In this presentation Samveet (Sampreet) outlines the evolution of AI in outbound sales from a focus on sheer volume to a need for relevance. He explains how early AI tools enabled mass email outreach, but resulted in decaying lists, low reply rates, and increased SDR workload. The core problem is a lack of contextual, signal‑driven targeting and fragmented tool stacks. To address this, Samveet introduces Tario, an AI‑enabled outbound platform that unifies CRM data, multi‑source signals (LinkedIn, news, hiring activity, company website), and personalization workflows into a single interface. Tario generates context‑aware email drafts, supports nuanced dialing with real‑time objection handling, creates AI‑summarized call notes, and pushes insights back into the CRM. The platform integrates bidirectionally with major CRMs (Salesforce, HubSpot, Zoho) and offers a ChatGPT‑like chat UI with enterprise security. A customer case study shows a reduction from 50 to 25 reps while doubling meeting volume, highlighting efficiency gains. Samveet discusses the evolving role of sales managers—balancing oversight of reps and AI agents—and emphasizes continuous playbook tweaking. The discussion also touches on future unified multi‑channel messaging (email, LinkedIn, text) and the cat‑and‑mouse dynamic of spam versus buyer trust. The session concludes with a call to action to book a demo at tario.ai.
Chapters
Here I'd like to focus on, you know, what we've seen happen over the past few years, especially starting since the advent of AI has come in. We've gone from volume to relevance, and a lot of sales leaders are struggling with the fact that they have a lot of volume and a lot of noise going on, but not enough focused presentation, uh, focused efforts. So volume was the answer that AI gave initially, right? But now it's the problem. We all were excited with the prospect that we could just send out, you know, 5,000 emails with one SDR or engineer manning it, and that will get us our pipeline started and make sure that can take us for, you know, enough meetings to then convert into closed ones. But that's been much— uh, the results have been counterproductive. Most email lists have started to decay, and the cold email rate drops over 5 to 6% every year. And the SDRs, far from being replaced, have to spend a lot more time to just juggling through all the activities that are done by these AI agents or automated processes. And the net result has been much, much worse. Not only is the pipeline getting weaker, but it's also harder to figure out what is working and what's not working because we are sending out so much information with very less feedback and very less control on who we are reaching out to and when we are reaching out to. AI was a double-edged sword, and unfortunately, until now, bulk of the outcomes in sales have not been up to the mark. Inboxes aren't immune to all of this, and buyers are tuning out. Everybody is changing the way their email filters work. Even dialing people has become harder with voicemail drafts. So, you know, where are we at? The shift isn't about, you know, figuring out that, okay, I'm just not going to use AI and sell a lot lesser. The way we thought of AI improving volume is not correct. We must think about how it can help us be more relevant at scale. So we need to spend a lot of effort on targeting. So we must find out who the person is that we want to sell, and do we have the right signals? Signals can come from various sources. It's not just, you know, what they said on LinkedIn or what the company— what the company's website saying. We also must look at hiring signals, figure out the need that they will— that might have, what people— what other people in the same org are talking about. I think context-aware messaging is the most important thing. We must earn the conversation just to make sure, like in the old days, if somebody was manually writing an email, the same amount of effort must be spent. Everything must be reviewed before being sent out, and we must not just have volume as the metric. Currently, a lot of people just look at reply rates or the amount of emails they send, or how many emails are being opened, and trust that, as you know, brand identity being created. But we must move to, you know, much more concrete metrics in terms of tracking how our pipeline is doing. We should be looking at the closed ones in the meeting book and also being able to attribute this to, you know, how what exactly happened in the messaging so that this worked, right? We must constantly tweak our playbook. We can't assume the initial playbook is correct or change the playbook just based on some signals, but we must be aware of the metrics that we're trying to target. The reason we created Tario, right, we found a problem in the market where most sales teams and sales reps specifically are just struggling between, stitching different tools together. The prospect on different tools, they, you know, send emails and do research using AI, maybe on ChatGPT or some other platforms. They send out emails from one service, dial from another service, and all of that, you know, then goes— has to go back into the CRM. And this is a very disconnected workflow, especially in the age of AI, right? Your AI agents must have context of everything that has happened. So to work on that, we built Tario to connect all of these pieces together. I'd love to move over to the demo and show you some of the snippets of how we are trying to connect all of these pieces together. So in terms of in our sequence, how do we figure out what the right metric is or what the right system is? Let's look at how we actually personalize and how Tario uses your data grounded in your CRM knowledge and your company knowledge to personalize and how it's, you know, previewed to feedback. So here we can see, right, uh, we're just looking at a sample email for Aliens Alias that's created, and you can see all the sources it's looked at. It's publicly available LinkedIn information, news reports, and your own company's website. And it's also, you know, telling you what, and it's attributing to where it's finding that information from. So we look at, you know, the value prop that we can provide what the company is doing, which industry they're in, and you know what probably is some issue for them. But let's say, you know, we want to shorten the email and add a more appealing CTA. This isn't just possible if you're giving a few dropdown suggestions on how to craft your agent. What we allow you to do is, you know, be a lot more intricate in terms of how you are customizing things and, you know, what your call to actions are and what your company's values are. So all of these are obviously grounded in the knowledge base that you create, and that's, you know, tweaked over time continuously. We want you to, you know, work with the agent on a day-to-day basis just to be able to make your messaging absolutely correct, right? This is where we can solve the first part of the problem that I spoke about. Where we know we are reaching out to people at the wrong time, where messaging is very generic or just based on random AI signals found on the internet, and that's what's leading to, you know, the lower reply rate. So with just doing these small changes, we've seen at least a 2x difference in the number of emails that have been replied to, just simply because, you know, this does look much more similar to what an SDR would write, and not just in terms of the style, but also in terms of, you know, who is trying to target specifically. I'm sorry, I just missed out on this. Yeah, next up, I'm running out of time, but I'll just show you how we dial and how our reps will be, you know, supercharged by dialing using Tario. You can see, you know, we— a few different things going on here. You must know what the local time in which the lead is. We're all working across time zones, and we must be able to power dial, right? The connect rates are low, So we allow you to power dial where, you know, you, you keep dialing until somebody picks up and the moment they pick up, you're not caught off guard, right? You have the suggested next steps, what the objections will be, some discovery questions, and you know what the solution fit is like. And this is absolutely critical because, you know, in today's time, even though we're dialing a lot, it should not seem like the calls are nuanced or the person is unprepared. Once the call is done, right, the old way was to just look at the recording, but we wanted to create AI-generated summaries for everything so that we figure out in this call what the rep do correctly, where is the deal at, what the rep improvement suggestions are, and some next steps. And finally, right, as sales managers and directors, all of us need to focus on getting effective reports. Gone are the days where we need to stitch around SQL ourselves and figure out how things work. All we ask for you is, you know, you chat with the system and this is an agent that's sitting across the board including access to your CRM to figure out, you know, what the activity reports have been, what is working, what happened in those calls, what are some of the things that need to be done and the recommended actions, right? And this wasn't possible if you had tools that are disconnected together. The way we have an agent layer sitting on top of all of your outbound processes is how we can make sure that, you know, not only is all the action relevant and connected to each other, but also when you're reviewing things, you can get an overall, you know, glass view of what's going on. I think that's my time. And Jared, hopefully, you know, they can go to our website, www.tario.ai, and book a demo with us to see more.
Can you share one quick customer success story?
Yep. So one of our customers, right, they went from, you know, 50 reps to 25 reps in, in about 2 to 3 months. And what they saw was it wasn't just the reps, uh, you know, reduction in headcount or, you know, managing attrition better, but what they also looked at was, you know, the volume of outreach increased significantly and the number of meetings booked was almost double of what they had with 50 reps. The way they did that was with, you know, getting a small number of, you know, power users, power BDRs that they had, looking at, you know, their strategies and emulating all of the agents in that manner. So I think the way they did it was they tweaked their best winning play moves and made sure that, you know, most of the people were able to follow it. And all of the messaging was grounded in those success stories.
Tell us also about the integration into existing CRM systems. How does that work?
Yeah, so we, we integrate into most existing CRM systems including Salesforce, HubSpot, Zoho, and we connect both ways. And we also make sure that all of our AI summaries are also pushed back to your CRM and the AI agents have access to it while writing communication messages or while creating reports.
And what feedback do you get from sales reps that use your system?
So I think one is a relief because a lot of the systems are now been built, you know, 15, 16 years ago. And just in terms of the experience that they have, they have, they have a lot more time freed up. But also the UI and all the information shown is a lot more intuitive because they want everything to feel like ChatGPT. And we're trying to emulate that experience with having enterprise-grade security as well as it being grounded in workflows and processes. But they can treat it as a pure chat interface that they're using.
And what— how does the role of the sales manager change because of that?
So I think the manager will now have, you know, two main roles. One is managing what the reps are doing on a day-to-day basis, which is hopefully, you know, talking to prospects and talking to customers. But the other part is they must manage the things they're replacing, right? They can't just let that be on autopilot. So they have to be able to manage the agents and what they're doing. In conjunction with, you know, the way the sales reps are working with the prospects on day-to-day. And this has to be a constant feedback loop with, you know, time put into the platform. The reason a lot of failures happen in implementing AI solutions today is because the initial success is— the initial success factor is there in terms of how effective it is, but they aren't spending enough time tweaking it to make sure the results continue.
Right.
Tricia, what is your take on this new solution?
It's fascinating to me because the challenge of how we keep up with everything, and I'm curious, how is it affecting your guidance in terms of how to use email most effectively? You know, what you just said, Samveet, is everything. If the message isn't right, it doesn't matter that you're doing it faster or you know, um, or that you're hitting more people. Um, so what is your best guidance in terms of the go-forward of how we're communicating, uh, with our prospects and with our clients?
Yeah, so I think the best thing to do is to, you know, take a look at what has worked for you and what is getting you the most number of replies and start scaling that. And scaling that in a very measured manner, right? If you're working for a certain, you know, ICP, just increase that ICP a little bit and your lead count a little bit and see what the result is and keep going on. Don't take that something that's worked for, you know, 100 leads and scale it to 10,000. I think scale it to 150 first and then 200, and that will be a lot more efficient in terms of also protecting, you know, your mailbox and your technical domain or technical markers like your domain health, etc.
It's so interesting. I find the whole communications landscape— I'm curious what you think is coming next, uh, because I prefer text. It makes my life so much easier. I've had to follow that with WhatsApp, which means I have two places to check when it's something urgent. And then I've got all the email, all the social handles, all those things. And to me, it just seems like there has to be some kind of total unification of all of that or total customization. And I don't know which. What are your thoughts?
Yeah, so I think we're seeing a unification in terms of, from messaging point of view, where I need to, you know, send out messages. I'll have all of my channels, you know, combined, including, you know, messages on LinkedIn, text, email, all in one platform so that I can see it easily. From a buyer point of view, it's a bit different, right? They want to make sure that, you know, everyone wants to talk to salespeople, but in a limited manner. We need to make sure that spam doesn't happen. So it's going to be, you know, a cat and mouse game where, you know, the more advanced we get on one part of it, people will start abusing it and, you know, buyers will then, you know, start reducing their trust on that, that part of the platform.
Incredible, Sampreet. I'm so grateful.
I want to just summarize this in my words, Sampreet, and correct me if I'm wrong. So that if somebody looks into Tario, the number one benefit is that you can scale outbound autonomously without adding to the headcount. So absolutely, the, and the benefit of that is you've customization towards the ideal customer profile so that you really engage the prospect in their language. The second one is the personalization at scale. So by analyzing the CRM deal history and the past interactions, the AI then automatically generates the right outreach. And I think the third benefit would be that Tario reclaims the time that salespeople need for high-value selling.
Correct. Yeah, I think the most important is, you know, getting that time back, and all of these are just means of getting to that outcome.
Wonderful. Well, thank you for an excellent presentation. Very clear. And I recommend we just put that link in umbrella.io. So go to, uh, uh, tario.ai.
Bookdemo.
Yeah, thank you. Here we go. Thank you so much.
It was a great, you know, session, and hopefully, you know, very successful conference.
