tool because we needed this internal thing to do it and there was I'm going to tell you the tool. was uh for YouTube and we wanted to optimize the process in terms of the research of the YouTube and so on and we saw tools on on the internet we said there's this tool this tool this tool this so none of them fit our needs so we just created a tool and it took us like two hours to create by the team and it we're using it every single day now the point I'm trying to make over here these tools are very easy to create and are getting easier to create with each launch for example opus 4.5 just launched and it's 80% I think on S swe benchmark meaning that we're almost close to 100%. So yeah many factors over there but the point being that AI is getting better.
These tools are going to pop up every single day by many people making those tools. What matters the most now is systems. How do you connect or integrate systems together? Meaning that you have right now an HR system within your company or you have an email system.
Let's say you have Outlook or you have Microsoft or whatever system you have within your company right now. How do you get tools or how do you get these tools that you're going to buy to integrate to form a system all together to give you the outcome that you want? Because a single tool won't get you the outcome that you want, but a collection of tools working together which we call a system to get you that ROI that you want. So now I think it's time to think about systems more than tools or this amazing tool or whatever that comes out.
And just a point over here and this is an observation as well. There are many companies or let's let's take evaluations or let's say many companies right now if you go on X or TW or if you go on the news and so on this company is valuated about $1 million or this one or $500 million and so on. How does an evaluation or the enterprise value of this company come from from where does it come? It's an agreement between the investor and the company.
Right? So these tools right now are essentially agreements between investors and companies. And what usually they do when they want to push this number up is that they start making marketing news, giving a free credits to their users or they start pushing a lot of things that is not sustainable in terms of the growth of the business. And what ends up happening they build out this hype around themselves to push the company forward in terms of that equity or the enterprise value. and then they say, "Okay, we're an amazing company." Then they start pitching to the Middle East regions.
And that's a that's a pattern I've seen a lot where when they come here, it's usually after a planned chess play of pumping up that value, heading to the Middle East to get some funding and then going back. Okay, now this wasn't sustainable. Let's go back a bit, fix the systems and businesses and foundations we have and start pushing forward. So this is another point on the tools that is very spo not spoken of but it happens very frequently.
So essentially what I said here audit build a system and then start adding the tools to get you the ROI you want. And the tool collection approach usually does not work. This worked maybe let's say in 2022 2023 or 2024 but right now this does not work because the tools are like every other day new tools coming out. So AI is an accelerator and not a replacement.
Let's talk about multiple things within this sector. So point number one is the media is telling you that AI is or there in in this sense they're telling you that AI just came in and this ex company such as Amazon laid off 30,000 employees or Microsoft laid off all this and it's because AI is replacing people and AI got to a point and I'd look two levels under that and not just one level. Why did they get rid of the employees? And usually comes down to the capex and OPEX and why what's the least let's say ROI driven operational cost they have to increase their capital expenditure which is capex to build infrastructure or build something around that company.
So I'd think about that sense more of what's the news is telling you because AI right now is an accelerator. It's not a replacement of humans or employees and so on. Yes, certain tasks are replaceable. Yes, certain microtasks but if a person's sole job is to do that that task, this is an opportunity to upskill and not be a replacement as AI.
So this is a very important point I'm trying to make that people are telling you that AI will replace employees and AI is a hype and so on. But if there's massive demand for these infrastructure and I always look at the infrastructure companies which are let's say the clouds the the the AWS and um the Azure Microsoft Azure and so on. So if you look at all these companies if there's high demand and not enough supply that's not bubble that's not a definition of a bubble. A definition of bubble is when there's no demand there's high supply but it's complete opposite.
So I I think of AI as an accelerator as I go through all of this. So this is also just a tiny framework in tool buying versus uh going through the framework path which essentially I tell you go through the audit process and don't just start with the tool itself. And finally how do you want to apply this? These seven principles that I'm telling you right now are insights that we've gathered throughout this year for the companies that we have in the Middle East with the cultural things with the politics with everything that has been happening.
It has been a learning journey for us and take these seven principles or frames in mind to have once you go and integrate AI within the businesses over here and this is something we are taking into 2026 and forward as well and of course we're going to learn a lot of new thing things as we go forward as well. So, this is the exact framework we're using for every single client.