AI for Growth: Lessons from Leaders · 2026-04-08 · 37 min
Key moments - from our scoring
Substance score
52 / 100
Five dimensions, 20 points each
Tamara Asselta brings a practical lens to how early-stage companies should approach AI adoption. Rather than treating AI as a complete replacement for human work, she emphasizes a hybrid workflow approach: use tools like Claude for coding, Lovable and Ready for UX design, and Notebook LM for comparing data-heavy sources like insurance quotes. She walks through specific use cases like Zapier automation for sales workflows (lead capture to follow-up scheduling), Canva for quick graphics, and the critical role of human review before shipping any AI-generated output. A key tension she highlights is the MIT research showing people over-relying on AI at the expense of critical thinking. For B2B operators scaling teams, her FACTS framework for prompting (Format, Audience, Context, Task, Sources) offers a structured way to get consistent, bias-free outputs across different LLMs. She's also cautious about AI-generated content saturation - noting that on LinkedIn and elsewhere, generic AI writing now blends into the background, while authentic human voice remains differentiated.
Notebook LM is specifically designed for extracting and comparing factual data across multiple sources - Tamara used it to compare insurance quotes from multiple providers in table format to find the best coverage and price.
Zapier has a beginner-friendly interface that walks you through workflows step-by-step; a realtor can set it up to automatically send follow-ups, create calendar bookings, and populate forms when leads submit a contact form.
FACTS stands for Format (how you want the result to appear), Audience (who or what expertise you're drawing from), Context (background goals and constraints), Task (the action verb describing what you want), and Sources (2-5 high-quality examples to guide the AI).
Use Claude for coding because it's advancing faster than ChatGPT; use Lovable or Ready for UX design of websites.
AI can hallucinate or produce biased outputs, and AI-generated content now looks generic and saturates social feeds - human review ensures quality, accuracy, and your unique brand voice comes through.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode covers practical AI tool recommendations and the FACTS prompting framework, which are useful but relatively surface-level. Most insights are introductory-level (e.g., 'use AI for mundane tasks, not critical thinking') without deep investigation of tradeoffs, failure modes, or quantified outcomes. The guest mentions an MIT study about over-reliance on AI but provides no specifics. Tool recommendations lack strategic depth on when/why to use them beyond basic use cases.
what we have to be very careful with, and I say this to a lot of the businesses that I'm working with, is you need to also still make sure that your collaborating with that AI, that you're making sure you're not just solely using it as your go-to and you're not just letting something ship out without reviewing what that work looks like
I just saw one from MIT that they're getting some early data where people are relying too much on AI and not using enough of that critical thinking
The FACTS framework for prompting (Format, Audience, Context, Task, Sources) is presented as original to the guest but is a repackaging of existing best practices in prompt engineering. Most tool recommendations and use cases (ChatGPT for coding, Zapier for automation, Canva for design) are standard industry knowledge. The 'human touch' and quality-control themes are well-trodden. No contrarian positions or first-principles thinking emerges.
So the acronym is FACTS. So, and how that acronym goes is the prompting best practices, right? So it starts with F, which is format. The next one is A for audience, C for context, T for task, and then finally S for sources
Cloud I actually would say is advancing way more beyond chat GPT. I there's a lot of articles about that now and I won't sit here and focus on that
Guest is a consultant advising early-stage startups on operations and team scaling with a recent side venture into AI adoption. She has no demonstrated founder/operator credentials, no exits, no revenue milestones mentioned, and no evidence of building at scale. She appears to be a consultant/coach observing clients use tools rather than a practitioner who has shipped products or businesses using AI at meaningful scale. The authority derives from advising others, not from direct execution.
my latest venture for the last ~year or so has been starting up my own side consulting business. And now I'm doing that fully. helping small startups and mid-sized businesses to help build their operations, build their teams, and keep them motivated
I've been working with some teams lately where They've been feeling maybe a little disjointed
The episode provides limited concrete data or named examples. The guest mentions specific tools (Claude, Notebook LM, Canva, Zapier, Lovable, Ready) with brief use cases, but no metrics, timelines, or measurable outcomes. The insurance comparison example is personal, not business-relevant. The realtor Zapier workflow is generic and not tied to specific results. No revenue figures, conversion rates, time savings quantified, or customer case studies provided. Mostly generic advice about 'automation workflows' and 'faster outputs.'
Proposals that would take hours are taking 32 seconds, which is just unbelievable
I had to renew my insurance. I have three motorcycles, a car, homeowner insurance. So I wanted to make sure that I was getting the best deal and the right coverage. I went around to every single insurance company and I took quotes from every single one. I put all those PDFs that they sent me into Notebook LM
Host asks clarifying follow-ups and pushes for specificity (e.g., 'Talk about what are some of the specific tools...', 'could they just log in and be able to do this on their own?'). However, follow-ups are often surface-level and don't challenge weak claims. When the guest makes broad assertions (e.g., MIT study on over-reliance), the host doesn't ask for details or sources. The conversation affirms rather than probes. There's minimal productive disagreement or skepticism about when AI actually creates business value.
can you be a little more specific? Talk about what are some of the specific tools that you see some of these companies using and maybe how they're using it
So somebody who doesn't know what they're doing, could they just log in and be able to do this on their own? Would it be obvious for them to do it?
Computed from the transcript - who did the talking, and the words that came up most.
Tamara Asselta is a strategic operations and business growth leader based in Portland, Oregon, with over 15 years of experience turning complex organizational challenges into scalable, revenue-driving, high-performing systems. She is the founder of Stratas Consulting, where she partners with startups and mid-market companies as a fractional strategic leader stepping in to build teams, coach managers, implement emerging technologies, launch new services, and guide organizations through growth and change. Before founding Stratas, Tamara spent nearly eight years at Adswerve, where she built the company's Adobe Services practice from the ground up, scaling it to over $8M in annual revenue and earning Adobe Emerging Partner of the Year two years in a row. As Senior Director of Professional Services, she led a 30+ person organization, managed a $13M+ consulting portfolio, and served as a trusted advisor to C-suite executives across Fortune 500 companies in healthcare, finance, retail, and tech. Her career has always been defined by the intersection of people and performance.
Transcribed and scored by The B2B Podcast Index.
Megan Driscoll: Welcome to AI for Growth, Lessons from Leaders. Today I have Tamara Asselta joining me. I'm super excited to have this conversation about how AI is impacting her business and sort of more broadly how it can impact yours. So thanks for being here.
Tell me a little bit about, start with just tell me a little bit about yourself so the audience can hear and a little bit about your business too. Tamara Asselta: Sounds great. Hi everyone. Hi Megan.
Thank you for having me. I am based in Portland, Oregon, and I've been here for about 20 years now, which is wild for me to say. I'm originally from the East Coast, so I have a little bit of the West and East Coast vibes, which basically how I explain that is... Megan Driscoll: Nice.
Tamara Asselta: I will run you down on the highway, we're going fast, let's keep moving, but then we're gonna have a hug later. That's how we do West Coast, East Coast. So yeah, that's translated into my business is I am a go-getter, but I also really, really enjoy helping people. And my latest venture for the last ⁓ year or so has been starting up my own side consulting business.
And now I'm doing that fully. helping small startups and mid-sized businesses to help build their operations, build their teams, and keep them motivated and help them find what I'm calling their sweet spot. So lately I've been very much in the AI startup space because there's a lot of those happening right now. And I'm helping people to not only move forward with their product, but make sure that the people who are involved in that space have what they need and are coming together as a team.
So I've been working with some teams lately where They've been feeling maybe a little disjointed. They're not really sure how to get to the next level because oftentimes we start with these businesses. We have this idea. We start running with them, but then we need to scale.
And so then when we get to that scaling point, that's when founders go, I don't know how to do that. I don't know where to go next. I just had this idea. I'm really awesome at that idea.
I don't know what to do with the operations. And so that's the space that I tend to help with and where it ends up going a lot of the time, which is my favorite spot. is really coaching the leaders and helping them find where they need to be in order to get the most out of their skills and move forward with that company. Megan Driscoll: Yeah, well, and also to, we as founders, ⁓ leaders, know, CEOs, you wind up a jack of all.
Everyone kind of has to be a jack of all trades when you start, but. very quickly, you need to utilize team and sort of offload the things that you're not good at. And it's always a question of like, what are those things? so identifying, you know, what those things are.
So tell me a little bit about how, like, so you're kind of getting a bird's eye view into small businesses and how they're scaling and growing. Tell me a little bit of how AI is impacting those businesses, like what tools they're using, like what are you seeing in the marketplace? Tamara Asselta: Yeah, so, well, I have a lot of thoughts on this. But yeah, but where to start is, I mean, the obvious one is how AI is impacting businesses is we are now in a place where we can get things a lot faster.
Now there are pros and cons to that, right? Because our expectations are increasing in speed. Megan Driscoll: Great. That's why you're here.
Tamara Asselta: However, what we have to be very careful with, and I say this to a lot of the businesses that I'm working with, is you need to also still make sure that your... collaborating with that AI, that you're making sure you're not just solely using it as your go-to and you're not just letting something ship out without reviewing what that work looks like. So I think that's kind of the hard thing a little bit right now with AI is our expectations are way up here. They're super high.
But we need to also make sure that we're doing our due diligence to check that work and still get the same quality that we expect. And there's different ways to go about that I'm sure we'll talk about here in a moment, but I think there's a lot more accessibility. to how people can show up in their businesses and free up time for more of that critical and strategic thinking. And that's a piece that worries me is to make sure that people are doing that because there are studies coming out.
I just saw one from MIT that they're getting some early data where people are relying too much on AI and not using enough of that critical thinking. So to summarize, think what's really happening for businesses right now is that expectation of speed, which again, it's a great thing because we can get certain pieces out the door. We can save time for ourselves on those more mundane, maybe administrative tasks. But I think that's the key that I like to help people understand is use them for those kinds of workflows.
use them maybe for that really advanced critical thinking that makes you unique and makes you understand the context. Megan Driscoll: Yeah, so can you be a little more specific? Talk about what are some of the specific tools that you see some of these companies using and maybe how they're using it, maybe not as you would advise, or maybe how they're using it ⁓ and it's working, right? What tools are you seeing are you excited about that you're thinking and what ones are like, ⁓ well, people are too heavily relying on this and it's causing problems.
What specifically do you think? Tamara Asselta: Sure. Yeah, I mean, I think I'll start with the obvious ones. So we have chat GPT, right?
We have cloud ⁓ and cloud I actually would say is advancing way more beyond chat GPT. I there's a lot of articles about that now and I won't sit here and focus on that, but I'll just say that those specific tools are great for coding, especially Claude, right? Claude is the one to go to for coding. I there's some really great ones, ⁓ lovable and ready, who are really great for the more UX design of websites.
⁓ Notebook LM, really great when you are looking for factual data from multiple sources. So an example I like to use, because I used it this way myself recently, I had to renew my insurance. I have three motorcycles, a car, homeowner insurance. So I wanted to make sure that I was getting the best deal and the right coverage.
I went around to every single insurance company and I took quotes from every single one. I put all those PDFs that they sent me into Notebook LM and I said, this is what I want to focus on. Can you put it in a, please put it in a table. I want to compare the price.
I want to compare the coverage. Here's the coverage I have right now with Allstate. break it down for me, tell me what the pros and cons are, and that tool is really good for getting that very factual data. You can still use that with Claude and some other ones, but I find Notebook LM is really the one for sources.
So it's really about doing your research and understanding why you might want to use a certain one. What I'm finding too is that there's workflows that we really should be creating with these. So for example, I will go to lovable already. I'll create a framework for an app for a website.
But then I'm going to go and pipe it over into Claude and have Claude finesse the coding. Then I might go use Canva, which by the way, Canva is really great for certain quick graphics, things like that. ⁓ But going back to your question, if you want. really expert design and branding.
You can use Canva as a starter. You can use some of these tools as a way to get ideas. But I think if you are a business that's really trying to capture a whole brand, that should just be your starting point and then go give it to probably somebody who specializes in branding and design that can really take it to the finish line. ⁓ But that's, mean, that's...
Megan Driscoll: Yeah. We're still not at a place. What you're saying is we're still not at a place where you can fully just say, here, hand it over. And what you give me is going to be great because I mean, you get mixed, you know, you get mixed results, obviously.
Tamara Asselta: you get mixed results and for quick and dirty it's great. Yeah, and I think that's, you know what? That's the key piece, Megan, is if you want quick and dirty, go AI route, get it there, get it to your foundation, get it to start thinking of the ideas, but the output, it needs to be human reviewed. It needs to have the human touch because...
Megan Driscoll: But for quick and dirty, it's great for quick and dirty. some companies need quick and dirty. Tamara Asselta: you can start to see what's happening. If you're paying attention, we're starting to get to a point where everything's looking a little saturated.
You can see the writing that looks AI. You can see the graphics that look AI. At first they were fun. Now we're scanning past it.
Megan Driscoll: Yeah. Yeah. Tamara Asselta: We're going, ⁓ it's another one. It's another one.
And what was resonating is that unique voice. The people who stand out on LinkedIn, I can tell you, they use AI, but they have their voice in it. Megan Driscoll: Yeah, totally. OK, so tell me a little bit about, ⁓ I sort of think about AI, there's a variety of different ways that clients can use it.
use it like we just talked about for marketing, right? And so for ideas generation, things like that, clearly for proposals. ⁓ And you know, actual work product that they're sending out. Proposals that would take hours are taking 32 seconds, which is just unbelievable.
Beyond the ChatGPT, the Cloud, things like that, are there tools that you like for organization or for communication amongst team? What other off the track tools do you think are worth looking at? Tamara Asselta: Yeah, so Notion is really great for organizing, ⁓ getting better structure, understanding what your templates are. Zapier, ⁓ my gosh, that one's actually really mind blowing for workflows.
to put automation, right? Automation of those workflows. What's been fascinating that I've been playing with a little bit is I'll use something like Claude to talk about the workflow of what am I, Megan Driscoll: Mm-hmm. Automation.
Tamara Asselta: am I trying to accomplish? And I say, help me create the right prompt for that tool. to finesse for that tool. Now I'll get into some of my secret sauce of prompting here in a little moment because there is a structure to get to a better place with your output and to make it happen a lot faster.
But that's basically what I've been doing is I'll say, here's exactly what I'm looking for. Here's the workflow. This is how it's used. Here's why it's used.
This is what the output is. This is the expectation. Help me create that. And then I put that into Zapier and everything's beautiful.
And it's just, yeah. Megan Driscoll: I know. So maybe we could just spend a minute on that, because I do think it can be like, I use it on a regular basis, obviously, for pretty much every automation that I want to put together. You have to have that account, and you have to be able to use it.
So why don't you just kind of give a high level, this is something that a company could use it for very basically, even if you don't have any really good skills. Yep, great. Tamara Asselta: Yeah, let's say you're a realtor. Okay.
You're realtor, you have somebody come into your website who does the sales contact form. You can have Zappier ⁓ send that person immediately a follow-up email of some sort, you like book a call or have it do the call booking for you. Maybe you already had slots on there. I mean, there's lots of ways you can go about that, but basically you automate that whole touch point process of the sales.
leads comes in, there's a follow-up, there's an ⁓ automated call that's created, maybe there's a scheduling thing, maybe it's through Calendly, you can do some of those. You can even input other tools into this. Then it gets set up. It says you can automate the follow-up of that.
You can automate here's the questions I have for you. Can you fill these out ahead of time? Can you fill out this form? That gets then put into your system and it's all in one place.
You can see the journey of this one sales lead all the way through ZAP.ba by ⁓ setting up that workflow and just automating all that administrative task. Megan Driscoll: So ⁓ could, and I'm asking because I do know some of the answers to these questions, but I like this conversation because we haven't really talked about the automation on the back end. So somebody who doesn't know what they're doing, could they just log in and be able to do this on their own?
Would it be obvious for them to do it? Tamara Asselta: Sure. Yeah. Yes, it's a very simple setup.
I think it's actually one of the easiest ones that I have seen. And that's always, I will say that's something to call out though, that's always a pro and con to certain tools is the easier they are to some extent, sometimes the more limitation there can be. but not always the case. With Zapier, I'm actually, that's probably the one I'm most impressed by is not only do they make the UX really easy for somebody who's like, I don't know where to start.
How do I even set up, you know, a workflow? It will actually walk you through the first time you go in there, it'll say, we're gonna start here. I want you to set up this. and then go to the next step, so on, so on.
And then it'll walk you through your first workflow. So you can even do a test workflow. Just set something up that maybe you're curious about, something that you have to do a lot. Think about things that you repeat, that you're like, this is the same step-by-step process I do every time and see if it's possible in Zappier.
like it truly, it will walk you through there. There's a bunch of tutorials. It's very simple. I've watched people who are not tech savvy, set workflows up in there.
Megan Driscoll: Yeah, agreed. It is a great tool and I think it's a lot of, it is overwhelming for people too. I mean, what I find too, and I think this is part of the problem, it saves us time in the long run, but people don't have time to do it. It's like, they log into that and they're like, ⁓ I don't want to, I don't want to do this.
I don't want to do this. It is. And I find that this, you know, there's, it's kind of a niche that I'm filling in my role in that people don't want to do it. They want it done, but they don't really want to do it.
Tamara Asselta: 100%. It's daunting. Yeah, it can be. Megan Driscoll: And they don't even want to bother with like thinking about how to do it.
They don't want to watch a tutorial. They don't really care. They just know that it sounds good and they want it to happen. And so, um, I think there's a real need for that.
The kind of thing that I do and that you're doing, right. Which is to say, okay, I'm going to do it for you. It's easy. I'm going to do it for you and I'm going to set it up for you you're going to be so happy.
You know, that's the kind of thing that forever. Tamara Asselta: Right. Yeah, and then they'll have that, you know, even if you have somebody else set it up for you, the long term, and even if that person's somewhat expensive or whatever, you know, that that cost setup is now you're saving all of that time and money for that automation. Megan Driscoll: Right.
Yeah. Well, and exactly. Tamara Asselta: And that's something, you you talk about accessibility of knowledge, right, for some of these things and just being overwhelmed. One of the things I've been thinking about doing, not to just...
Go do these 20 minute long YouTube tutorials or anything like that. That's not what I'm trying to get into but I've been thinking about doing these little snippets just for people to be educated on what is possible simply even if it's something as simple as just knowing that Zappi exists or that Canva exists and how you can use it and you could still they could still hire somebody like you Megan, right that does the more complicated workflows. But I think there's power in knowledge of these things in just knowing where to start, even if you aren't the one who ends up creating those so that when somebody comes to you and says, hey, I wanna use Claude or I wanna use Canva or I wanna use Notebook LM, you'll at least know what they're talking about and why that's used versus a competitor.
Megan Driscoll: Right, 100%. Yeah, and think it's, ⁓ yeah, and I think it's, in some cases too, what I find is it's almost so early that people don't even understand how amazing it is until after it's done. And they're like, wow. And I find that about, like, lead gen, I do a lot of like cold email lead gen using AI.
And people are like, it's like Instacompany. You know, within like three weeks, you can have like a fully operational CRM that's running workflows, you know, that's doing work behind the scenes and has been populated with like 10,000 contacts prospects in, and that I've already received multiple emails from you, right? It's like Instacompany and people are like, whoa, you know, ⁓ Tamara Asselta: All Right. Megan Driscoll: But there is a little bit of a leap of faith too, because you have to trust that this is gonna, it's like you're almost trusting the person a little bit, like this is gonna work, right?
And then when you see people's eyes light up, like, this would have taken me two, five years to do. And we did it in six weeks. So let's. Tamara Asselta: Right, and I think that's a big power in it, right?
Is that solopreneurship, it's not, there's no excuse now to go do that idea. The only challenge that I'm seeing that will happen is again, going back to, you wanna make sure that you still have a human approach somewhere along the line, that somebody knows who you are, because... Megan Driscoll: Yeah. Tamara Asselta: it's starting to feel a bit too automated in some areas.
Yeah. Megan Driscoll: Yeah. Well, and I think too, like especially as it relates to Lee Jen, you know, ⁓ the gold or I think where the value that I provide in that piece is I personally, human, am using AI tools to develop these lists of people we're gonna go after. But I'm the one who's deciding who makes it on and off that list.
I am the one who understands the search and understands what we're looking for. And so if you've got a good target, everything is better if you have a good target. But if that's kind of crap, the whole thing is crap. But you have to really put that human touch on.
And then also developing that Tamara Asselta: Right. Megan Driscoll: like what are we gonna say to the client that really speaks, what are we gonna say that represents your brand, right, that will resonate, that will produce the outcome that we want. And that's all human, and there has to be some human connection. I can use AI to help me write those, but I have to really strip them down and really apply my understanding and the client's understanding of what they're doing.
So you're right, like. And I've said this before, I actually don't think that people are going to be replaced by AI. I mean, there'll be some businesses and some roles, but there's no way. There's no way.
I want to talk about prompting because I know that's like your, like, you're so good at it. And I want people to, because I do think that people think they're good at prompting, but I think that they don't know what they're doing. So I want to spend some time on that. So tell us about prompting, about like your best practices, kind of how you go about, how you approach it.
Tamara Asselta: Oh, 100%. No way. Yeah. Yeah.
Yeah, so I have this acronym and I should have looked before. not, you know, I always say, hey, do these things and then I don't do them myself. So one of the things I should have done was checked to see if this acronym already exists, but I did not. So we're just gonna pretend in this moment, in this safe space that I created this.
So the acronym is FACTS. So, and how that... Megan Driscoll: It doesn't. This is just yours.
Tamara Asselta: acronym goes is the prompting best practices, right? So it starts with F, which is format. The next one is A for audience, C for context, T for task, and then finally S for sources. Now that is in the summary, your structure for creating the best prompt.
And the best prompt gets you to your output quicker. It then also allows to make sure that we're avoiding any kind of bias or hallucinations that occur, which we can, if we have time, we can talk a little bit more about that. But essentially you're just trying to make sure that this prompt is including everything that you can provide to AI because these LLMs only know what you put into them and what they've been trained on already. Megan Driscoll: Yeah.
Yeah. Tamara Asselta: So that's why when you go to different tools and you, and I suggest to do this by the way, is try your prompt, same prompt in different tools and see what the output is because it is going to be different. Now there will be certain overlaps because there is some overlap in what they are trained on, but they are different. And it's a fun little exercise to do to see which ones actually do resonate with you the most.
So starting with format, what does that mean? Megan Driscoll: Yep. Yeah, let's talk about format. Yeah.
Tamara Asselta: It's how you want the result to appear. So you would say, create a table, create a bulleted list, create a presentation. It needs to be editable, right? So on, so on, so on.
What is, how do you want it to look? How is it going to be presented to you is the format. The audience, this is the next one, is what expertise am I drawing from or who am I speaking to? So for example, act as a seasoned marketing director or a doctor who knows a lot about cancer research.
Or maybe you're speaking to a woman in their 40s. Maybe you're a teacher and you're talking to fifth grade students. That's the audience when we talk about, and it can be who you're speaking to. It can be what the persona is that we're trying to be.
The next one is context. Context can overlap with a lot of the other ones, right? ⁓ Context can be, who is the persona? What is the format?
What kind of task are we doing? Who's the sources? What are we trying to resemble? It's background information.
It might be objectives. Perhaps you have a boss who says, we're trying to increase our sales by 25%. How many of us are in that space right now? Probably dealing with that.
And you want to give that the context. Perhaps you are creating a report and you're trying to analyze certain things and you wanted to tell AI, our goal is to get to 25 % increase of sales by Q4. What is this data telling us right now and how we're going to get there, right? That would be providing context.
So it's any rules that the output must follow. It could even be the amount of words this thing needs to be, right? Like let's say you're trying to create an Instagram or a LinkedIn intro and you know some of those sections only have so many characters. The task itself is the action verb to describe what you want.
So that could be create a project timeline ⁓ that is in a presentation deck, right? So the action is create the project timeline. Now, Megan Driscoll: Mm-hmm. Tamara Asselta: That's not to be confused necessarily with format, right?
Because creating a timeline could be in lots of different formats. So that's where you have to get specific when you're talking about a task, you need all these other pieces. And then last but definitely not least, your sources. And now other people call these other things, they might say references, resources, right?
I just call it sources. Your sources are two to five high quality examples of what you're looking for AI to guide upon, right? So for example, if I really like a website, I'm creating my, let's say I'm becoming. I'm starting my own consulting business and I want to create a website and I know there's these two websites that are doing really well and they're pretty close to where I want to start.
They might not be exactly my branding, they might not be exactly the word usage I want to have, but structurally they're pretty, they're pretty spot on. So I will put those two sources, I'll say this website and this website. Megan Driscoll: Yep. Tamara Asselta: I want to follow the same homepages or the about page or maybe it's the home page from this one the about page from this other one ⁓ maybe there are presentation decks or word docs or ⁓ pamphlets or whatever or lots of other different sources that you've been provided that now you can input into that and say, I'm taking the ideas from all these different parts or holistically across all of these, can you create something?
So that's what we talk about when we are saying sources is what are the things like? this that you're you're envisioning that could help guide AI and if you follow all this structure I guarantee your outputs will increase by accuracy ⁓ a thousand percent we'll say. Megan Driscoll: Yeah, yeah, would bet too. And I don't think everybody follows the same format.
So I do like that you have the acrobat, and I'm sorry that it was already taken. Damn. So in terms of prompting, ⁓ what is your fave? I know we just talked a little bit about you think Claude is kind of advancing beyond like chat GPT, and we talked a little bit about that.
But let's say you wanted to do this for like, Like what do you think is the best for like a marketing and like shiny documents and things like that? Would you go to Claude or chatGBT? Like where would you go for that? Tamara Asselta: If I was going to do marketing materials, yeah, and for the fancy stuff, honestly, I wouldn't do one tool.
I would probably do Claude, a little bit of Notebook LM, and Canva. I would probably start in that realm because... Megan Driscoll: Like materials and yeah. You do vote two or three.
Tamara Asselta: I would want to see the difference of the input for Notebook LM or the input and then the output between Notebook LM and Claude, especially if I had pamphlets already or certain branding that already existed because I do feel Notebook LM. has just a slightly different view on some of that. ⁓ It's better for the context, but Claude is going to be better for producing the code for the creative output. But Canva's going to finesse it.
Yeah. Megan Driscoll: Yeah. So talk about that. Let's walk us through.
Let's be specific. my audience is a lot of small businesses who are not working with they are. They're doing chat GPT for the very basic stuff. You say coding to them and they're like, ⁓ So talk a little bit about maybe you can simplify that, just one process.
Tamara Asselta: Right. Sure, so let's say I wanted to create a marketing pamphlet, a one pager. Okay, and the materials that I have right now is here's some of the language we want it to include, here's our branding toolkit, and this is the audience we wanna target. Okay, great.
I will put that same information into Claude separately and into Notebook LM separately and I'll see between the two of them which output I prefer. Or, you know, what are they telling me that maybe I want to combine into Claude. Or just, yeah, exactly adjust. So, likely what will happen, because I've done this before, this is an example I actually have done.
Megan Driscoll: Yep, or adjust. Tamara Asselta: I will take some of the pieces from Notebook, I will put them into Claude and I say, okay, I wanna go more this direction. And I'll say, create what this looks like. And so then Claude does the backend, they actually do the code development.
And what's important about that is because if you want to put that pamphlet as a visual on your website, you can. You can copy and paste that code and implement it onto your website. Now you'll have to know how to do that step, but likely somebody who's built your website would know how to do that. Or you can even ask Claude, you could say, how could somebody who doesn't know how to code make sure that the code gets onto the website?
You could start there. You could try it out. It makes it very easy for you to do this. So that's one way to do it if you're trying to get that visual onto your website.
If you're just trying to create a pamphlet and it's just going to be a link that exists on some other platform like Gumroad or whatever you're using, Etsy, et cetera, then you can put that then into Canva. You can say, hey, this is, you can download the PDF. You can say, hey, Claude, put this into a PDF. It'll put it into a PDF.
You can then put that PDF into Canva and say, I really like the start of this, but I want the design to be more XYZ. And so you can put in Canva AI, here's what I prompted to create this, but I want these kinds of colors, I want this kind of wording, and I want there to be a little bit more finesse, say mid-century design for this. And then you'll get a better product that way because Claude is, again, very, very good, very powerful with code. but I find that it gets a little limited when we get to the design part.
same with Notebook. Megan Driscoll: Yeah, well, I would say they are all kind of limited when it comes to design. One of the things I found, at least with Canva, if you don't feed it a baseline, if you don't feed it something that you're starting from, the AI is terrible, actually. ⁓ Tamara Asselta: ⁓ yeah, it's like all over the place.
You'll get close, but then it will put weird shapes in certain places and then you have to edit all the text anyway, but that is, it takes forever. So you really, I find getting to a starting point and that's actually maybe a good way to summarize something like Claude and Notebook LM is those spaces are just to get you started somewhere. Megan Driscoll: Yes. It takes forever.
It's not. Yeah. Tamara Asselta: It's really, and sometimes they are to finesse, know, they go both ways. Like I said earlier with lovable and ready, Claude tends to be my end output when it comes to coding, but not with design.
And we're just scratching the surface, right? Like there's just so, there's so much out there for these different workflows and. Megan Driscoll: Yep. Yeah.
Tamara Asselta: I would just encourage people to look up some of the ones we've talked about if you're not already familiar. I think most people, if you have been starting to use these, you've probably started somewhere in these spaces. But just look up. hey, this is a task that I do all the time, is there some kind of AI tool that can automate this process for me?
And I guarantee one does exist for that particular step. Megan Driscoll: Yeah, for sure. Well, and I mean, I there are like, 5,000 new products coming on the market every month. And they're saying that there'll be 5,000 new products coming on the market every hour in the next two years, which is, know.
Tamara Asselta: Yeah, it's crazy. Yeah. Yeah, yeah. And not all of them are, ⁓ you know, as complex and as in-depth as others.
Some of them are definitely doing smart things where they're grasping into a niche, right? And it's like it solves this one problem and that's perfect and that's great. ⁓ But yeah, I mean, honestly, even if you just put in a prompt in Claude and said, Megan Driscoll: Yeah. Tamara Asselta: What are the top 50 AI tools for creativity, templates, presentations, automation, data workflows, data presentations, ⁓ engineering, ⁓ interior design?
Any topic that you could think of, put it in there and then just start there. Just say, give me those top ones or only give me the top three of each one. and ones that are maybe less expensive for a small business. If you're a starter, you're by yourself or you only have two employees.
Maybe you don't have revenue yet. You're very much in the early startup phase. There's a lot of those tools doing trials for two weeks a month. Or they're really cheap right now.
So, yeah. Megan Driscoll: Yeah, and or they're cheap, really cheap. Yeah. Well, this has been an amazing conversation.
I have loved talking to you. I'm so excited about multiple things that you said, but particularly the facts. think that's like a really like easy, simple way for people to remember like what they should be doing every time and then just making it a habit, right? That of good prompting, which obviously is gonna have make better outcomes.
So, all right, so tell me a little bit of how people can find you. Like, where are you? What's your website? you know, how do you make sure you spell your name?
Tamara Asselta: Yeah, yeah. So my name is first name is Tamara T A R A not ERA. So ARA. Last name is a CELTA a SS ELTA and you can find me on LinkedIn just type in my name.
I'm the only Tamara CELTA that exists thankfully. So hooray. ⁓ And then I'm also I just started I just switched over my newsletter from substack to beehive maybe a little bit of a plug for them. But my new letters now called YouGrow and I also have an Instagram that I just started that's also YouGrow so you can find me in those spaces and then last but not least you can also find me through all those avenues on my website at stratusconsult.
com and it's stratus almost like the cloud but it's S-T-R-A-T-A-S so should be pretty easy to find. Megan Driscoll: And it's Stratus Tamara Asselta: Yeah, just, was consulting was already taken, so. ⁓ Megan Driscoll: StatusConsult.com.
Awesome. Thank you so much for coming. I really appreciate it. And we'll go look you up.
Tamara Asselta: Yeah. Thank you so much, Megan. Have a beautiful day. Megan Driscoll: You too.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.