In this episode of Software Spotlight, host Michael Bernzweig sits down with Peter Juhasz, the founder of Syrvi AI, to explore his journey from a challenging business setback to launching a successful AI-driven company. Peter shares insights into how Syrvi AI utilizes artificial intelligence to revolutionize B2B sales, focusing on precision targeting and building trust and authority in the field. Listeners will gain valuable understanding of the unique challenges and opportunities in leveraging AI for tangible business outcomes.
Peter Juhasz is the founder and CEO of Syrvi AI, a company dedicated to leveraging AI to transform B2B sales processes. With over 20 years of experience in business development and marketing, Peter has a deep understanding of how AI can be used to build effective sales systems. He is passionate about helping businesses achieve predictable and scalable pipelines.
[MICHAEL]: So, when Peter Juhasz started playing around with ChatGPT, he realized, you know, just how big AI could be for businesses. It wasn't long before he sold all his previous ventures to dive headfirst into Syrvi AI, focusing on harnessing AI for real business use cases. And what's really interesting is how he took this leap after initially planning never to start another business again. I'm your host Michael Bernzweig and this is Software Spotlight.
--- INTERVIEW ---
[PETER]: I know a lot of the founders and executives listening in today have been on their own journeys with their organizations, whether it's founding and taking that entrepreneurial journey or, you know, working within the organization to get to where they are.
[PETER]: And it's not always a straight path to success, as we both know.
[PETER]: So I was hoping you could share with the audience a little bit about your personal journey leading up to starting Cervi.
[PETER]: Thank you very much, Michael.
[PETER]: So I'm Peter, Peter Givas, co-founder and CEO of Cervi.
[PETER]: And the journey started, although we started about two years ago, Cervi as a business.
[PETER]: But I'm always saying that the journey starts always a lot sooner, starts since we born.
[PETER]: So my journey was going to many successful and unsuccessful, less successful businesses, journeys, business journeys.
[PETER]: And learning that running a business is not easy, whether you, whether you start a startup or whether you buy a business, we don't, we don't both, we're still doing it.
[PETER]: Each of them requires a different level of expertise and gut feel as well and toughness, a lot of toughness to run your own business for sure.
[PETER]: You can never know that the product and the service you're going to be launching is going to be successful.
[PETER]: People will like it or people won't like it.
[PETER]: So it started since I was born.
[PETER]: But my business journey is about 20 years ago, back in my country, Hungary, been through many business journeys and I came to the UK when I lost everything and thought I'm never going to do business again.
[PETER]: And then eventually started to work for other businesses and eventually I said, OK, I think it's not for me.
[PETER]: So I missed, I missed the stress perhaps, and I missed the excitement and started my own businesses.
[PETER]: As well, so in the last 10 years, we had many businesses, we were a top group of businesses in property, hospitality sector, service departments, professional businesses, lead generation marketing agencies.
[PETER]: And that's how, where we get the knowledge of business in general, and especially on the side of sales and marketing and outreach.
[PETER]: So we're doing outreach and we're doing lead generation for clients, SEO services and other marketing services for almost 10 years now.
[PETER]: And when AI came first, I think it was 2023, no, for us AI come first, obviously it's a long, long, long time ago.
[PETER]: It's been here with us.
[PETER]: And we started to play around with chat GPT, that's where we realized that there's something that's here.
[PETER]: And there is a, there is a, this can go somewhere, this can be big.
[PETER]: So since then, the only question was how we can harness the power of AI in business use cases, real business use cases, not just flashy, creating a, you know, MVP and showing on YouTube, but rather something which actually delivers results for a business, actually helps generate more leads, build pipelines and close deals.
[PETER]: Because that's what we do.
[PETER]: That's what we've been doing long, long, long time.
[PETER]: And that's how, that's how we created Servi.
[PETER]: That's how we started to play around, start to build systems for us internally.
[PETER]: And then we seen it's working, then we sold all the previous businesses and got full time into, okay, let's do how we can do this thing, how we can build something, which actually generate benefits for other business owners, rather than they need to start everything from scratch and build their own tools and everything else, which is just so time consuming and just so difficult.
[PETER]: So how we can build something that they can just literally plug in, and then they can enjoy and expect the results.
[PETER]: Yeah.
[PETER]: And I think, you know, that's really interesting because we've seen so many of the apps that we rely on today.
[PETER]: I mean, even Loom or Trello or some of those early solutions that were created as internal tools, you know, for organizations and, you know, really added value to the founders of those organizations, then they launched them commercially.
[MICHAEL]: Now, let me ask you, Peter, are there certain types of organizations that seem to see the most benefit from a solution like Servi?
[MICHAEL]: Like, are there certain industries or types of organizations that you've seen over the years?
[PETER]: Yes, we see a complete change and big, big changes in a B2B, in a complete B2B sector.
[PETER]: So because that's where prospecting really makes a difference.
[PETER]: That's where you can build a pipeline from different databases.
[PETER]: You can extract data, you can play around with it, you can, there are still companies using spreadsheets, downloading data and just making phone calls.
[PETER]: It's great and it still works.
[PETER]: I'm not against it at all.
[PETER]: There are better ways and there are more smart ways, but it still works.
[PETER]: So almost in any B2B sector where you would need to prospect, where you need to get the data of companies, searching for them.
[PETER]: OK, this is my company.
[PETER]: This is my ID client profile.
[PETER]: Let's go to the website.
[PETER]: Let's see what they do and let's give the cold call or let's send a cold email.
[PETER]: So anybody who does those kind of things, there's tools which we developed and can be developed.
[PETER]: It's very useful.
[PETER]: Renewable energy sector is where ESG and climate carbon consultants looking to connect with the ID clients.
[PETER]: Solar installers, especially commercial solar installers, very, very useful.
[PETER]: We have a particular product developed for, for example, solar installers, a Google map agent who will not just find companies, but find buildings on Google maps, identify those big buildings with big roof sizes, qualifying against the solar installer criteria and then the enrichment and finding out who the owner, who the building of the owner, owner of the building, who is the company, who is the decision maker and then sending emails.
[PETER]: So for those guys as well, for those industries is really, really useful to connect them with more ID clients, more warehouses and more high energy use companies because with the latest legislations in the UK and in EU and in the US as well, net zero goals, sustainability, ESG alignment, it will be really, really important.
[PETER]: It's not just about reducing your energy bills, but it's also your carbon footprint.
[PETER]: If you are part of a chain of bigger companies, bigger corporates, and you are not adhered to certain rules, then you will find yourself in a very disadvantaged position.
[PETER]: And this in 2025, 2026 going forward is going to be more relevant.
[PETER]: So in those sectors as well, really relevant to connect those guys with normal businesses, especially businesses on a high energy use to help them, to guide them through this process and make sure that the carbon footprint is reduced.
[PETER]: Now, I know a lot of organizations, as we're into 2026, are finding that, yes, with AI, they can eliminate a lot of the research and the cold outreach and all of that.
[PETER]: But once the organizations have been contacted, there's a lot of gatekeepers between the individual that you may want to be speaking to and that final bringing an account to board as a new subscriber or a client or what have you.
[MICHAEL]: So what are the steps after outreach that you found most effective for organizations that are doing well using your solution?
[MICHAEL]: It's a very, very good question, Michael.
[MICHAEL]: So is, shall I quote, quote or not?
[MICHAEL]: This is always the very first question from clients.
[MICHAEL]: And the short answer is no, you don't need to.
[MICHAEL]: If you have a right system in place, you can target the right people.
[PETER]: That's so critical because when I'm saying segmentation, before we set up a campaign, enrichment and segmentation, what I mean by that is there are XYZ company, Oasis, Michael is an owner.
[PETER]: But there are other, maybe we can't go from there, managing director, maybe CEO, CEO, maybe manager on a manager level as well, sales director, commercial director.
[PETER]: So when I'm saying segmentation, I mean this segmentation as well.
[PETER]: So not just segmentation within a particular sector or industry going down the niches, but it's also meaning going into different decision makers who are either we call the ultimate decision makers or influencers.
[PETER]: People who can influence decisions because they have the decision power to make certain decisions within a certain budget or they can go to the CEO, they can go to the owner and they can accelerate.
[PETER]: I don't know how many emails on a daily basis we're sending and receiving from, it's just literally coming back as forwarded to someone else within our organization.
[PETER]: If whatever you have to say and tell to your ideal clients is interesting, even if we find the wrong people within our organization, if there is something good in your offer and your service, internally, those people, if we target multiple people, if we target multiple people, they're going to be sending it to the right people and it's going to be in front of them.
[PETER]: So that's going to give an extra trust on your email because someone internally forwarded your email.
[PETER]: You will not just find a company and a website and an inquiry email and sending an inquiry or calling a central number, most of the time we can find and hit the right person or multiple persons within our organization and someone will be interested.
[PETER]: If whatever I have to say, you have to say and tell is interesting.
[PETER]: And also, that's why we're implementing LinkedIn as well, because we are targeting them on LinkedIn as well.
[PETER]: So you are messaging them on LinkedIn, hi, Mike, I sent you an email, I'm just following up on LinkedIn.
[PETER]: Is it a good time or this is what I have to say or whatever you have to say on LinkedIn.
[PETER]: So we do a very soft approach on LinkedIn.
[PETER]: But as you can see, multiple touchpoints within companies and multiple touchpoints with every single decision makers or influencers is going to be take you, put you in front of those people.
[PETER]: And even if you want to do a cold call because you need results quickly and you want to increase this point rate, when you make that phone call, most of the time we find mobile numbers or even if it's just an online number and you definitely want to call Mike.
[PETER]: You can say in a phone, hi, Mike, I sent you two emails, three emails, I follow up with you on LinkedIn and I just wanted to make sure that you receive my message, I just want to have a chat with you kind of thing.
[PETER]: And even if it's a gatekeeper, if you are a good caller and you definitely want to do a call, it's smoother to get into those.
[PETER]: So really, a lot of it comes down to crafting the right message so that when it does arrive in front of the right person, you're there.
[PETER]: And Peter, I really appreciate you taking a few minutes out with us before your presentation, which I am very much looking forward to.
[PETER]: Thank you very much.
[PETER]: Thank you very much for the invitation and for the opportunity as
[MICHAEL]: And you know, what struck me there was how you turned internal tools into a full-fledged business model. I'd love to hear more about how that evolved.
--- PRESENTATION ---
[PETER]: Our next speaker is Peter Jouvec, he is the founder and CEO of Cervi AI, an AI automation agency that builds service as software sales systems to help SMEs turn cold leads into predictable qualified pipeline and revenue.
[PETER]: Thank you Michael, thank you very much for the intro.
[PETER]: Hi I am Peter and today I will show you an example of what we call as an AI revenue engine and let's jump into it.
[PETER]: So we are here to discuss how SMEs can build enterprise level systems without the enterprise level budget.
[PETER]: So it's about leveraging AI to create what we call as an AI revenue engine.
[PETER]: This isn't just about the concept, it's a practical framework to achieving predictable and scalable pipeline.
[PETER]: We will explore how to move beyond fragmented sales efforts, build a cohesive high-performing system.
[PETER]: Focusing on smart integration, not particularly how we throw money to everything, but it's a rather strategic way of utilizing AI to implement all three strategies.
[PETER]: So my next thing is, let's address the elephant in the room.
[PETER]: The SME sales paradox, you need predictable revenue.
[PETER]: Just like large enterprises, but you operate with a lot smaller budget normally.
[PETER]: So enterprise level, they have the capacity and the budget to spend anything from 50 to 200 or even above 1000 pounds or dollars on a monthly basis or annual basis depending on the size of the company.
[PETER]: So SMEs normally and SMBs in the US are working on about 5-10% of the budget and it's literally impossible to achieve that sort of level of outreach with a predictable pipeline building and with all the automations necessary.
[PETER]: It would just take too many, too much human efforts or to expand the tools would be too expensive.
[PETER]: So it's not just about the cost challenge, it's about strategy as well.
[PETER]: Traditional sales tools often solve one problem, but they're creating another one.
[PETER]: And if you are just hiring a lot more business development or sales development representatives, then you again, more often than not, you're creating more problems than you actually solve.
[PETER]: So in the next slide, I'm showing the three very important elements of revenue engine, which is if they are working in isolation, which normally what companies and businesses do, they either do content alone or they do email alone or they do LinkedIn alone, neither of them are going to work.
[PETER]: So email itself, it's like spam most of the time, LinkedIn without email and content, just noise and content alone is not effective.
[PETER]: It takes too much time and too much effort and more often than not, companies are just going to give up because they expect the inbound leads, but content alone will not solve the problem.
[PETER]: Although it's a really, really important part of the revenue engine.
[PETER]: The next slide will show how there's different core elements we call working together as an integrated system with a compound effect.
[PETER]: And I think that's critical, cannot emphasise enough how important that these different elements actually working together as a whole system.
[PETER]: So the revenue engine as a concept is a strategy, it's built around outreach, which is email and LinkedIn.
[PETER]: Email is great, but with LinkedIn outreach, we essentially accelerate the outreach efforts.
[PETER]: So if you are sending an email and following up with someone on LinkedIn, it's more touch points, it's build trust and it's more opportunity for a positive response.
[PETER]: On the other side is build long-term authority, positioning you and your company as an expert on the field.
[PETER]: It builds trust, it builds authority and eventually is going to be accelerating all the outreach efforts, which you will be doing on email and LinkedIn, plus in itself is going to be start to generate inbound leads because how LinkedIn works, as we are reaching out to the particular ICP ideal client, LinkedIn algorithm prefers to show your content as well towards more to the similar people from the similar group.
[PETER]: So that's how you're going to start to generate, after a certain time, leads from people you haven't even targeted before, but they are in a similar ICP because algorithm is going to start showing content for them.
[PETER]: And obviously we can go, I could go more on how we combine this with SEO website as well.
[PETER]: So content blocks and in the latest strategies, answer engine optimization and generative engine optimization.
[PETER]: Well, I don't want to go into that today, but that's something is also very important.
[PETER]: So how people can find you on your website, how they can find you by the AI search engines, like chat GPTs and perplexities, if they are searching the type of service you are looking for.
[PETER]: So it's more about accelerating the outreach efforts because they're positioning you as a thought leader, but it's also in itself content generate inbound leads on LinkedIn, from LinkedIn and also from other sources as well, if you want.
[PETER]: Now there is another thing, which is a lot of, it's a very new, not new, completely new, but for SMEs and SMBs is quite new, is voice agent.
[PETER]: So how voice agent technology can help to implement in integrating a part of the team, not replacing the human teams.
[PETER]: It's very, very important.
[PETER]: Voice agents, AI agents, the purpose of them is different from what we humans need and supposed to do and your team supposed to be doing.
[PETER]: But everything else is automated already.
[PETER]: As this point is coming back, qualifying all these points before the human team get involved, it's so critical.
[PETER]: So your best people not wasting time on unqualified leads, because let's be honest, are you generating leads?
[PETER]: Probably 70, 80% of them will be either time wasted or not really ready, depending on which stage of the funnel they are.
[PETER]: And that's where voice agents can save you a lot of stuff.
[PETER]: So we're using it for inbound leads, qualifying them by email, even by calls.
[PETER]: And also in a nurturing phase as well, once we had a call with a client or we sent a proposal, then as per the nurturing sequence or lead scoring sequence, voice agents can make outbound calls as well to try to bring back clients to the second call to close them.
[PETER]: So that's the voice agent technology.
[PETER]: That's how it accelerates sales and also saving human teams from a lot of efforts.
[PETER]: Now the next one is the intelligent outreach.
[PETER]: So we're going back to how we're executing outreach.
[PETER]: So anybody can set up an email campaign.
[PETER]: Anybody can set up a LinkedIn campaign manually, or there are tools available from instantly for emails and you know, there's many, many other LEM lists, there's many other tools.
[PETER]: You can do email LinkedIn campaigns.
[PETER]: So that's not a hard thing.
[PETER]: But a really magic happens is then the precision targeting, which is so important.
[PETER]: But in order for you to do that, you need, you want to target 1,000, 10,000, 100,000 of companies, B2B companies, and you would need to research every single one of them manually to find the companies, the pay points, who they are selling to.
[PETER]: It's really important because the ICP of the ICP.
[PETER]: So you need to find out who they are selling to.
[PETER]: And based on all those information, and also based on the segmentation, you're going to fall out of segmentation, which means that company size, the sector, the niche, in a lot of industries, there are multiple niches, okay, what those companies actually specialize on, if there is a specialized niche they are targeting, company size, title, within a title, the company's multiple titles, you would send a different message for a CEO of a 200 company from a company has 30 employees or 10 employees and the founder led company.
[PETER]: So the message, the angle, it's different, the tone is different, how the copywriter agents write the copies.
[PETER]: And that's where the precision targeting and enrichment come into place.
[PETER]: So before we send a single email, or even before we write a single copy, we just need to make sure that we know who we write and what we write to.
[PETER]: So this is where, before email, this is really important because enrichment process, which is a precision targeting, and then the orchestrated execution, but after segmentation, we're scheduling them to email and then we're scheduling them for LinkedIn as well.
[PETER]: So everybody who we send an email, we're following up with them in a predetermined sequence, following up with them on LinkedIn as well to increase touch points, increase trust.
[PETER]: And in either scenario, we're also developing content, which is resonating with the audience.
[PETER]: So your ideal client, let's say you are selling, you are, I don't know, a commercial, be selling a commercial lead generation for commercial solar installers, you're obviously going to be talking about how we're helping the commercial solar installers, for example, finding those high energy buildings and warehouses, for example.
[PETER]: So the content we're developing is very relevant to your ICP we're targeting to.
[PETER]: Personalization, when we're writing, our copyright agent writes the emails, other than all the information and all the segmentation, which I mentioned, it's very important that they are using it to get all the information about the company, about the person they are sending an email, everything which is available online or website or company, LinkedIn bio, the copyright agents we use, if you find a good angle.
[PETER]: And this is how we personalize every single copy, and we'll be sending it to businesses and different people.
[PETER]: Very important when it comes to how the AI agents using information and how they know if we have, let's say, set a multiple four or five different ICPs, how the AI agent will know what my ICP, who they are selling to, or how it knows about all the information and how it knows about what kind of tone they want to write an email.
[PETER]: So every business and every company and everybody who will be working with all our clients, for example, they all have the different view on the tone of the email should come across their style, their writing style, and everything else is so critical, you want to represent your brand.
[PETER]: So you don't want some kind of AI agents completely goes out of context and write some gibberish.
[PETER]: So you want to make sure that whatever you send to your clients is representing you, representing your brand and your company.
[PETER]: And this is so important contextual retrieval.
[PETER]: That's why we are using, for example, vector databases.
[PETER]: So we have different playbooks, scripts, best performing templates for different sectors and niches.
[PETER]: And when we have, for example, a client who wants to target, let's say, warehouses, manufacturing companies, then our agents knows exactly what kind of playbooks and what kind of, who's the client, so what's the style and tone of the client wants.
[PETER]: On a sector, it knows what kind of playbook he needs to pull.
[PETER]: And all this information is happening almost like real time and can write 10, 15 different variation of copies for the same campaign and using different offers, different style, different angles.
[PETER]: And as we are sending the first couple of hundred, couple of thousand, this is why we always do the first three months is always testing what works and what doesn't.
[PETER]: If something doesn't work, we completely remove from for that client and the things are working, we double down on whatever it works.
[PETER]: And because data is keep coming back and the copywriter agent knows which angle, which copy, which subject line was working, which style was working, and then it's going to be start to write more and more and more on that style and with that angle.
[PETER]: So the responses are increasing more and more and more all the time as we're doing it.
[PETER]: This is why analytics is so important and this is why vector database is so important because all the data from previous companies, best performing templates and everything else is available for copywriter agent.
[PETER]: The next one is the content, which is really relevant for the content.
[PETER]: How content fly because the content flywheel is essentially accelerating your outreach efforts.
[PETER]: So content positioning you as an authority and we often come across clients making comments, I just want results.
[PETER]: I just want you, for example, to send me X, Y, Z emails and just generate X, Y, Z leads.
[PETER]: It's so important to understand if you're executing an outreach campaign with an agency or whether yourself, but position yourself to your market, to your target market and write content at least two or three times if you can do more, which positioning you, which helping them first of all, whatever you want them to have it, it's going to be a helpful content for them and it's also going to be positioning you as an expert in your field and whatever emails you're going to be sending or whatever LinkedIn message you're going to be sending, it will be resonating a lot more with them as well.
[PETER]: So as you can see in there, this effect is so important.
[PETER]: So as exponentially growing your authority and normally this is the exact same exponential growth, which we predicting in terms of your leads of the number of leads you are generating from your outreach efforts.
[PETER]: And also really important is not just from your outreach effort, but also it's going to be start to generate leads from inbound efforts as well, which I can show you in the next one.
[PETER]: So inbound, warm leads generated from outreach and content pillars.
[PETER]: It is so critical to understand that content is going to be accelerating your outreach efforts and in itself as well, will drive and bring warm inbound leads as well.
[PETER]: And that is going to be the voice qualification.
[PETER]: Once you have leads and leads coming in and you literally can't handle it anymore, then a voice agents can handle it, can qualify and access the fit, urgency, normally service fit, budget, urgency fit and all the rest.
[PETER]: It absolutely can be done with an AI agent.
[PETER]: And once the conversation and the qualification is successful, then it can book a call with your sales team, with yourself, however you want to do it.
[PETER]: Then the next one is again, as it illustrates the full revenue engine effect.
[PETER]: So how each rotation builds on unstoppable momentum.
[PETER]: Outreach creates those initial touch points and fills your immediate pipeline.
[PETER]: Content then builds credibility for those touch points, making your outreach far more effective than it would have been otherwise.
[PETER]: Finally, the voice agent qualifies those engaged leads, ensuring your human team only speaks to truly ready prospects.
[PETER]: As the cycle repeats, your authority grows, your outreach becomes even more effective and your inbound lead generation becomes even more effective as well.
[PETER]: So your entire system just keeps accelerating and accelerating.
[PETER]: By achieving this requires avoiding some common pitfalls as well.
[PETER]: Normally the biggest problem we see people, they're expecting absolute unrealistic expectations and miracles within four weeks, two weeks sometimes, this revenue engine can bring results in short time, after four weeks, eight weeks, 12 weeks, continuously.
[PETER]: But the first 12 weeks, we always say it's testing.
[PETER]: It's bringing results already and you already have leads.
[PETER]: You can already sell closed deals, but it's normally testing what works and what doesn't.
[PETER]: And after the initial three months, it's scale.
[PETER]: How big you can scale is depending on your ICP, your ID client profile, depending on the sector and the niche you are targeting, is it a highly fragmented, is not much fragmented, but the total addressable market, and also what's your budget, what's your goal, how many people, how many prospects you can serve, we can, you know, there are certain sectors and niches and you can have five, 10, four calls per day, if you want.
[PETER]: If you want to go completely big and crazy, or if you have just one need, one or two calls per day, then you have just one or two calls per day, and then you close accordingly.
[PETER]: Not everybody is going to be closed, we all know, but this is why it's so important to build up the momentum, build up the leads, build up your pipeline.
[PETER]: And once you feel, okay, what's your closing rate, then it can be implemented, voice agent can be implemented, who helps you and your team to be even more effective.
[PETER]: And I think that's in a nutshell, so that's what it's about.
[PETER]: If obviously any questions, we're always happy to help, but normally that's the most important thing.
[MICHAEL]: So, this is the live panel that Peter was part of at our recent Software Oasis live AI event. These are the real questions asked by the audience in the room.
--- Q&A PANEL ---
[PETER]: Okay, I hope everyone has been enjoying the event so far.
[PETER]: I know we've had a lot of questions come in from the live audience, and I see that we have a lot of organizations represented from all around the world.
[PETER]: So, we'll try to get to as many of these questions as we can over the next 25 minutes.
[PETER]: I'll start by introducing the panel of speakers that you've heard over the previous few hours.
[PETER]: We have joining us from RegPAC, the founder and CEO, Asaf Taras.
[PETER]: Next, we have joining us from CHERI, the Global Head of Corporate Development, Kevin Stoffman.
[PETER]: And finally, we have the CEO of EQAI, Harsha Mokarla, and he is joining us live as well.
[PETER]: So, as we're getting started, I think what better place to start than with a question from Liam, who is joining us live from Chicago, Illinois, and his question is for Asaf.
[MICHAEL]: The question is, for a services business that's terrified of losing its human touch, what's a low-risk first workflow you might automate so the team feels relief from the grunt work without freaking out about robots replacing them?
[MICHAEL]: Okay, so first of all, thank you for the question.
[MICHAEL]: It's a great question.
[PETER]: I think the most important thing to understand about automation is if you do it in a humanized way, you're actually going to be more connected to your clients rather than less connected.
[PETER]: The basic things that you can automate right away is confirmation emails, approvals, things like that.
[PETER]: Think that someone is registering and selecting a specific course or specific activity that they want.
[PETER]: An email goes out or an SMS or both that tells them, yes, you have registered for this.
[PETER]: Once a payment goes out, a receipt goes out and tells them, yes, you have completed the process.
[PETER]: You have paid and you're enrolled.
[PETER]: Let's say they can't enroll and they need to do the same thing, notify them that they're on the wait list.
[PETER]: Then after all of that, once they complete the process, tell them you've completed the process, everything's fine.
[PETER]: This will eliminate a lot of back and forth for your team.
[PETER]: And then any specific actions that you do, you make it so that you do the actions in the system and all the notification and all the communication with the client happens automatically.
[PETER]: That way you're basically creating an automation of processes and automation of communication with your clients and making sure that you're giving the same level of service to every single person.
[PETER]: And it's not based on someone remembering to do something or not remembering to do something.
[PETER]: I think that's the easiest, most frictionless way to start automation.
[PETER]: And from there, you can go a lot.
[PETER]: You can continue to much higher levels where specific situations just create a game plan that goes into place.
[PETER]: But I would start just with that.
[PETER]: And then your team, once they see how much less back and forth they have, they will embrace it themselves.
[PETER]: You won't need to push them towards it.
[PETER]: I think that that makes a lot of sense and it sounds like some really easy starts to take a transition into automation.
[PETER]: The next question came in from Vancouver, Canada, from Sophia, and this is for Harsha.
[MICHAEL]: Sophia is asking, how do you convince a leadership team that a single blended CAC number is dangerous and that they need segmented, margin-aware CAC without overwhelming them with complexity?
[MICHAEL]: So it's a great question.
[MICHAEL]: And in particular, a fairly relevant question if you work with stakeholders who are not the most financially savvy or don't think about numbers day in, day out, that kind of thing, for example.
[PETER]: For me, I think the best ways in which this has worked is by genuinely tying it to some kind of financial or business outcome, if you will.
[PETER]: The thing that most leadership, most leaders at most businesses understand is their P&L, is the sort of bottom line, top line parts of their business in terms of what comes through.
[PETER]: And the moment you can demonstrate that all activity, all marketing activity is not created equal in that, there are parts of the business, there are parts of that activity that are generating or contributing less to your P&L than other parts, for example, and that there is room for improvement there towards people meeting their goals, that tends to go a long, long way.
[PETER]: The more these concepts like cost of acquisition or ROAS or any of these kinds of constructs get stuck in marketing speak, the less likely you are to get leadership to understand what it is, the more you can tie it back to contribution margin, the actual sort of EBITDA or financial metrics that a business cares about, the more likely you are to succeed in getting people to care about it.
[PETER]: And so what I would do is go talk to your again, your friendly neighborhood finance person and say, hey, how do I translate my concepts around say marginal tax or return on advertising spend or those kinds of things into a financial metric that can be reported out along with the rest of the financial metrics the company cares about.
[PETER]: And the more you're able to do that, the more you tend to get buy-in interaction.
[PETER]: And the last thing I'll say is, I don't know what Sophia's job function is, but if you're a marketer and you're in a position where you are actually talking about this and trying to convince your leadership that this is something that they should care about, kudos to you, Sophia.
[PETER]: You're already ahead of 90% of the pack in terms of folks who actually think about these problems and challenges that way.
[PETER]: So kudos to you for even attempting to go down that path.
[PETER]: Most marketers don't like going that far with these metrics.
[PETER]: Michael, I think you're on mute.
[PETER]: Thank you for that thoughtful answer.
[PETER]: This next question I really like, it came in for Kevin and let me just read this.
[PETER]: This is from Imogen from London over in the UK.
[MICHAEL]: And Imogen is asking, what's a practical first step for a bid-sized real estate firm to build a shared property hierarchy before they even think about layering AI on top of it?
[MICHAEL]: Sure.
[PETER]: First, I would just do a quick add-on to the answer that Harsha provided to Sophia.
[PETER]: CAC in and of itself is almost a worthless statistic unless you understand the relevant lifetime value you're going to get from the same customer you're acquiring.
[PETER]: So a large amount of money you're spending acquiring said customer is fine if the lifetime value that same customer provides in revenue is massively larger.
[PETER]: The same way, a small CAC may not be that great if you're not going to get much revenue from that same customer.
[PETER]: So always be thinking of those two statistics in tandem of how much it's going to cost to attain and then retain the customer and then what type of revenue you're going to get in tandem.
[PETER]: And that's regardless of what industry you sit in.
[MICHAEL]: Now, as to Imogen's question, in the world of real estate, you could take that same question and think, where do I get revenue in my business?
[MICHAEL]: Is it from recurring revenue on rental income?
[MICHAEL]: Is it fee income for certain amenities and bookings and events?
[MICHAEL]: And depending on whether your revenue is time bound or recurring, that will prioritize what data is critical to your business.
[MICHAEL]: And once you've determined what the critical data in your business is, you need to then have an agreement across your midsize firm what the definitions of those data points mean.
[MICHAEL]: What does property actually mean?
[MICHAEL]: What does asset actually mean?
[MICHAEL]: What does unit actually mean?
[MICHAEL]: What does lease or let actually mean?
[PETER]: She's in the UK and we call it leasing over here, she calls it letting over there.
[PETER]: We call them brokers over here, she calls them agents over there.
[PETER]: If you're in a multi-regional business or you're in a multi-property business, you need to have standard definitions because when you record all of that data without standard definitions, if you put AI on top and you tell the AI agents to go prioritize work orders or execute payments to vendors based on prioritization or recommend rental rates for a potential tenant, it might get those recommendations wrong if you don't have a standard data model to work from.
[PETER]: So the first step is just agree on what is critical in data elements to the business and then standardize the terminology across all of the departments in the business.
[PETER]: And when I say that, I mean people who are responsible for finding the deals, people who are responsible for underwriting the deals, people who are responsible for closing the deals, and then people who are responsible for managing and operating those deals once you've bought them.
[PETER]: And then lastly, the people who will then advertise how you've performed on those deals to go raise more money to do it all over again.
[PETER]: Makes a lot of sense and appreciate you're jumping in on both of those questions.
[PETER]: I think that adds a lot of value.
[PETER]: The next question came to us from Toronto, Canada for Asaf, and this is from Arav in Toronto.
[MICHAEL]: And the question is, when you start automating the common case, how do you explain to staff that automation is there to remove the grunt work, not to quietly eliminate their jobs?
[MICHAEL]: Okay.
[PETER]: I think that's less a question about automation and more a question about company culture, I would say.
[PETER]: Any new addition that you add to your company can be viewed by your employees as a threat.
[PETER]: But at the same time, it can be viewed by your employees as a possible solution or improvement to your business.
[PETER]: So I would say that this has little to do with automation, more to do with culture.
[PETER]: You need to create a culture in your company where people understand that their knowledge, their ability is valuable, that computers cannot replace people.
[PETER]: I'm sure Harsha can talk about how AI cannot replace human understanding of data and the assumptions that you bring to it, which will decide what output you'll get.
[PETER]: So the human ability is much more than just the grunt work.
[PETER]: If your employees think today that that's what they are, people that just do grunt work, you have a bigger problem than automation.
[PETER]: But if you manage to give them the sense that they are valuable, they are people that can think, they are people that can create, they are people that can cause things to happen, then they will embrace any new technology that will come up by any new feature or ability that you will create.
[PETER]: Perfect.
[PETER]: And the next question came in right here from Grace, who is in Boston, Massachusetts, and it's for Harsha.
[MICHAEL]: The question is, when every ad platform reports amazing tech, what's one practical way to reconcile the numbers so the math actually matches?
[MICHAEL]: Thank you for the question, Gershon.
[MICHAEL]: This is again, in some ways, it's a nice call out to Asaf's point about humans and grunt work and that kind of thing, for example.
[PETER]: My recommendation, Grace, is to don't try to make that math math.
[PETER]: The right answer here is to ignore what the platform tells you in terms of what tech is and to create what is a true actual reality picture based on your data for it.
[PETER]: As I talk about often, the platforms have a role to provide or data to provide in terms of cost metrics, how much money are we spending by campaign, by geography, by tactic, all of that kind of thing.
[PETER]: Typically, the campaigns do a poor job of reporting on the outcome metrics in terms of conversions and sales and revenue and those kinds of things, for example.
[PETER]: My suggestion largely is to ignore the platforms in terms of that and create the structure for yourself where you have the source of truth in terms of those customer and revenue metrics and use the platforms for the role that they're intended to play, which is around cost and spend and platform driven metrics and then use your reality to calculate those metrics, if you will.
[PETER]: So I wouldn't even spend time trying to reconcile the two.
[PETER]: I would try to create an abstraction and a source of truth that is your own and that is not controlled by the platforms.
[PETER]: As I've said before, the platforms have a vested interest in making those numbers look to their advantage, and you should not trust that.
[PETER]: You should trust your own data for that purpose.
[PETER]: And again, thanks to automation, thanks to all these tools, this is now well within reach for everyone, for companies large and small, the ability to put that source of truth together is right accessible and right handy today.
[PETER]: Perfect.
[PETER]: And the next question came in, and this actually came in a few times from several different individuals at the event today.
[PETER]: So I'm going to pick this one from Zara, who's over in Edinburgh, over in the UK.
[PETER]: And this question is for Peter from Survey.
[MICHAEL]: The question is, how do you keep an AI revenue engine focused on pipeline quality, not just blasting out more generic outreach faster?
[MICHAEL]: Hi, everyone.
[MICHAEL]: Thank you for the question.
[MICHAEL]: Well, thank you.
[MICHAEL]: Very good question.
[MICHAEL]: And it's a really important one as well.
[PETER]: Targeting is very critical and crucial, how we're targeting when we do outreach.
[PETER]: And it all comes down with us when we build a revenue engine.
[PETER]: It all comes down to strategy first and really nailing down on ICP, ID client profiles.
[PETER]: I mean, there's so much thing we can nail down and segment even within certain profiles.
[PETER]: We can do four, five, six levels of segmentation to really nail down who is my ID client and then how we could target them.
[PETER]: So it's really nailing down segmentation strategy and ID client profile to make sure that when we do a bunch of hundreds, thousands, hundred thousand, it doesn't matter how many, email, that each of them should be landing in the right people inbox and be the right message as well, which is so important.
[PETER]: That's the whole enrichment and whole segmentation comes down to.
[PETER]: Fantastic.
[PETER]: And Peter, I see quite a few questions that came in from you.
[PETER]: So I wanted to, while you're with us right now, share a second question.
[PETER]: Noah, who is here in the U.S.
[PETER]: in New York City, has a question.
[MICHAEL]: When you connect inboxes, CRM and content, what are the biggest data quality traps that can quietly poison an AI revenue engine?
[MICHAEL]: So then, depending on the desktop you're using and what automations you're using and all the rest, yeah, it's very critical.
[PETER]: I mean, you definitely don't want data which is going into the wrong place or, you know, any wrong data going to the wrong place and all the rest.
[PETER]: That's the question.
[PETER]: So depending on the stack, the stuff you use, the CRM you use and the whole system, the whole, the reason why we're calling it a revenue engine, because you're connecting multiple things to the CRM.
[PETER]: CRM for us is always the central hub.
[PETER]: But for data, we're using something different.
[PETER]: So the answer is really depending on, you know, what system you are using.
[PETER]: But, yeah, it's really, really critical that there is only one source of truth always when it comes to data, whether it's your CRM, whether it's your, you know, some other database, Postgres database or whatever you're using.
[PETER]: But there's only one source of truth.
[PETER]: And then that's how you manage across your ecosystem.
[PETER]: So we're using four or five different systems and you're always one source of truth when it comes to data.
[PETER]: And this next question came in for Kevin, and I love this question.
[PETER]: This is from Carter, who's in Dallas, Texas.
[MICHAEL]: And the question is, when you standardize property data definitions, how do you keep power users on board who feel their local naming conventions are special and can't be changed?
[PETER]: Carter, I'm also a Dallasite.
[PETER]: So ping me after this.
[PETER]: I'll share some of my longer opinions over a cocktail.
[PETER]: You don't have to sacrifice any local market knowledge in what you would consider some sort of proprietary competitive advantage in the way you define something without also realizing that if you're in multiple regions and you're across multiple property types, you might be in multiple fund structures or fund mandates.
[MICHAEL]: But once everything gets up to the corporate level for reporting, there needs to be a standard, right?
[MICHAEL]: And so if you want to raise a bunch of capital and you want to satisfy the needs of your LPs and the cost to provide information to regulators, getting to that standard quickly and efficiently is paramount.
[MICHAEL]: So you may have something unique in the Dallas market around zoning or permitting or identifying foot traffic or demographics, but that shouldn't change the definition of what lead to lease means or what economic occupancy actually means.
[PETER]: I see this a lot in multifamily.
[MICHAEL]: Someone says, well, what's occupancy?
[PETER]: I was just at the city REIT CEO conference where we talked to publicly traded REITs and they're all reporting average occupancy.
[MICHAEL]: And some would say, well, what does that mean?
[MICHAEL]: Does that include the model units, the down units, anything undergoing renovation, properties that are under construction and that have started lease up, but they're not ready.
[MICHAEL]: So it's true that there's nuance there, but you got to get to a standard because you're going to do public reporting.
[MICHAEL]: And someone might futz with the data to make themselves look good, but getting to a standard puts everyone else in the same playing field.
[MICHAEL]: It just makes you outperform.
[MICHAEL]: The next question came in from Manchester over in the UK from Chloe.
[MICHAEL]: And this question is for Asaf.
[MICHAEL]: How do you balance automated payment plans and eligibility checks with the need to handle messy real world family situations in enrollment?
[MICHAEL]: Okay.
[PETER]: So the way that we build it in RegPacket, and I think that is the correct way to do it, is again, regarding the previous question about the common case.
[PETER]: We have an algorithm that calculates the risk of the user and the just-in-time funding for the organization.
[PETER]: And then it offers a payment plan or an installment plan for that user based on those factors.
[PETER]: So let's say someone gets the option to pay three, six, and nine installments, and they pick one of them.
[PETER]: Now, life happens and suddenly in the third installment, they can't pay or something, or it's rejected.
[PETER]: So you have two basic ways to take this, either the automated way or the not automated way.
[PETER]: The automated way is that the system will try multiple times until it will basically give up on that specific installment.
[PETER]: And then it will add that installment to the next installment.
[PETER]: Let's say the assumption is that they don't have money for the third installment because of, I don't know, they had some kind of expense that they needed to do.
[PETER]: It fails, and on the fourth installment, it will basically take the third and the fourth installment.
[PETER]: So that's like a good way to just automate it and not think about it.
[MICHAEL]: But if you want to be more hands-on and communicate and connect with those families, you can connect with the family and tell them, okay, what can we do?
[PETER]: You have the ability, and this is something that I believe that all automated systems have.
[PETER]: They need to be automated, but at the same time, allow all the administrators to always overwrite stuff.
[PETER]: So you can take them off the plan, put them on a personal plan.
[PETER]: You can pause the plan.
[PETER]: You can basically do anything you want.
[PETER]: Now, if you have a system, and this is where automation gets a bad name for itself, if you have a system that does not allow people to override the automation, that's where your problem is.
[PETER]: You always should be able to come in as a human and trump the automation, say, no, stop.
[PETER]: Now I want to do something else, or I want to do something manual.
[PETER]: And if the system doesn't allow you to do that, that's the problem, not the automation.
[PETER]: Now, today, from my knowledge, most automated systems or most high-level automated systems allow that.
[MICHAEL]: And if you have, especially in payments, because people are super touchy about payments, and rightfully so, right?
[MICHAEL]: So if that's not the type of system that you're working with, then you should replace it, because you should have that ability.
[PETER]: So this next question, I think, touches on something that a lot of organizations are dealing with in the audience.
[PETER]: And this is from Oliver over in Dublin, Ireland.
[PETER]: It's for Harsha.
[MICHAEL]: The question is, how do you approach CAC when you're experimenting with AI-driven channels where attribution windows and touch patterns don't fit your old model at all?
[MICHAEL]: It's a great question, right?
[PETER]: And I know it's framed in the sort of AI and that sort of thing.
[MICHAEL]: But in general, when you're experimenting with new tactics, when you're experimenting with new channels, how do you think about CAC in relation to those?
[MICHAEL]: What do you think about CAC when, say, for example, you're doing connected TV or channels that are non-traditional, different than some typical sort of framework that you've actually built?
[MICHAEL]: But first, it's important to acknowledge and understand what the dynamics of sort of the customer experience within that channel look like, right?
[MICHAEL]: So is this a channel that drives direct sort of click-based traffic?
[MICHAEL]: Is this a channel that has sort of second-order impacts in terms of impressions and brand impacts and those kinds of things, for example?
[MICHAEL]: Important to understand sort of what is typical customer behavior on these channels.
[MICHAEL]: Important to understand how you expect that channel to actually produce value for you and where you expect it to produce value for you.
[MICHAEL]: Does it actually impact things like how many people are searching for my brand, for example, right?
[MICHAEL]: Does it show up in how many people are searching for my product?
[MICHAEL]: Does it show up in direct traffic?
[MICHAEL]: Does it show up in phone calls and things like that sort of in my business, for example?
[MICHAEL]: And it's important to build your experimentation in light of that, in that sort of your data is set up to read that.
[MICHAEL]: Your experimentation is set up to read that, right?
[MICHAEL]: Mechanisms like match markets, mechanisms like pulsing spend on and off, for example, all of these are, you know, and many, many more methodologies like that are sort of what you should approach it with in terms of understanding, hey, when I go into that new experimental channel that, you know, I expect to drive customers a different way than what I'm actually used to, for example, am I set up to be able to read that in a fair test versus control kind of manner, right?
[MICHAEL]: And this kind of harkens back to the question we had earlier as well about, you know, how do you sort of measure that from the platforms and things like that?
[MICHAEL]: This is exactly why building the data infrastructure to where you can actually read all of this at that type of granularity, you can actually understand the second order impacts of, you know, increasing activity in channel X over channel Y.
[MICHAEL]: It's exactly why those things become that important as well.
[MICHAEL]: Setting up that data infrastructure becomes that important as well.
[MICHAEL]: So again, it's not, there is not a sort of, you know, straightforward cookie cutter answer for this.
[PETER]: I think that measurement strategy of how you'd measure CAC on some new experimental channel should be derived from the experiment itself.
[PETER]: The experiment should contemplate what my CAC measurement for that sort of channel or experiment look like.
[MICHAEL]: And the last point I'll make is there's also a question of thresholding within that, which is to say, hey, what do I expect, you know, this new channel to perform like, right?
[MICHAEL]: Obviously when you enter a new channel, there's room for experimentation and learning, which means, you know, it won't be efficient right off the bat.
[MICHAEL]: You'll have to learn and optimize your way into it.
[MICHAEL]: Therefore, again, making sure you're setting yourself up to the right thresholds for, hey, how much money am I going to spend on this channel?
[MICHAEL]: What am I, you know, what am I, what am I comfortable losing, if you will, how inefficient am I okay being while I go through a learning phase of that channel?
[MICHAEL]: That's important to establish upfront as well while you're experimenting with this.
[MICHAEL]: And AI or not, these have been principles that, you know, are true for new channel entry, new market entry for, you know, for as long as there's been marketing innovation.
[PETER]: So as we're coming to the close of this Q&A panel, I have two final questions.
[PETER]: One is for Peter.
[PETER]: Peter, this is coming to us from Freya over in Brisbane, Australia.
[PETER]: I've been there, a fantastic area, lots of amazing wineries.
[MICHAEL]: What leading indicators tell you your AI workflows are actually compounding into a healthier pipeline rather than just creating busy work and noise?
[MICHAEL]: And that's for Peter.
[MICHAEL]: Okay, yeah, thanks for the question.
[MICHAEL]: Very, very, very good question.
[MICHAEL]: So it all comes down to how you, again, how you put that revenue engine to the AI revenue engine together because she's very right.
[MICHAEL]: She's right.
[MICHAEL]: So it's very, very important to understand so it's very, very important to how you put the whole system together and how you use, do you use analytics in the first place?
[MICHAEL]: So how you measure the different attribution, the different data coming from different sources, depending on what you use the revenue engine for.
[MICHAEL]: So whether it's email, whether it's LinkedIn outreach, whether it's the content itself, measuring engagement and all the rest.
[PETER]: So depending on the activities which is going and the type of channels we are measuring and the data we're getting back is going to tell you whether you are in the right place or not.
[PETER]: It can be measuring the availability if you're talking about email.
[PETER]: It can be, again, connection requests if you're talking about LinkedIn, can be engagement followers.
[PETER]: Depending on what you want to measure, the most important thing is which we want to achieve, at least for us, is what kind of end results we can get in terms of our pipeline, healthy pipeline, those people who are engaging with us versus the effort going into the first place.
[PETER]: I mean, how many people we need for that and what kind of, what sort of activity is going into that.
[PETER]: And if you measure it in the right way and you measure all the channels and all the activities, then it should tell you that whether you're just generating leads and keeping yourself like a busy fool or that pipeline is really healthy and people actually engaging with you and that's going to be, can be measured by on your closing date.
[PETER]: Perfect.
[PETER]: And Kevin, I'll toss this final question out to you.
[PETER]: This is from Amelia over in Auckland, New Zealand.
[MICHAEL]: And Amelia's question is, for portfolios spread out across several countries, how do you design a data model that respects local quirks without losing global comparability?
[MICHAEL]: Sure.
[PETER]: Amelia, the first thing to do is what I mentioned is first standardized definitions.
[MICHAEL]: So it gets a little more complicated when you're going to think about what is assets under management mean, right?
[MICHAEL]: You need to now convert local currencies into whatever you're going to report out to your investors in.
[MICHAEL]: If HQ is in Auckland, you know, you'll report in that local currency and you'll convert everything else.
[MICHAEL]: You might have a different lease flow in the way you report what is a lease in New Zealand versus what might be happening across the pond in Sydney or what's happening to the Northwest in Singapore or what's happening in Japan.
[MICHAEL]: But at the end of the day, it's still about what revenue are you tracking that's repeatable from tenants?
[MICHAEL]: What revenue are you tracking that's from event based things like, you know, booking amenity space or what is capital budget spending versus what is recurring operating expenses?
[MICHAEL]: You can still get to those standards, same P&L like terms.
[MICHAEL]: So on the financial side, it's much easier to get to a standard where you're going to run into more nuance is on the operational side.
[MICHAEL]: So why are we outperforming or underperforming a benchmark might be very different in certain regions because operating metrics are different in different regions, right?
[MICHAEL]: You might interact with brokers or agents in different ways in different regions.
[MICHAEL]: You might leverage working with local municipalities in the way you get permitting or zoning approval in very different ways in different regions.
[MICHAEL]: So all of the nuance is going to happen on the operating side, and you should continue to respect that while understanding that the standard for global reporting should be done on the financial side.
[PETER]: I would like to thank everyone for their time and presentations today.
[PETER]: And for anybody that's listening live that would like to reach out to Asaf, Harsha, Kevin, Peter, or any of the other speakers that you've heard during the event today, that big orange button at the top of the site will get you directly in contact.
[PETER]: And if you're part of the community, you can reach out you can reach out as well.
[PETER]: So I'd like to thank everybody for their time on the panel today.
[PETER]: Thank you so much.
[PETER]: Appreciate it so much.
[PETER]: Thank you.
[MICHAEL]: Wow, that was such an insightful conversation with Peter. I really appreciated how he shared the journey from losing everything to creating Syrvi AI, and the importance of precision targeting in B2B sales. It was really great having you on, Peter. Thanks for sharing your story. If you enjoyed the episode, make sure to subscribe to Software Spotlight for more stories like this. Until next time!