Sales Calls, Call Tracking, & Machine Learning Tech with Mo Asady of Convirza

In this edition of the UpTech Report, we meet with Mo Asady, the CTO of Convirza, to learn about how marketing teams are tracking and evaluating sales calls using AI.

If you only have one or two sales calls to make, it’s easy to evaluate what happened after the call. However, if your team needs to examine the trends across hundreds of hours worth of conversations, artificial intelligence is likely a better solution.

Through speech recognition, natural language processing, and ultimately, machine learning algorithms, Convirza’s tech is learning about how your team does business. Then, it’s analyzing the calls, providing actual scores about lead quality and a team’s phone skills.

In the near future, this call tracking technology will even be able to help guide the sales conversation itself, giving the salesperson specific directions about which parts of the script to hit next.

It’s not just about going down a sales checklist, it’s about understanding exactly what is going on in the conversation, in real-time, and then using this information to make data-driven marketing decisions.

Mo Asady is a technology executive with over 20 years of tech leadership experience. He started his career as a full stack developer, progressing to technical architecture, technical project management and finally as a seasoned CTO.

He has worked across different industries, starting at Microsoft where he led teams within the Dynamics CRM suite. Moving to OTT and video streaming in Sweden, later he introduced digital solutions to the logistics and supply chain industry both in the US and in Dubai.

Today he is the CTO of Convirza, changing the way call tracking and AI aids advertisers to better understand, plan and execute their marketing campaigns.

Mo is a husband and father of two, he lives in Orlando, FL.

Convirza is the first enterprise call tracking and call optimization platform. To address the growing demand for intelligence from phone interactions we combine traditional call tracking with the powerful analysis of phone calls. Convirza uses speech recognition technology and sophisticated algorithms to gauge lead quality, measure conversions and take action with fully-baked marketing automation. We help sophisticated marketers make better decisions, improve close rates and customer experience while increasing revenue. 

The company serves leading national and international brands in automotive, healthcare, hospitality, home and financial services. Designed for businesses of all sizes, media/publishing agencies, call centers, and more, it is a call recording solution that helps track and analyze calls to understand caller intent and behavior.

DISCLAIMER: Below is an AI generated transcript. There could be a few typos but it should be at least 90% accurate. Watch video or listen to the podcast for the full experience!

Mo Asady 0:00
We cannot say the same about when the conversation moves off. So the second the customer picks up that phone. All that tracking is got

Alexander Ferguson 0:14
Welcome to UpTech Report. This is our applied tech series. UpTech Report is sponsored by TeraLeap. Learn how to leverage the power of customer stories at Today, I’m excited to be joined by my guest, Mo Asady, who’s based in Orlando, Florida. He’s the CTO at Convirza. Welcome Mo, good to have you on.

Mo Asady 0:34
Thank you. Thanks for having me.

Alexander Ferguson 0:36
Now, how we understand for Convirza, the problem that you solve just starting off, they kind of set the stage for the rest of this, I understand it’s a high level looking at your website, enterprise call tracking and call optimization platform. But truly what’s what’s the problem that you guys are focused on?

Mo Asady 0:54
So in today’s online world, there are a lot of tracking solutions for when the customer is online. We cannot say the same about when the conversation goes offline. So the second the customer picks up that phone, all that tracking is gone. So what we have come up with his his creative ways to maintain that web session, the continuity of the online track and connected to that conversation. That’s the problem that most of the companies that work with us are trying to solve.

Alexander Ferguson 1:32
Being able to bring attribution to your efforts to what the sales folks are doing with your marketing efforts. And that’s why I love people I’ve loved being able to have internet technology of tracking links, etc. But you’re right, when you go to the offline of talking to someone you lose that visibility, but new technology has made that possible how to understand, let’s dig into the technology itself. How does it work? And also, how is it different from other options out there.

Mo Asady 1:58
So the company is 20 year old and technology terms, that’s a dinosaur, it’s a very old company. And we we started with just classic cold tracking. So we will provide phone numbers that customers could use in their marketing campaigns, and that all these recordings would be all these calls would be recorded and analyzed by humans. As technology evolved, we took out the human factor out of that. So right now AI is listening to these conversations. And it is able to look at probably 50 different indicators of what happened on that call. And did that call result in a conversion? Answering the big question of ROI. Because when it comes to marketing, it’s the one big question that all customers are are wondering I spent X did, how much did I get back? So we help them understand the amount of conversions that resulted out of that marketing campaign. We also identify any lost opportunities, how many leads came in, but did not end up converting? And why did they not convert? Was it as a training issue with your agents, that they just miss certain key aspects in the conversation? Did the customer acquire the contact information of the color that they own the conversation and really lead it in the right direction was any competitor names mentioned, etc, etc. There are so many indicators that we look at. But ultimately, at the end of the day, we help them better understand the data around the marketing campaign and be able to tweak it the next time because marketing is all about planning, executing, measuring, and then doing again,

Alexander Ferguson 3:56
that cyclical nature for sure. For those that are familiar with our series, we usually like to look at the technology piece, the business piece, and also the journey that you’ve been on and the business itself. Let’s stay on the technology just for a little bit longer. And you’ve you’ve mentioned a lot there, and there could be a lot to unpack. Let’s start with you said the 50 different answers, what are some of those attributes that that your algorithm your, your, your, your AI is looking at?

Mo Asady 4:26
So again, it looks it looks that I think the most important is tying it to the money to these 10s on marketing campaigns. So our ys is right at the center there being able to identify conversions, being able to tie amounts to these conversions and say you convert it X amount that is extremely important. I think what is unique about about conversa is that we are able to identify not only actual conversions but lost opportunities saying there was 50,000 to be made on this campaign, which you completely missed. And we do that almost in real time when I say almost four seconds behind the call. So the second call is done. We, the AI looks at it and identifies a lost opportunity. And that triggers a notification to the floor manager to the sales manager, whoever is working on the on the on the client side, to let them know that you’ve just lost an opportunity. Because this is the person their contact details, pick up the phone and close that opportunity because something did not go as

Alexander Ferguson 5:36
planned. Is it looking at is simply alright, I don’t it’s listening, natural language understanding. No, thank you. I’m not interested, or is it under? Is it looking beyond that of no way?

Mo Asady 5:49
Dr. Okay, way, way beyond. So keyword spotting is the old way of doing things. And that’s how we started, we were looking for certain keywords within the conversation and flagging them. But now the AI has evolved. And what it does is it looks at the entire conversation and getting the entire context of what really happened. One, one really cool feature that we is in beta right now we’re going to release it probably in a couple of weeks, is taking a 30 minute conversation and summarizing what happened on that call in just a few sentences. So it would say customer X called you spoke about the weather just to be nice. Then you got it to business. They mentioned a competitor name they mentioned where they saw your ad, and ultimately ended up making an appointment, converting buying, etc.

Alexander Ferguson 6:50
That that immediately makes me think of how some AI has been used to condense like earnings calls, etc, into like little news headline articles. Automatically, you’re doing now the same for sales calls.

Mo Asady 7:02
Correct? Correct. Correct. That’s also saves all of our advertisers time because, again, classically, they used to listen to all these calls. And that’s extremely time consuming, especially for a campaign that has 10s of 1000s of calls, you would you would start sampling and basing your entire report on samples or random samples. But instead now we allow them to just take a look at a you know in a class at a summary, and be able to understand what happened on the call. And more importantly, to focus on the calls where things did not really go as planned.

Alexander Ferguson 7:41
For for this to be able to to focus to use as you’re saying, you mentioned also giving suggestions. Are you building around it’s like here’s now go do this with this, or is it more of a dashboard like, here are the results of what’s happened and then they have to infer the

Mo Asady 7:59
Okay, so the way looks today is a dashboard, the way it looks today. And that’s an important keyword. So it will give you all the insights about all of your calls, almost in real time. But I’m not a big fan of the word almost there. What I want to do is do it in real time. So what we’re working on is a dialer where a call is outbound, or inbound, an agent is speaking with a customer. So once we are at the point where we are able to get the transcription of the call live, and it’s available. And it understands not only word by word, but it’s changes the context based on historical sentences being set. So it will update itself. And then that will map into all of our indicators live and be able to coach the agents, in a way telling them where to take the conversation, which key points they missed, and how to own it back again.

Alexander Ferguson 9:05
Oh, sec. So you’re saying during the live call the agents having a visual coach by this AI of what to talk about next based off of what it thinks is happening in the conversation?

Mo Asady 9:16
Correct? Correct. We’re, it’s not yet announced. But we’re 90% done with it. I expect it to be done in a few weeks.

Alexander Ferguson 9:24
Wow. Wow. So there’s, there’s a lot to look to even talk about that. What does the future look like as a salesperson in a role where you get live coaching by this artificial intelligence, and good things, bad things, thoughts around that that could come from that whether it’s well understood thoughts, you have preconceived notions of being I’m being told by an AI what to do versus the reality of the data that’s there. But if I if I take a step back at the business use case, I mean, the marketers want to know what’s working, what’s not working and the sales leaders you’re Working with both marketing and sales leaders, I guess to for this type of tool to be taken advantage of.

Mo Asady 10:05
Correct? Correct, we work with both sides of the equation. And just to to go back to that point where it could be taken a negative way where computers are telling people what to do, it’s not really bad because a good salesperson is a good salesperson. But if you simply just forgot to collect the contact information for the caller, or you did not mention your your business name, or a competitor was was mentioned, and you needed to turn that conversation into a positive conversation. It will get tips and tricks, but then a good salesman always be a good salesman, those jobs are there to stay.

Alexander Ferguson 10:49
I love the distinction there break down because we as humans, we always forget something. And that should be the role of technology is that it aids us in being able to do our jobs better than then even without it. Absolutely. This the base where we’re technology is today, it wouldn’t have been possible to do this. 10 years ago, five years ago, let’s let’s take a little bit, I’m curious on your journey, in brief, like, where did you start in technology.

Mo Asady 11:19
So I graduated from Stockholm, Sweden, that’s where I got my bachelor’s move to London, start as a developer, became an architect, moved back to Stockholm, worked a few years from Microsoft, then decided to try something new, went into video streaming online, spent five years there. Then another new thing and thought, let’s use technology for logistics. I did that for five years. And now I’m back in the telecom industry. So a little bit all over the place all over the world.

Alexander Ferguson 11:55
What what what about coming back to now telecom? Like what’s, what interests you? Why are you looking into this and excited about this role,

Mo Asady 12:03
new solutions to old problems, this is an ancient problem, it’s always been their phone call. I mean, phones have been around for a very long time. But there are new solutions to that problem. AI is definitely a huge aid. It’s, it’s taking over, it’s making it cheaper to understand the data, it’s making it cheaper to generate the data. And now we’re able to present it to our customers in ways that not only let them understand what they’re doing and how well they’re doing it, but also how to make it better. Because here, here’s a thought for you, when when you think of customers or customers who pick up the phone are very unique. I don’t think they belonged in the lead category. They’re more on the on the prospect side of things, because they’re either new customers that are looking to convert, they are demonstrating that will to convert by picking up the phone speaking to a real human or existing customers that have a real problem that needs to be solved immediately. And they are two sides of extremes that need to be handled very carefully. Because an angry customer will definitely result in a bad review. And we all understand how important these reviews are. And a hot prospect that is looking to convert needs to be handled in a special way. And if you compare them to form submits emails, etc.

Alexander Ferguson 13:48
It’s being able to understand as quickly as possible who this person is and how to deal with them is going to make the difference in better outcomes.

Mo Asady 13:57
100%. And the way the way we do that just back to technology for a minute is we tie that phone call that person to a web session. So again, we’re we’re good tracking them online there. There are tons of tools that do that. Were able to know how many pages they visited, how long they spent on each page, what they clicked on, etc. But then when they pick up the phone that connection is lost. What we have come up with to solve that problem is called DNI dynamic number insertion. What that means is you will be displayed a unique phone number. It’s not the big toll free number that the company uses but instead just a unique number that is tied to you as a person. When you call we know your entire online history. We still continue to see your online history even if you’re engaging on a on a conversation and you click on the website looking at something we’ll still be able to see that and know that it was you and then once you hung up that got that phone call, then that number goes back to the pool is recycled and used for the next person.

Alexander Ferguson 15:07
That ability to know who that person is the the what they’ve been looking at the website helps that agent potentially do their, their their job better. I am I know of the industry of this industry, but there I don’t know all the pieces. What I’ve heard is that it’s growing. So I’m curious for you like speech recognition is comes kind of like table stakes. Everyone has a technology, the ability to listen and understand how do you guys compare to other solutions that are that are growing up? Gong and other things like that?

Mo Asady 15:43
Well, in terms of speech recognition, it’s, it’s a very tricky subject, because Dart so many accents take mind, for example out there. And we what we do is we do the converse, a test, I think we have employees that speak in total, over 20 different languages. So when when we speak English is going to sound extremely different where across multiple time zones, and we we tested ourselves to make sure that is able to pick up what we’re saying. If it does that, then definitely it will do very well up there. So that’s what we have done historically,

Alexander Ferguson 16:26
we don’t using your own algorithm for speech, or no, we don’t,

Mo Asady 16:30
we don’t know, it’s one of these things that, you know, if you cannot do it perfectly, it’s better to just use a third party and focus on what you do best. I’m all for shortcuts. When, when, when when possible.

Alexander Ferguson 16:44
Yeah, as I said, I think I said earlier is it’s it’s almost become table stakes, you have the big three or four that are doing this that have pouring research in and being able to have this and you can build on top of it. So where where is your UC as your sweet spot compared to the again, the other solutions that are that are arising out there.

Mo Asady 17:03
So it’s all about the telephony knowledge. And we’re experts in that matter, and combining it with new technologies. So again, I came from industries where I’ve never worked with telephony, I had to learn it the hard way. But then combining that with new web technologies and AI and machine learning technologies is what made a difference. Because there is a generation shift happening on the telecom arena of I would call it where you have your hardcore telephony engineers that work only with sip. And now everything has changed to API based telephony, where you can do a lot of things live as as the call is happening. We’re talking about the ability to do web lookups. And scraping and collecting certain information live as the call happens, because it’s always trigger action. As long as you have the trigger and the action, you’re able to do whatever you want.

Alexander Ferguson 18:09
Did you say API phone calls? Is that is that the term? So 20? That is that is the term? Yes. How long has that been around in existence?

Mo Asady 18:19
Probably five years,

Alexander Ferguson 18:21
five years. Okay. And what what? Is that becoming adopted everywhere? Is it just become the standard? Yes. What’s what’s kind of then what are you excited about? We look forward from here and the opportunities that exist with API phone calls and the interconnectedness that that means with technology. What gets you excited,

Mo Asady 18:47
again, is that it’s an ancient problem. And there are limitless solutions to that problem. I think there are so many things that no one have thought about yet. I don’t believe phone calls are going anywhere. We will still use them for when we need them. And again, people pick up the phone are extremely different persona, so it should be handled accordingly. So they’re technologies evolving on all fronts. There are ways to integrate it to some some other cool things that are happening elsewhere. We’re just excited about the future. And we’re there is no clear leader in this industry today. Let’s be honest about it. So what we’re looking to do is to continue growing by just being a leader in technology first, and proving ourselves and growing our presence and market share.

Alexander Ferguson 19:50
If you think for these marketing leaders and sales leaders that could be using our technology, not others. If you would give them I say words of wisdom. But it’s more than it’s more than that. I mean, for them to be able to use technology to do their job better. I mean, what some of the things that they should be keeping in mind.

Mo Asady 20:13
Just like any other tool, we’ve listened to our customers, because, as a technologist, it’s very easy for me to assume that I know the solution to a problem. But unless it’s proven, by putting in the market and getting feedback, we’ve been doing that for 20 years, we’ve been collecting feedback from our customers and making sure that we build exactly what the customer wants. So it is a tool that saves them a lot of time. We don’t want to add complexity to their daily routines, we understand how busy they already are. So we want to make sure that we’re providing them a tool that saves them time, first of all, and gives them all the data that they need to be able to make more qualified decisions. And then their daily jobs.

Alexander Ferguson 21:06
I I’m I’m fascinated when technology can come in and actually help people do more with their jobs to be able to to elevate the the opportunities because I think that should be the role of technology, not remove people from the job, but actually help us do our job better. And I am fascinated with the ability for tech for AI or artificial intelligence to understand human conversation and be able to understand the nuances that originally just people like that, you’d have to have a person, truly listen and get the nuances. And I feel like we’re getting to that point, that’s where you kind of started our conversation, you were digging into the technology that it can start with just keywords. And now it’s actually something so much more. If I can just spend another moment on this just are you are able to understand sentiment and even other things like that, like what were if we were looking at what to what extent can truly an AI understand human conversation.

Mo Asady 22:06
Absolutely it does. It does understand sentiment at not only on a an entire conversation level, but also at a sentence level. So if you say, I’m not sure about that, it would pick that and say you’re more on the negative side, because you’re not even neutral there. And it’s it’s, it’s very fascinating. And we’re planning to take it a step further, potentially, because most of sales calls are moving from typical phone conversations to what me and you are doing right now everything is moving to zoom, Google meat, etc. So we have an additional layer of information that is not being utilized, which is the video track of this conversation. So we’re working on a technology that is going to look at facial expressions and being able to aid the the the agent or the salesperson in this case than just reading the room and knowing what how the person in front is feeling about a certain product or service.

Alexander Ferguson 23:24
It’s I think video calls only become more natural and normal. But that is a whole nother level of challenge to understand the nuances and are people paying attention. I just had a recent conversation on on uptake of another person talk about this exact same thing. It’s like if someone is looking at you, or if they’re looking at, you know, their their screen over here, but you’re actually over here. But does the computer understand the difference between the two? There’s a few challenges to solve.

Mo Asady 23:50
Absolutely, absolutely. But the great thing about machine learning is, all you need to do is show it some good examples of the problem and the solution, then it will figure everything in between it will figure how to get from the problem to the solution. And that is the key to building. I’ll circle back to summary which I think is a very interesting feature that could take a 30 minutes conversation and turn it into just a few sentences that tell you exactly what happened. In order to be able to generate that very condensed piece of information. You need to fully understand everything that’s happened on the call. And keyword spotting would definitely not get you there.

Alexander Ferguson 24:42
GBT three, I’m talking about the exact name is that playing a role in in some of the stuff have you got are you guys using it or what are your thoughts around it?

Mo Asady 24:52
So we use a lot of technology. This is always about being up to date. What’s new out there. And what I can tell you about our ai t is they are very good at reading new publications and papers out there, and understanding what the brightest around the world are doing. And just jumping on it and, and figuring out how to do it on our own. It’s just an interesting team. They are hungry to learn more. And I think that is the key.

Alexander Ferguson 25:30
I feel like with with the open source community that is growing as a base, and then for the businesses being okay, how can we apply that to solve a specific problem challenge in a workflow, and you guys are focused on that for for this enterprise space of exactly from call tracking to from coming to a website to understand how that sentiment comes and that entire workflow, it’s different than another use case, which could be something very completely different, but it’s applying it to solve this unique challenge. That that’s where it’s nice technology. Absolutely, yeah,

Mo Asady 26:03
it’s all starts with understanding the problem, you need to, to understand and own the problem, to be able to figure out solutions. It’s AI and ML is a lot of trial and error. That’s just the way the industry works. We have these extremely bright engineers that are doing complex math that I do not understand in most cases. But at the end of the day, they keep tweaking it until it reaches a level where both us and our customers are happy. So we’re hitting 95% accuracy on pretty much everything that we do, which is extremely high in the in the ML world.

Alexander Ferguson 26:48
I feel like about a year ago 80% was was really, really good. Like that’s, that’s like, yes, we got 80% there, and that’s a good thing, but to be able to get to 95% That’s impressive.

Mo Asady 27:01
What 80% You’re still 20% wrong, so it wasn’t so acceptable. We pushed it to 95. And we’re still 5% Wrong, we will keep

Alexander Ferguson 27:10
pushing. And that is the the future and exciting this of AI and I’m excited for where you guys are going to. For those that want to learn more about converse, you can head over to that Thank you mo for sharing your story and the future of where we’re headed. This is exciting.

Mo Asady 27:30
Thank you. Thanks for having me.

Alexander Ferguson 27:31
And we’ll see you all on the next episode of UpTech Report. Have you seen a company using AI machine learning or other technology to transform the way we live work and do business? Go to UpTech and let us know


YouTube | LinkedIn | Twitter| Podcast

Future of Education after Covid-19: Is Virtual Learning Here to Stay? with Narine Hall of InSpace

Past, Present, and Future of Data Analysis Technology & Automation with Ravi Shankar of Denodo