AI in SaaS: Automating Operations & Driving Data-Driven Growth with Jose L. Fernandez

Thank you. Thank you for having me. Delighted to be here with you today.

So working in this giants, tech giants I realize there’s a lot of focus in big enterprise bringing AI to the big enterprise. There’s not as much focus in bringing AI to mid market and SMB in general.
I thought there’s a big opportunity because mid sized companies, they go to trade shows and they go to events and they hear all the time, do AI or be a digital illiterate. And they go back to the office, call the IT guy. I say, let’s do some AI project and the poor IT guy says, yeah, when I fix my cyber security, the problems with the database and a couple of other applications, then I can start looking into that.

So it never happens. So that’s when Salesforce comes in my company. We help out, we basically, we help educate executives. So teach them exactly what an executive needs to know, not too technical, but also not about tools. Executives don’t need to be training tools. They need to know the concepts and the general direction of the technology.
And then we workshop to understand what areas of the business can benefit from this tremendous way, which is AI.

AI is already changing the way companies operate, the way companies are run and big companies are making huge investments.
Already. So if SMBs don’t get up to speed, they’re going to be run over. Some of the areas in which I see more movement are process automation. Everybody has processes and those processes, at least some of the steps can be performed by a robot very efficiently in a very predictive way. With a very low error rate that adds efficiency and people have realized that it’s not replacing people.

That’s not the idea. The idea is to liberate valuable time from people so they can do tasks of greater value. So people are not being replaced by this technology. People are being [00:04:00] Empowered by this technology because they get to spend time when where it matters most if you’re in a sales organization You don’t want your salespeople Spending a lot of time in the office Filling reports and running processes you want them in front of a customer.
So if you automate as much as you can of the back office work, that’s what you are getting more time in front of the customer. The other area of great development is on applications that work on the data. People have been told that there is gold in the data. But nobody’s saying how to mine that goal out of the data.
So there’s applications that do that can absorb data sets and allow executives to have direct conversations with the data. That is amazing. Imagine not needing an intermediary to speak with your data as an executive or as a business owner. That’s fantastic. We used to see that in science fiction.
pictures only. This is now reality and you can do it. Ask questions. Okay. What are my top performing products? Now the same thing, but by region, what are the products that are experiencing the greatest XP improvement in terms of revenue, in terms of gross market. And what are the pros that are experiencing the greatest deterioration in margin and why?
Is this a cost situation or is it simply that my sales team is over discounting? And the ability to do that in a conversational mode is a game changer. Many SMBs are already absorbing and implementing this kind of technology. And it is fascinating. It’s fascinating.

Yeah, in marketing primarily everything that’s process based can be automated. You can automate with RPA, Robotic Process Automation, but this is also the era of the agents, the AI agents.

There are platforms of agents all over the place. So many of the marketing tasks can be automated. And not only automated, but automated with AI tools. So that not only it’s a decision tree based on rules, but it’s based on AI decisions. So that’s a much more powerful approach. Everybody’s saying that 2025 will be the year of agents.
Agents is not something of the future. They’re available right now. Dharmesh, the founder and CEO of HubSpot, launched very recently a website called agent. ai. That is a network of professional services by agents. That is a network of professional services by agents. And I think that’s going to be a big explosion.
You can do agents also on ChatGPT and many of the other platforms out there. And you don’t need to be a programmer in order to build them. You just need to have the logic. So if you are a marketing executive, you can build the logic. very easily as if you’re drawing it on paper but build it into an agent.
And then you have people or you have agents, digital agents helping you out. So that allows a marketing team to be hybrid, some humans, some digital agents. All working together in order to run a marketing campaign, for example.

International expansion is a combination of science and art.
There’s no universal recipe that works. So first of all you use AI to produce content. If you’re a marketing agency you have your assets. And those assets represent style they represent the way you portray yourself professionally. So you can train a model to continue producing content in the same style with similar formats.
So you can have a model that continues to produce content within the same parameters style in general. So that’s one way. Other way is with AI, you can very quickly identify what are the entry points of a new market. Let me give you an example. If you are a foreign company trying to come into the U.S. market, people who have tried that know it is very difficult to do. It’s not easy. So the first thing, and this is not AI related, you need to understand the culture, the U. S. is about what is not about what you say. It’s you do is about what you do. So people like to look at your references.

They like to speak with your customers. And every industry is a different path. So if you’re going to bring software that has already installed an install based from In Spain, for example, then what I do is, okay, what are some of the Spanish companies that are already established in the U S and let’s start there because their references that we have in Spain are going to make sense to these companies because they come from Spain.
And then once you have a couple of those, then you have local references. And then you build from there. And for that it’s very valuable to have channel partners because the channel partners have local presence. And If they are in an adjacent field they’re knowledgeable of the technology you’re promoting, then that’s definitely the path forth.
And you use AI in all the process with the small tasks. You can use AI in so many different ways. It’s incredible.

So the first thing is. Don’t start doing AI just for the sake of doing AI without understanding the universe of what it is. So you have to have a literacy plan for the company. And it doesn’t mean that everybody’s going to go take an eight hour training and that’s it. No, it means that executives Need to learn the basics, the fundamentals, and they need to know about use cases.
For example, then it needs to know about engineering, data science and those fields, then the business leaders in the company, they need to know about use cases, what are some of the use cases? And when people study AI, they understand that there’s two different things. tools for the company and tools for individual productivity.

So the most common mistake is company adopts individual productivity tools and they say, Oh, we are already doing AI. Yeah. Doing AI at a personal level. Yeah, to become slightly more productive and that’s fine. That’s all good, but that’s not doing AI for the business. So that’s the most common. So first piece of advice is don’t think that because you took a training to do, for example, better chat, GPT prompts.
that you’re doing AI, you are using AI as a personal productivity tool. But that’s not AI for the business.
Then you have to be updating your knowledge on a permanent basis, you have to keep an eye open to see what’s around what your competitors Or, and your colleagues are doing don’t wait till one day you realize everybody is ahead of you and it’s too late. There’s a very interesting survey by Pete Marvick that was released a couple of months ago.
This is the CEO survey they do every year. And the survey goes to the CEOs of the Fortune 500 companies in the US. And the conclusion is the more people adopt AI projects, the more comfortable they feel and the more rapidly they adopt newer technologies. And it makes sense because people start with hesitation.
Ah, let’s see what it does. Then if the outcome is positive, okay, let’s do a little bit more of this. Oh, maybe we can do this other project. And they get going but it takes time. So if you wait till your competitors are already running, then it might be too late for you.

Yeah. So the other very common mistake has to do with data. Some companies try to implement very ambitious projects on top of data. That is. not very well structured and it’s very disorganized. That’s one mistake. At the other end of the same spectrum, you have the other problem, which is Oh, my data is not fully organized, so I’m not going to do anything with AI at all, and I’m going to wait till I fix it.

The two extremes are problematic. You need to find the middle ground and start implementing AI solutions with the information you have, given that the consistency of the data is good enough. It doesn’t have to be perfect. And waiting too much is a mistake. And trying to do too much with inconsistent data is also a mistake.
You have to start doing stuff little by little with what you have. And yeah, you need to start cleaning up your data in order to have a sustainable path.

So it depends on the type of application they want to they want to launch. So my take is you have to be specific. in the in with the first application and make sure that the data set that application will use as input is available. That’s all you really need to make sure that there’s availability.

If you wait for perfection, Then you’re never going to launch and you’re going to be late to market. And if you try to do something that’s company wide, that’s also problematic. Because you’re going to invest a lot of time and money. On building something that’s going to have to be updated and changed as you fix the underlying data.
So you have to be specific. Pick a project that covers a particular need. In the case of sales and marketing, for example you can do a campaign manager, for example. You can launch, agent that helps the marketing specialists run campaigns and take some of the intermediary steps in an automated way, for example, that doesn’t present a conflict with the way your underlying data of your company is structured.
So it gives you time. To do that in parallel and take care of that problem in parallel.

Yeah. Targeted marketing is becoming more prevalent because. It’s easier to read the audience. And once you have the data of how your audience behaves, You can segment based on data, not based on definition customer segmentation and market segmentation are not new concepts. Yeah. What’s new is the fact that it is now done based on data, not in definitions.

And you can do segmentation and you can do micro segmentation. Which is even more powerful once, once you have micro segmented your market, then you can very easily see what’s working in those customers in which you have the greatest presence and share of wallet. And what’s the difference between those and the ones that have a very low share of wallet in which you have a very low share of wallet, what’s the difference?
You can leverage AI. to understand that and design a tactical response to try to flatten that curve of share of wallet.

Yeah. I like reading books, but in this case, unfortunately, by the time a book is released, it’s probably outdated.
So books, unfortunately, are not the best source of this. It’s at a philosophical level. Yeah. That’s all right. You have to be paying attention to the blogs. And there are so many that you have to be hyper selective. Yeah. What I’ve done is I’ve made a list. Of who are the leaders I could trust as far as the quality of the information that is released.
The founder of Coursera, for example, Mr. Engy, has a lot of credibility. So he publishes a number of videos in YouTube and in other platforms. So that’s one of my sources. There is a couple of newsletters on AI that I have subscribed. And basically what I did is. I joined six or seven AI groups in LinkedIn and I subscribed to their newsletters and then I subscribed from the ones that didn’t have much value and kept a couple that are more aligned with what I want to read about.
And then I spend about an hour a day reading. And over the weekend, I tried to catch up with the latest and greatest stuff. So this weekend, for example I was watching an interview with Sam Altman about what he thinks about agents. and what he thinks about artificial general intelligence and when artificial general intelligence would be ready.
I also took a look at an interview with Musk and his take on AI what he’s doing with Grok, for example, It’s remarkable. It’s amazing. Then I took a look at an interview with ham the CEO of HubSpot. The way he integrated ERP with CRM and the community or the platform for agents he has created with Agent ai.
I think it’s brilliant.

Yeah. So I try to follow only those extremely influential and credible leaders in AI because there’s too much information, too much noise and much of it is worthless. So you’re being, you’re going to be wasting time. And whenever I see an email with a very catchy subject line, it’s 90 percent of the time it’s just garbage. It’s just garbage. Yeah. I believe you have to start subscribing to newsletters. And then de subscribing from the ones you don’t like, and look up for the greatest leaders in AI. The CEO of NVIDIA is also someone. That I pay attention to because he tends to be very sincere when he speaks. He won’t give you too many specifics because obviously he represents mostly a hardware company.
People will argue Nvidia is also a software company, but not at the application level. But I feel his interventions in social media are very accurate and well intentioned. Sam Altman the guy has a lot of credibility. He’s also very sincere. I take it like, like that.

Yes.

The most? What surprised me the most? What habit

What habit? Not responding very fast.

A habit that I despise the ums when I’m speaking the pause. I think that’s not annoying, but it makes your audience lose the interest. I didn’t realize I did so much of that until I watched some of the videos or interviews or blogs.
So I’ve been working on that, trying to not have those,
That often, because yeah, it makes the audience lose interest.

Yeah. Doing the dishes going cleaning up my email.
That’s super boring. I get, even with all the filters, I get too many emails that don’t even deserve reading. So that’s a waste of time.

I love leadership.

Okay. And everything that has to do with emotional intelligence. I think it’s super relevant in these times. Having the Courage to understand that our fate is 100 percent up to us, each and every one of us. That’s high interest for me. I’m also interested in stand up paddle boarding and pickleball.
I’m playing pickleball like three or four times a week.

Thanks.

Oh, my latest it was refine this email response for me. And then I copy paste pasted something that I had written.

Yeah And Yeah, the GPT is a better writer than I am, I have to say.

Thank you for having me. It was a pleasure speaking with you.