Go back

#598 - How I Use Claude to Book 4-6 Meetings Every Week (Live Walkthrough)

57m 4s

#598 - How I Use Claude to Book 4-6 Meetings Every Week (Live Walkthrough)

Alex Murphy, a senior account executive at 30 Minutes to President's Club, demonstrates how he uses Claude to consistently book four to six outbound meetings each week without sacrificing his regular AE responsibilities or personal time. His approach is built on a structured "intelligence layer" that ensures Claude has all necessary context about his company, sales process, and book of business before generating recommendations. Each week, Alex exports data from HubSpot—including all company accounts, MQLs, engaged contacts, and historical deals—and uploads these as CSV files into Claude's project files. He also maintains a set of markdown playbooks that encode his outbound process, ideal customer profile, and messaging strategies, many of which were co-created with Claude and refined over time based on results. The system identifies hot accounts by prioritizing first-party signals like high content engagement or course purchases, supplemented by external triggers such as new leadership or funding rounds. Claude then generates hypotheses about prospect problems and suggests outreach angles, which Alex validates and executes using tools like Sales Nav and Gong. After each cycle, he feeds outcomes back into Claude, creating a flywheel that continuously improves targeting and messaging. Key practical tips include managing project storage by converting PDFs to markdown, pruning outdated files, and using dedicated projects for consistent memory. Ultimately, Alex emphasizes that this workflow is practical, not perfect, and encourages others to adapt it to their own needs.

Transcription

12007 Words, 62697 Characters

English
What's up gang? My name's Alex Murphy. I'm a senior account executive at 30 minutes to president's club I use Claude to help me book like four to six outbound meetings every single week on top of my normal AE work And I'm gonna show you how to do that So if you're like feeling like you want to try to use AI to book meetings But you don't know how this is how I do it and it works really well for me Listen, I am a full cycle AE so I got to close deals. I'm a dad So I can't work all day and all night. So this is not the perfect Claude explainer video if you want to go find that There's like a hundred of those on YouTube go find one You're gonna see me in this use a bunch of different processes and files and like Sort of terminology and jargon. I've got that built into Claude already in a markdown document format that you can actually put into any LLM so that you don't have to rebuild this you can just use what I've done So we're gonna include a link to that in the show notes So you can go grab it plug it into Claude and you'll be on third base eat the meat spit out the bones take what works leave What doesn't and and hopefully you know start to book some more meetings with less pressure you ready to jump in okay This is my workflow now this looks a little complicated. It's actually very simple I'm gonna walk you through the whole workflow and then we're gonna go beat by beat so the first piece to this is what I call the intelligence layer The big issue if you just go into Claude or chat Gbt or Gemini or whatever and just say hey book me some meetings It's gonna just pull all of its training data and take a wild guess at what needs to happen to book a meeting That is not helpful. So what you want to build and what I build the first thing that I do is I make sure that Claude has all the key information About my company and my sale and my book of business in order to then go do what I'm asking you to do So a couple of key pieces of data that I am feeding into Claude on a regular basis. I do this once a week So every week I am Sharing from HubSpot my territory my accounts my mqls my all my deals There's a bunch of playbooks that I'll tell you about as we go through here that are in the Claude project as Dot MD these are like that stands for markdown document This is the sort of the AI language show. I'll show you what that looks like. This is all intelligence in here then Claude has that and it runs a process that we've agreed on so it's gonna take that data It's gonna give me the hot accounts with the signals. It's gonna hypothesize problems. I can use my outreach And then I also use it to gut check that I'm not missing any key prospects at those hot accounts So that's like running giving me all the input I'm taking that input and then I'm actually working it and that's where I go to sales nav zoom info Gong combine them all together and then I go do my thing and as I do that There's gonna be outputs that come back meetings are gonna get bugged people are gonna respond to different stuff Positively some of it won't get responses at all and then I go back in and I feed the intelligence layer Information about what happened and it just creates this flywheel then the intelligence layer gets smarter My processes get tighter my data gets better and then the next time we go through it improves the output I'm more effective as a prospector feed it back in and so on and so on and so on and so on and so on and so on and so so that's That's how this whole thing works. Let's dive into the first step of it. Okay zooming in on the intelligence layer Which is not a whatever that's my term for it a couple things that I want my My data to have so I'm gonna actually walk you through click by click here again We use hubsbots so the first thing that I want my My AI to have is my territory accounts mqls and my deals before I even download those though an important note So here's clawed here's my buddy clawed it knows who I am welcome Alex I have a dedicated project where all of my 30 minutes to presidents club stuff lives So I've got projects for 30 minutes to presidents club and then I've got like other stuff here like my bills and my LinkedIn writing my this other project that I'm working on just in my free time like so all of this is happening inside of this Project which is nice because that means clawed has this memory bank that it builds over time So that helps the intelligence layer automatically I've got really clear instructions. So clawed knows like you are Alex Murphy's dedicated sales assistant I am an executive at 30 NPC. Here's the what I'm trying to do Here's the ground rules and this builds out over time - I had clawed right this for me So this is kind of my first step. I dropped in my company website I just sort of talked through some stuff and had it recorded and gave it to clawed as a starting place And then as I work there's all these project files and I'll show you this But this is the like specific intelligence layer So you're gonna see me dropping files into the project This is like the permanent memory bank that clawed will look at and never forget so the chats themselves are like very temporary the stuff gets lost easy anything I really want to remember permanently. I'm dropping into a project file So let's start with the hub spot intelligence layer So four different things I need to get and in fact two of them the account tearing in my territory are all in one sheet So the first thing I'm gonna go look at I'm gonna go into hub spot Here's my house by instance and this is my Alex's territory report. I am literally going to export This is Alex's tam. You know what I actually do though is I have all of the accounts and this is like probably overkill But again, this isn't about about being efficient. This is about what works So I am exporting all of my company accounts including the ones that are tagged with my name as the owner and some of these are are tagged with you know My other sellers on my team I'm exporting that I'm telling us about export that it's gonna email that to me in a minute So I'll download that in a second I do all the accounts because there are a couple of times where there's an account with no owner in it So maybe it hasn't been named but it has like meaningful signals and we just missed it in our territory I want to make sure I'm at least considering those. That's one reason to do that. The other reason to do that is my view is It's not that expensive like I'm not hitting usage restrictions on cloth You do have a limited amount of searching you can do. It's pretty light for cloth to sift through a whole CSV I'd rather have more information than less because I never know like what it's gonna find and so I could do you know 245 accounts that are in my name or I could do 19,000 accounts some of which are in my name some of which are not all of which I'm down for those to be candidates to to be accounts to be worked and I'm eligible to work them Given our rules of engagement So that's why I'm pulling all of it The next thing I'm gonna get is our mqls and so for us we have mql reporting So I have a little mql. I'll go to my engage context. I've got a little mql report here My actual like official like from marketing mqls. I work those pretty regularly So I sort of use that I sort of do that outside of the actual Like workflow. So yeah, for us what we're logging here We track you know 30 mpc makes a ton of content and people download that content and they'll you know Fill a form with their email and their title and their name and when they do that it signals to me That they are downloading our content. I'm selling our sales enablement program with our courses So if a bunch of people from an account are downloading our content I know that's a pretty hot signal some of those people if they're a high enough level in the sort of the right buyer persona will turn into an mql So that's a good list. I work those right away super quick. That's a little bit outside of this flow What I do want though like you can see I'm working a few here and the rest I move so quickly through them that that they are either Deqed or moved into an opportunity pretty fast What you're looking at here is this is everybody else. This is everybody from every account who has downloaded 30 mpc content I want Claude to use this as one of the signals that it looks at some of the other signals that I look at for which accounts should I work Those can be found on the internet. It's going to check on those in a minute This is proprietary to us and so I wanted to factor this in so I'm going to go ahead and export this sort of all these people that have engaged So that Claude can match the name of the company against the territory and start to tell me which of those is the most likely to buy from from Also or be interested in having a conversation last piece is the deals one of the downsides of Working for a long time in sales org is there's a lot of closed-loss deals There's a ton of valuable information in there But it also can create some noise and so what will happen at first when I started doing this is Claude would give me like Hey, you should work this deal and I'm like yeah, thank you cloud. I actually have a deal going with them right now So that I want to fill it to that noise out But I also know that people that closed lost six months ago eight months ago ten months ago Maybe ready to come back for a conversation and I don't want that data incorporated so the last piece I'm exporting from HubSpot is my own my deals so This is my current active pipeline These are like hot deals that are going right now But I actually want to just look at all the deals we've ever run on this product ever And I'm going to export all of these so that Claude has data on every deal we've worked the status of that deal And can factor that in when it's deciding which accounts. I'm going to work next So I've exported all these I told HubSpot to export these. I'm going to go to my email. It has emailed these to me So I'm going to download them And I'm going to bring them over to here is my next export great. I'm going to bring them over to my Claude project files cool One more waiting and there it is. All right I've downloaded this cool. I've downloaded these That the very first thing I'm going to do is put them into my project files so that Claude can remember them Now the you can see I've used 9% of my uh, oh those are zip files. Uh-huh. Just kidding. Let me open these up Cool bang bang. All right great You see like clot there is a certain amount of storage So every once in a while I'll go through and prune old files that I'm not using I don't worry too much about it until I'm up to the like closer to you know 60 70% of the project used It's a lot of like CSVs and MDs which do not take a ton of space what you want to be careful of with this is like PDFs That kind of stuff is like very expensive So it's better to drop a PDF into a chat and then have it tell you what's in there and turn that into a markdown document and pop that in your Project files just as a way to like manage your space that way you don't have to spend a bunch of time taking through but you'll see My most recent exports are here my most recent engaged contacts are here my most recent accounts are here I am going to get rid of the old versions of this and you can see like this These are my three most recent I do want to kind of get rid of the old ones only because I don't want hubs for it to get confused So this one's dated May 18 I'm gonna ditch these Get rid of these because I don't want Yeah, I don't want it to get confused about what it is that we're actually trying to do And start feeding me old data. So okay Cool it has my my most updated data awesome So my intelligence layer is filled out the other thing I have here are Playbooks that will tell cloth these are like the instructions for the project these are gonna tell cloth what to run So I have playbooks for my outbound process I've got playbooks around ICPs that are ideal folks in like what are the main problems that those people are trying to solve with us I've got a very specific play the health and service problems I've got a messaging matrix that helps me attach the right sort of solution message to the right person Based on what we know about the account and then as I book meetings the Emails the phone calls the LinkedIn messages whatever that book those meetings. I also save those into cloth periodically so that it knows what's working and can Can make recommendations on that basis? So some of these I build as we go the outbound process document. I'll actually show you what this looks like so these are all like I mentioned These are all mds and so if I look at Here's my outbound OS markdown document So this is 209 lines you can see like what this is clawed created this for me. It's not like I typed this out I basically typed out here's what I want to do and then every time I run this session I update it and I have called update the document so This is my operating manual for booking meetings one workflow on Friday the execution and then here's like my goal Here's the inputs. Here's the outputs. Here's what it should do. Here's the steps. So I don't have to tell cloud what to do every single time it knows So you can see like I just I can type in I'll do it right now I'm gonna have a run outbound and it's gonna start running this process I'll get the weekly prospecting prep going let me check the data for files first Let me stop it because I want it to do that. Yeah, but you can see like I've got it trained to where it'll immediately go So that's what that outbound process is all of these are marked on documents. So who buys and why that's something we fill out over time Who at the org what type of persona is most likely to buy from us for us? It's sales leaders like directors vp's plus of sales or Enablement leaders people who are responsible for training the sales team will partner with us normally. It's a combination of both What's nice is as we go because I'm feeding clawed data about who's booking meetings? I have an increasing body of data about like who is most likely to book meetings and which of those meetings are most likely to close So it's starting to understand like who should I continue to target like as we go if I start to see like ooh Enablement leaders are booking meetings way more often than sales leaders That's good information for me to have so I can be really selective about how I spend my calories during the outbound Sessions in my week. Let's talk problem-prompter and messaging matrix for a quick second. So these This is a shameless plug, but it's legitimately what this is you know a 30 minutes to presidents club has this really bad-ass program with courses on Cold calling cold email adjacent bay and so there are a ton of rich data here I watch these courses. I took these courses. It's important for me to know the principles here in my brain I don't want to just outsource that to an AI But it also really speeds things up if the pieces of this that I'm gonna use most frequently are In-clawed so it knows that I'm asking for so there's two major examples of that the first one is problem-prompter So this comes from Jen's course Her very first section of her course is about what it takes to get an executive's attention and this is like very directly related to prospecting and so she talks about building your executive messaging and In the next lesson she talks about getting executives to reply and so I watched this whole thing I absorbed it be you can see there's a lot of good stuff here And there's even a transcript so what I did is I actually just opened up the transcript to hit command a on my Mac copied everything here Dropped it into a chat and said this is the problem-prompter play ingest this because we're gonna apply it to 30 NPCs sale Club doesn't need me to explain what 30 NPCs sale is because it knows so I did that and it created a markdown document Which I saved into the project so you'll see me run problem-prompter. I don't explain to club what I'm after It's gonna know because not only do I have a given it Jen's course information, but I've also then Refine that every session I basically refine based on what I'm learning so you got your intelligence layer It knows all about your business. It knows the process you're trying to run the next thing is to actually run it So there's basically four things that I have clothed do the first thing I have it do is assimilate all that data So what I'm really looking for here are two things Clot already knows based on the previous what are the signals that I care most about and What are the attributes of companies that I'm trying to reach out to for me? The specifics of that is I want to reach out to companies where they've got a lot of people downloading 30 NPC content from our Outreach we kind of know if nobody that count as downloading our stuff They're not gonna book a meeting with me what I'm looking for really is like that nice rich first party That's the most important thing There's also supplementary signals like hey you've hired a bunch of people or you had a new merger and acquisition or you've got a new CRO or a new head of enablement who came in somewhat recently those are all good signals a Fundraising round those are good signals that support the sort of main thing which is do you have a bunch of 30 NPC fans who are actively using our content? That's number one Today's show is brought to you by a line the a ideal workspace that helps sales teams close complex deals So here's a tip about multi-threading there are stakeholders in your deals that you don't even know exist like the VP of Finance Who's viewed your champions recap twice this week, but no one actually introduced you that is the type person who's Beacon around your deal and might sneak up to kill the deal at the finish line and Aligns a i surfaces it before it's too late so you can multi-thread to them We built a full multi-threading playbook powered by a line. It's free in the show notes Today's show was brought to you by outreach the agentic AI platform for driving execution across every stage of your deal Great sellers drive velocity at every stage of their sale and one thing that's worked well for me is Assigning red line deadlines to my prospects when we kick off vendor review even artificial micro deadlines for things like first cut of Red lines and security review helps my prospects get their internal team moving We built a guide on how to help you drive six and seven figure deals with our friends at outreach get it free in the show notes The worst discovery advice is just be curious and ask lots of questions look our prospects are not showing up to sales calls to answer our Questions they're showing up to get their questions answered and to figure out how we can solve their problem And this is what pipe drive does if you want to get clear complete sales visibility You can use pipe drives easy to use and customizable CRM that is designed to be simple from day one and stay easy to use as your team Gross we actually put together an entire resource hub with pipe drive to help you get to precedence club And you can get them for free in the 30 NPC pipe drive closers hub the link is in the show notes even better than downloading content It's like did they buy a course have they bought a copy of cold calling sucks, and that's why it works like I'm looking for that stuff Well, Claude's looking for that stuff and then attributes is pre-baked This is I'm looking for employee size so for me. I know my sweet spot is kind of in that like mid market kind of like 515 hundred or so employee range like that's pretty good You know b2b companies are definitely a sweet spot for us tech is usually even better than something that's like a non-tech so Again, I've kind of built that out over time, but Claude knows all this. It's assimilating that data Then what it's going to do is it's going to surface hot accounts. I'll have it do 10 accounts You know depending on the week I may be looking for 20 25 sometimes it's I just need a little top-up So I'll have to do five, but it's default is going to be I think 10 and it's going to surface those with the signals and the reason that it's surfaced them The other thing I will have it do is hypothesize the problems that those accounts may be dealing with And you know, so okay, we know what's true about them What does that probably mean? They're dealing with this is where Jen's problem-prompter and Jason's messaging matrix really comes into play Claude's using that to then answer the question of like what problems should I lead with in my outreach? And then the final piece is gut checking prospects. So Claude has what's called an MCP Couldn't tell you what that stands for, but basically what it means is Claude can talk directly to zoom info who provides all of our data And I'll have it gut check once I'm done adding contacts Hey, is there anybody I missed and Claude will go through and use it zoom info connector to say hey You maybe miss this person or that person. I've like found legit contacts who ended up being my champion that I missed on my first pass that Claude caught on the the backend that then book meetings with me. So that's a very important step. One thing you're not seeing on here is writing the emails. At this point, I do not have Claude write my emails for me. And honestly, even like hypothesizing the problems, I'm not really relying on that anymore. I used to, and then basically you do it enough. You start to notice the trends and the whole scope of like the problems I could hypothesize, the way I talk about our solution, like that starts to zoom in, zoom in, zoom in, zoom in, zoom in, zoom in. And for me, I don't have that many named accounts. I have like 300 named accounts. So I don't have to like do that. It actually is less efficient to have Claude do that. Then for me to just rely on what I know, which is if an enablement leader has sales reps that are downloading our content, that's a really good sign. Almost every enablement leader is overwhelmed and has way too much stuff to do. That's the problem I hypothesize. Our solution is pretty clearly addressing that. And I'll have a pretty standard CTA at the end of my email. So I don't need to rewrite every email. But that's my sale. I used to be, my first BDR job, I worked for a legal tech company, and there are literally like half a million law firms in North America that I could have sold to. And the signals were way less proprietary first party data. And the first party data was like really hard to find. It was a lot of clicks to find it. And so it made a lot more sense. Like if I had had this then it would have saved me so much time because it's such a huge book of business. And there was no designated territory. So it would have been way faster for quad to produce that because there's a whole bunch of different products we could sell. So there's way more solutions, there's way more problems. There's a lot of different personas. So I'm going to run this now. And then we'll talk about what to actually do with the data. So here we go. Let me go into my project, start a new chat. Here it is. I am going to start and I'm just going to say literally run the outbound process. Claude is thinking, while it's thinking it's going into, oh look at that, I'll run the weekly prospecting prep. First, the required input check. It's doing what I've told it to do, which it's going through and it's reading those CSVs I just dropped in. It's looking at the outbound process. It's looking at the data I've given it around, which signals to find. It's going to pull those signals and it's going to hopefully feed me 10 accounts that are really solid and have good signals. And then we will work those accounts. So again, Claude's thinking it's running some command. It's filtering out active pipeline. So I don't want to see any deals that I'm actively working. Obviously that's wrong. It sees that the deals export is from today's date. Cool. It's scoring signals from the account. So it's like finding the, looks like we have 80 candidates. And it's finding which ones have the most density of signals. And then it's going to generate account notes for me. Cool. Now it's sifting out, verifying the employee account numbers, which it's using, I believe it'll be using Zoom info for. It might be doing the internet search. We'll find out. We've got downloads count. So it's looking at who's like at how many people are actually downloading. It's giving me the batch. Here it is. All right. So I'm going to just eyeball this. What it just spit out for me is 10 accounts. Tier one, this is not an official tier. I don't really tier my accounts because again, I got like 250, basically the tiering for me is like how many fans do they have? So it's very simple. You, you, you may have like a tier one, tier two, tier three account. It would be good for cloud to know that. So when it says tier one, it's just talking about its own tiering. My whole process assumes that there's mistakes here. And we're going to go double check them. So for example, 800 employees, 1000 employees, 2000 employees, those are extremely round numbers. Probably not totally accurate. So I will be double checking these. But this has taken literally, I think it was 19,000 accounts and we did it down to 10 that it thinks I can work. So at a par, 800 employees, it scored at 6.0, which just means that like score high, this is a, I don't really care too much about that. It's kind of cloud language. They just raised a series G recently. They've got a growing enablement team. This is the name of an individual that we should blur out who used to work with me at a company that cloud knows I used to work at. That's pretty relevant. I'll go take a look. They've got funding, new enablement hire. That's a really strong cleanest stack in the pool highest priority. I won't go through every single one of these. I'm going to double just gut check this before I go work it. So at a par, I'm going to go into my CRM. Go into the TAM here. Let's just make sure that there's nothing weird about this. OK, so this is in my name. Ah, look at that. Account refresh date March 18, which means I started working this two weeks ago. So this is a great example of cloud not knowing everything it needs to know and enriching the intelligence layer over time. So March 18, I mean, that's three months ago. That's really recent. They also have one form submission. That's actually like a very weak signal. That means literally one person one time downloaded a form from us. I disagree with the cloud that this is a good account. So I'm actually going to a user tool called Whisper Flow, sponsor of 30 minutes to present. And I actually like to talk to cloud for me personally. That works really well. So I'm going to just dictate to cloud right now. At a par is not a good fit. They only have one form download. And the account refresh date is March 18, which is very recent, given that today is June 2. So that's a miss and not one that we should work. We should look for accounts that have not been worked at all that are in my name and not assigned to David. And we should look for accounts with a large number of form downloaders. So we're going to need to regenerate this list in a moment. But also I need you to make sure to update the outbound process doc to reflect that so we don't make them as big again in the future. And then my little dictation tool is going to pop that into my channel here. Great. Bingo. I'll get rid of this because that's my note to you viewer. Let me just do a quick sense. Check these two and see like, all right, cool. This is 1,000 employees. 14 content downloads on the account. Now we're getting somewhere. There's an enablement persona fan. Hey, OK, they downloaded the outbound sequence template. That is really, really rich data. Let me gut check that real quick. OK, no account refresh date, no account, no 14 submissions, beautiful. So I'm going to set the account refresh date today. This is where Cloud is great for sifting data. I do not use it as my data source of truth for a few reasons. It's really easy to lose key pieces of info in an LLM like Cloud. So I'm always updating my system of record, which in this case is hotspot. And the other thing is, Cloud's hard to query. So even like we're in this chat right now, but if I go back and go like, hey, what was the deal with SuperSide? Why do we decide to work that? It's going to take Cloud forever. It's going to burn a bunch of tokens. And it may or may not get me a good answer. It's certainly like, if somebody from SuperSide picks up the phone on a cold call block, and I forgot why I'm reaching out to them, there's no way I can query Cloud. So what I'm going to have it do instead is I've got this little account note field. This is where I keep my super tight. I can look at that in one second after somebody says hello on the phone. And I have what I need to actually book a meeting with that person, or at the very least explain why I reached out in the first place. So yeah, now I'm going to just double check SuperSide. Let's see. I'm going to go look at the engage contacts and just double check with a little quality that we actually have what we said we had. So then I'm looking back at these are people who downloaded. OK, so I have only one downloader, but this person's an enablement person. And they downloaded nine form submissions, most recently a couple months ago. That's like a pretty rich signal. So I'm going to then, yeah, great. Cloud should produce account notes for me. So I'm actually going to hold off right now. But that's that's some like how I'm thinking about this. So again, right now we're assimilating data and we're servicing hot accounts. All right, I'm going to do a you can like we can fast forward quick through this. I'm going to go just do a couple more just to make sure my real goal here is to make sure that there's not some other kind of mistake being made. I know if there's, you know, there may or may not be people who have a similar like read account refresh date. But there might be other mistakes that Cloud made that I just haven't thought about. So cool. So I'm going to go just do a quick sense check on the rest of these and make sure they're legit today. I sure I didn't show up. So here's a great example of this is a hot account. It's in my name for whatever reason it did not show up on my CRM like Tam report. I don't know why. Good news is I don't need to figure out why that's Revops job. I just need to work the account. And so this is another reason why I pull all the accounts in our CRM rather than just the ones that are assigned to me because you just never know what you're missing. And it does look like we've got some recent people that have downloaded. I'll just double check to make sure. OK, just one. It's not great actually, but apparently this person-- interesting. OK. Cool. Here's another account in a condo. Some of them in a condo had a kid in the same swim lessons as my kid. Probably won't mention that in my help on email. But look at that. All right, cool. We've got a refresh day. Great. Cool. And I do have some like form submissions, etc. Great. All right, cool. I'll set that one for today. I actually get this for guys doing so soft. I can't refresh today. Good. Cool. All right. So that's the tier ones, and then it's got tier two's. Single strong signal, high download volume. I actually think that's a stronger signal, frankly. Claw is like discounting it too much. I may or may not tell it that, you know, it's probably good hygiene to do it, but you always, I'm always making the decision of like, is it worth the effort to get Clawed perfect or do I just need it to be good enough to work? And the answer is usually I just need to be good enough to work. So okay, we're going to accept this one. Cool. So so far I've only rejected one of these 10, not bad. All right, let's go here, bang, bang. Cool. So yeah, I would normally finish this list for today. I'm going to just move on from here. Now Clawed is giving me some flags, make sure I verify some of the data at grade, the download accounts are a little weird, cool, and then it gives me the next step. All right, cool. So I'm going to tell it like, don't worry about it. Don't worry about populating more accounts for today. Here are the accounts I'm going to work that you can add to the next step. Super side, insight, software, and a condo, plural site, but we can leave the rest for later. So just pull them out for now. Now it's going to hypothesize some problems and it's going to also like, should give me like a little account note that I can add to my CRN. That nice thing is again, back to that like system of record thing. If I go back to my report of companies, if I sort by that refresh date, these are the ones that I've worked most recently, I should see a few, I'm actually missing a couple that are getting filtered out for whatever reason, probably because they don't have the right number of employees. There they are. All right, cool. No problem. So bang, bang, bang, bang, bang. Great. While it's doing it, it's diagnosing the problem that led to that account that got surfaced in appropriately, which is cool. I don't need to diagnose it. It's going to diagnose it for itself. That's one of the advantages of using the more expensive models, expensive in terms of tokens because they're smarter. They'll figure that out. I'm going to let it do its thing over here while it's doing that. I'm actually going to move. It's going to hypothesize problems. I'm going to actually get started on this next step, which is going into sales now. I'm just start to populate people because I know that's the next thing. So while Claude's working, I'm working as well. This is another like efficiency game thing like you can kind of work in lock step. So I for my prospects, I'm adding prospects. I basically got two kinds. I've got enable leaders and I've got sales prospects, like sales leaders. So I've got a sales and I have search that's a Boolean string with every possible job title of the sort of person I would want to reach out to an account. So I don't have the actual account loaded up. I literally just have these job titles and that's it. And so what I'll do then is I'll go like manually add in the accounts that we agreed to work. And I can't a superside plural site in site software. I'm going to add all those right now. Accurate. And a conda. In site software. Great. Plural site. Cool. And then there's one more I forgot. Super side. This is giving us a bunch of enablement leaders. So I can spot check these. Because he's pretty accurate. Here's a sales enablement manager. These are all legit people at these different accounts. So that's cool. Great. Let's see where Claude's at. All right. I'll look at that. It went and updated my prospecting play for me. That's so nice that it did that. You know what? Before anything else, I'm going to actually just download this updated thing. I'm going to change the date to reflect today's date, which is just my own little numbering system of health and keep things straight. I'm going to go to my project. I'm going to drop in the updated one. Cool. Bingo. There it is. And then prospecting. Pret. Only find the other document. Where is it? I just searched. I just controlled. I have to find the name. Oh, there it is right there. So I've got my outdated one. My updated one. You can see this one has 204 lines. This one is 205 lines. But more importantly, here's the date 2026 June 2nd. This one's date is 2026 May 12th. I'm just going to get rid of this one. Cool. So it's not perfect, but it's good enough. Okay. So great. Awesome. Cool. Now it's going to tell me what's the next step. Cool update is done. So we got this batch. Cool. The other piece of info that I want to give it is I'm going to add these prospects in just a minute. I want it to hypothesize the problems just to show you what that looks like. And I'm going to have a gut check the prospects while I am searching for them. The other key piece of info is how many people from LinkedIn are following 30 minutes to presidents club on LinkedIn. LinkedIn pretty aggressively blocks like AI bots from searching its data. And so this is one where I'm doing some manual work here to like give the best number. So I actually got ahead of myself before I even go in and look at the enamement leaders. I actually just want to look at everybody at these five accounts who are following my company and there's 62 of them. And so I'm going to go I'm going to go one by one and I'm actually going to just tell Claude the number at each account who are following us. And that way Claude is going to have that little extra piece of data. So I'm going to dictate that real quick. I'm going to really get it. Followers. Add to single followers. Add to six followers. Suicide. It's like a three-color. It's like 14 followers. Add to single. Now look at this. Look at that. Combine downloads plus followers. That's a really strong signal, which is very, very cool. Okay. The next thing I wanted to do, hey, want me to return, run the account research now? Yeah, run the account research in the problem prompter, please. Output an account note for each of these accounts. Oh, right. So now it's running the account research and the problem prompter on these four accounts. It cut out one of my accounts. I'll go back and fill it in later for the purposes of this video. Great. So what it's doing is it's actually pulling the named people who have downloaded stuff at those accounts. It's also, oh, look at this. Would you like me to enrich and have the engaged contacts? Yes, I surely would. So this is that MCP for ZoomInfo that I was talking about where it's, oh, look at this. We've got some people here. What it's doing right now is it's checking our data, the downloader name against ZoomInfo to see, is that person still at the company? Because if I'm about to go email a VP of sales and say, hey, this person downloaded our content and that VP says, who? Oh, they quit. I don't like them. And in fact, the fact that the download of your content means, I don't want to work with you guys. That's probably good for me to know. So yeah, Cloud is going through because it has that connection to ZoomInfo and a lot of different engagement tools. And like data tools have these MCPs. It's using that to validate that those people are still there, which is very cool. I forgot that I told it to do that. Pretty important. This is a great reason to do this and like iterate the intelligence layer is, if it was up to me to remember every single piece of this process, it's definitely stuff I would forget. But because I like how this iterative process where Cloud is learning with me as we go, it catches things that I would, I remembered one time as a good idea and then I forget the next time because it's busy. And, you know, I don't sleep a whole ton because my daughter's, you know, a toddler and, you know, so that's just what it is. So this is the beauty of this one. Cool. So now what it's doing is it's searching the web for that sort of list of signals that I've already given it that I know are important and it's sort of building an account note with some key research for each of those accounts. And I'm going to go in to my Sierra, I'm going to fill those in once it is finished. Today's show is brought to you by Responsive, which helps sales teams respond to RFPs, security questionnaires and buyer requests faster so deals don't die at the last mile. One great way to drive your deals faster is to run parallel processes instead of linear processes. For example, in vendor review, work with your buyer to kick off legal security and procurement review simultaneously instead of doing them one by one sequentially. We built a playbook with our friends at Responsive on different proposal accelerators that you can use to close deals faster. Grab it in the show notes. Today's show was brought to you by Outreach, the Agentech AI platform for driving execution across every stage of your deal. Great sellers drive velocity at every stage of their sale and one thing that's worked well for me is assigning red line deadlines to my prospects when we kick off vendor review. Even artificial micro deadlines for things like first cut of red lines and security review helps my prospects get their internal team moving. We built a guide on how to help you drive six and seven figure deals with our friends at Outreach, get it free in the show notes. The worst discovery advice is just be curious and ask lots of questions. Look our prospects are not showing up to sales calls to answer our questions. They're showing up to get their questions answered and figure out how we can solve their problem. This is what pipe drive does. If you want to get clear, complete sales visibility, you can use pipe drives easy to use and customizable CRM that is designed to be simple from day one and stay easy to use as your team grows. We actually put together an entire resource hub with pipe drive to help you get to precedence club and you can get them for free in the 30 NPC pipe drive closures hub. The link is in the show notes. Look at this. This again is more than I can ingest like while I'm in the middle of a cold email block or a cold call block. But this is really helpful for me to skim and read before I do that. I just kind of have it in my mind. So we don't go through all these but like here's plural site. Here's what's going on. This is all available on the internet except for the content downloader stuff. So I've got a strategic objective that it feels really confident in. Re-excelerate new bookings and retention after the balance sheet reset. I don't know what that means. Looks like we got cool. All right. They went through a debt. recap, all right, cool. So they're like shifting around some other financials. They're in a retention and growth re-exceleration push. That's a really positive signal for us. It means they're hiring a lot, which is great because that, I mean, that creates this issue where it's hard to onboard a bunch of new hires and we help with that. So this is already keying me into a really specific problem that I think I can attach in my outreach. So the sales org is under pressure to grow. Bookings with the same or leaner. Headcount, okay. We've got a bunch of people that are being like enablement, gets measured on whether the skills transfer are cool. What's the status quo? This is the next piece of Jen's problem-promptur thing. Like what are they probably doing now? They're a learning company themselves. So they may learn not lean on internal content. It's good for me to know that's probably an objection I'm gonna get at some point. So I should be prepared for that. I wouldn't really have thought about it except for this process. What's the 30MPC solution here? Like, okay, maybe they wouldn't buy like courses from us, but if we position this as the sort of 80% practice, 20% teaching model, that's a really strong positioning. And so when I'm sending emails, I'm probably not gonna talk about, hey, we have courses. I'm gonna say, we're an enablement partnership that helps you reps permanently transform their behavior to hit these really aggressive goals that you have. And we've got a CRM one-liner. Cool. So I'm not gonna run through every single thing here. And then look at this. It's given me these beautiful little account notes that I can pop right into my CRM. So here's the one for Pluralsight. Bang. Cool. And it's just literally it's given me like the top line thing that I'm gonna say when somebody picks up the phone to get a conversation started. So I'm gonna go add these into my CRM. I will double check these. And so it's important, like, I'm not gonna do that in this video for the sake of just you watching me, but it's always important like go double check and make sure it's legit before you start citing it to people. Quick zoom out. Here's what we've achieved so far. So again, we've got our intelligence layer which is getting more robust even as we iterate as we run the process. But we've told you everything about my current book of business for my CRM. I've got all these different processes to help Claude very quickly and efficiently do what I need to do in an outbound block. We fed all that in to the actual Claude Central engine. It's brought that data together. It surfaced to the best accounts. I hand picked a few of them. It went and then hypothesized the problems that they're dealing with. And it surfaced nice little account notes. And then I went in and rehydrated my CRM with that data so that it's easy for me to quickly pull. And I'm not relying on Claude to try to go query it in a "Hey, he's when somebody picks up the phone" or whatever 'cause that's deeply inefficient. So the next thing to do is start actually finding, "Who am I gonna reach out to at these accounts?" And so that's what I'm gonna do. So this is where now I'm going to use this sort of prebuilt sales nav Boolean search. So quick primer on this, you could do this a different way. You could go add in the functions and say sales and sales development and you could add seniority level and say I'm looking for VP of sales and certain keywords or whatever. I find that the most precise thing to do is a big Boolean string, which surprise, surprise, Claude showed me how to do. And let me actually go find my Boolean string so you can see what it looks like. Okay, cool. So yeah, that Boolean string is just a bunch of different possible job titles and LinkedIn headlines and it's searching for all of them. So now I'm gonna go add those companies that I was looking at before. So cool and a kind of superset of Pluralsight and so I saw for cool. Great. Cool, bang, that's seven people. Awesome. You know what's awesome here? I don't have to go one by one. I'm batching this, which is gonna make it way quicker. So we've got people, we've got revenue enablement. Great, great, great, great. Awesome. So this is where I open up the Zoom Info browser extension. So I've got this in sales nav. I'm gonna use sales nav for my actual prospecting or not yet. I found the key contacts next is to enrich those contacts with the browser extension of Zoom Info, which is really nice because it will just take a look at sales nav, see who's there and very quickly take those folks, check it against the Zoom Info data, and then I have it connected to Gong. I use Gong engage for my outreach. It will then send that to Gong and it's gonna give me the prospects that I'm trying to get in front of. All right, so let's do superset. I'm gonna have it just for demonstration purposes. I'm gonna have this, I'm gonna have, I'm gonna add a couple of people from superset. So I know that this person downloaded stuff from us and Claud knows it too. So using the Zoom Info MCP, I'll just have it, we'll look at flags and stuff too. This is very important. So okay, this is a little bit old, good to know. It's saying I should go do a quick check, which I'm happy to do, but for the purposes like, can you give me the email address of the superside contacts? Carolina, Todd Nick, and Tamara. So that's not all right. Quick little note, if you're trying to budget, you're doing a lot of this, you're gonna budget your claw tokens. It is more efficient to, like if you do something and it's like giving you the wrong result rather than stopping it and doing a new prompt, just edit the old prompt and rerun it. And that uses less. So hey, look at this, I have some, based on what Zoom Info shows me, I've got some people. So I'm gonna go actually just add these folks directly into HubSpot, I guess. Here's no way, great. (silence) Here's where we're at. So we went, we found the hot accounts. We found the folks in sales nav. We used the Zoom Info browser extension to add them into Gong Engage. Now they are in Gong Engage and it is time to actually do the awesome step of sending them a legit email and a legit link in message. So I'm gonna go to Todd Nick, who is a person who downloaded our content. What I like to do is now I'm here in Gong Engage. I actually like to go take that little account note and paste it right into the prospect notes. So when I'm calling, it's easy to just be like, hey, what's going on? And like I've got the info right there. So there it is, bingo, cool. Now it's time to do my next steps. So the way I've set my flows up is I have an enablement flow and a sales leader flow, two different flows. The first message that I like to do is a LinkedIn touch. And sometimes I'll go try to find a warm intro. This person downloaded content directly from us. So it's already very warm. And so I'm gonna go do a double tap. I'm gonna send a quick email. I'm gonna send a quick LinkedIn connect. And so with Gong Engage, here we are. Great, here's this person on LinkedIn. They are, who they say they are. Very cool. I'm gonna do a quick scan here. Sales and enablement specialist. Da, Cape Town, okay, interesting. I'm just looking for like, can I, is there some little PS I can mention? Nothing too crazy. Okay, cool. But I do have a good sense of who this person is and what they care about. I could go like, look at, if I wanted to really dive deep on this person, I could go look at all their interactions and go look at, like, what did they comment on lately? Interesting. And AI Serbian cowboy. Okay, very fascinating, right? So like, I could, I'm probably not gonna mention this 'cause that would read as weird, but it is helpful to just be like, is there any major like company news that I can attach to? In this case, there's not. And that's fine. So I am going to, mark this as a complete and then I'm just gonna go bang, bang. I'm gonna load up my little note. So I'm gonna connect. I've got a preloaded note here. It's templated, it's personal, it's accurate. And I'll, I'll send this right after I send the emails. Let's load it up, that's ready to go. And now what going to gauge is gonna do is gonna think for a second and it's going to give me the actual email task. There it is. All right, so now I'm gonna write this email. Great. And I'm doing this for one individual person. The nice thing is engaged. It'll be bucketed and batched. So when I write one supersight email, it's gonna be pretty similar to all the other emails that I write. So you can see this is my actual email template. This is built and based off of the learning from Jason and Jen's course into 30 NPC enablement programs. So it's already like a pretty solid email. Now there are some situations where I'll go a level deeper and I'll do like a real deep dive, super personalized email. This is like hot account, number one account, super warm, best person I would love, nothing more than to get them into an opportunity. I'll go a level deeper. Everybody else, this is a really solid email and I know it is because it's booked meetings for us. So in this case, I'm going to, I'm gonna modify this one. Our normal thing is like, hey, not sure if you knew but you got a bunch of fans. Well, in this case, actually, I know that this person downloaded some specific stuff from us. And so let me actually pull up here, gray. I'm gonna go back and like actually do my little check to see like what specifically did she download? And I can even ask Claude, hey, what did Todd Nick download? And Claude knows the answer to that. So that's beautiful. I can use that to like really personalize my note here. So I'm gonna go back and like actually do my little check to see, to change this up and be like, you know, they have, I think it said, I think we said 21 others. So cool. I'm going to update this up on sequence template on 317. Okay, cool. All right, cool. So I'm actually up to this. Saw you were digging into our outbound sequence templates and it looks like 20 others at SuperSide are following. Content. Cool. And then up here, I want to make sure my subject lines great. I've got this 30 NPC SuperFans. I'm just going to say like, cold, you know, cold outbound sequence plus 30 NPC SuperFans. Great. Here's my problem. We saw we have a signal. We attach it to a problem related to the enablement persona. I don't have to change anything about this. It is accurate. It's specific. I've got a solution here enablement teams will work with us. Nothing weird. You know, this is where maybe for this other account that's like a it is itself, you know, a learning development platform. I might tweak this, but in this case, this is great. Cool. And then I'll probably add like because they were looking at our outbound sequence templates, PS, I'm going to actually like do a more humanizing PS for what it's worth. Our program explicitly teaches every one of your reps. Yeah, a deal way to structure. Now bound sequence. Cool. Great. Bang. I got a nice CTA nice problem, nice solution, nice signal. This is a solid cold email. Great. Do a quick double check. Bang. Send it. Cool. Bang. And then I've got great. Cool. I just emailed you. Hey, hi. Bang, bang. Multi channel. Cool. In two days, I'm going to call. By then, I will definitely have forgotten what this whole thing was about, which is why I've got my little account note here. I know. Okay. Ah, yes. Yes. Yes. Outbound sequence. Simple. Great. Cool. Which I could have done there. So that is that last stage of actual execution. In a couple days, I'm going to make a call. By then, I will definitely have forgotten everything about this person, this company. Maybe I'll remember a little bit because I read through the problem prompt or deep dive. But for the most part, like I need to be able to run this really efficiently. So I've got my little account note. And then the actual sequence that I run, I'll show you. It is, you know, you can see here, I've got my flows in going. So this is my enablement leader flow. This is my sales leader flow. This is an old flow that I don't use anymore. So for my enablement leaders, there's 12 steps in here. So I've got my day one linked in an email. That's what we just did for this person. Day three, I'm going to call. I'm going to, I've got a little auto email. This is just a bubble up bump. Great. And got the the threat clause. Let me know if I'm off. And I won't call you again. And she, you know, it's a fun little thing. And then I'll do another LinkedIn touch here. If they've accepted, I'll send a note if they haven't, I might go engage with some of their content. And then wait a couple days. And again, triple tap, column, another bubble up email with a case study. So just a little bit more. And then like another LinkedIn touch, wait a few more days, four more days. Boom, auto email, call auto email, auto email. And then they'll be done. And the very last one is just like, hey, here's another case study. And we'll leave it right there. So now they're off to the races. What you just saw, a lot of this is automated because we've done a ton of work to validate that this is a sequence that gets us a really good yield on our inputs. So this whole process, this whole workflow takes the heavy, expensive, costly time consuming process of pulling all the data and structuring it and finding the right people and finding all the signals and figuring out the account. And it makes it really, really, really efficient. You've seen the various slowvers in today. But when I'm doing this, I can sequence 20 accounts and like do all the research and get everybody sequenced in like 90 minutes, maybe once a once a week, I am getting all of the prospecting inputs on Friday that I'm going to need to then run for the rest of the week. And then in gong, you can see like then all my to-do is be loaded up here on, you know, Tuesday and they're great. These are all people that I've loaded in. Here they are. And here's all the steps that are required of me. And I can just run that process. So it's a little slower at the front end, but what you're getting is a much smarter process that actually is faster in the long run. The very last step is iterate the intelligence layer. So this already we saw happened. So I won't have, I'll finish this process before I do this again. But you saw that cloud went and updated the process and gave me short cut. I can just see the output here. It gave me this little dock, which has an updated, you know, this is the this is the markdown document for the weekly prospecting prep process, my outbound process. So it updates this. At the end of a session, I'll go into cloud. I'll just say like, all right, please update all relevant process docs to, you know, not repeat any of the mistakes that we made this time. And ingest any changes or new learnings about how to make the process more efficient. Clouds going to go think it's going to think a little bit and it's going to do the work of updating my process for me. The thing that I have to be diligent and disciplined about is doing this every single session. Every time I do this process every week, I need to do this because if I forget all this valuable rich context about what not to do, how to improve the process with the output should look like I'm going to I'm going to miss that and we'll have missed this really critical step of iterating the intelligence layer, which deprives me of this awesome like flywheel that works really well. So cool. Here it is. Hey, look at mistakes this session. We made some slips. We have inefficiency. There was a Boolean search that we had to like talk through. Like so it's it's finding. I don't have to tell it what to do. It's going back through our chat that we just did and it's finding everything that should be better about the process and now it's updating the process docs. So when I do this in a week, it'll be smarter and it won't repeat those errors. Maybe it'll make new errors. Maybe it'll you know, whatever that, but that's the whole point is we iterate as we go. And this can really absorb changes in our data too. Like if I all of a sudden realize like, oh my gosh, there's a person I'm missing. I don't have to create a whole new process. I can just fold that into this process. So you could see what it's doing right now. It's like writing rules into the document. It's going to give me that nice markdown document, which I'll then drop back into the project. And then we'll be done. Quick reminder, before we go, all of the process documents that you see me update, including the one I just did, the updates I just put in there, that will be in the show notes for you. So you'll be able to download that, pop it into your LLM of choice, and you'll be starting with all of the work that I've already done. And then you can tweak it from there to be specific to your sales. So definitely recommend grabbing that instead of trying to do it on your own. Okay, so Claude has now run through and it's going to give me a quick spit of like, what did I change and why? Here's stuff I changed. I can look through. It's possible there's going to be a mistake in here. So it's good to like double check this. And then look at it. It gave me two different things. So it updated our prospecting prep doc. For whatever reason, it's not dating this properly. I'm not going to bother with that right now. So I'm going to download that. And then our also my like specific account research play, which is slightly different. That feeds into the prospecting doc, but this is the actual process. I'm going to download that too because it made some updates there. And just as a like a side note, the way that I keep these organized is with dates. So I use year month day, because I find it to be the fastest. So I'm going to turn this into today's date. A time of recording is 2026 0602 because it's June 2nd. So for both of these documents, I'm going to start it with that. So that way, I just drag those. I just updated those things. I downloaded. I dragged them into the project. And now I know like, this is the one I did earlier in the session. Let me get rid of that. Account research. There's probably an old. In fact, I know there's an a lot of that account research. It's old. Get rid of it because I don't want to have a bunch of conflicting account. Oh, account research. March 21. Get rid of it. Oh, look, account research, plat, blue, get rid of it. All right. Cool. Now I have one definitive source of truth updated account research document. That has everything that we just did. Everything from the previous sessions and all the updates that we just made. That's it. Easy, right? Simple. The crispy, right? Well, it's not simple for me, but it hopefully will be simple for you. So go to the show notes, download those docs, load them up into your LLM of choice, ask it to explain them to you. Tell them what you want to change, have it update the documents, and then go books and meetings, my friends. Peace be with you, Adios. I got to go be an AI bro and turn my left arm into a robot arm so that I can do more prospect. Today's show is brought to you by Unify, which helps you prioritize the right accounts at the right time. Try this. Stack your tier one highest intent leads at the front of your morning dial block. So if someone signs up for a trial, visits your pricing page, dial them first before any other accounts, before you open your Slack, your email, before anything. Unify makes it stupidly simple to surface those tier one accounts automatically. We built a guide with you to find on how to get 100% of BDR's to quota in 2026. Go check it out in the show notes. Great sellers drive velocity at every stage of their sale, and one thing that's worked well for me is assigning red-mine deadlines to my process. when we kick off vendor review. Look, our prospects are not showing up to sales calls to answer our questions. They're showing up to get their questions answered and to figure out how we can solve their problem. And this is what Pythrake does. If you want to get clear complete sales visibility, you can use Pythrake's easy to use and customizable CRM that is designed to be simple from day one and stay easy to use as your team grows. We actually put together an entire resource hub with Pythrake to help you get to president's club and you can get them for free in the 30 NPC Pythrake Closers hub. The link is in the show notes.

Podcast Summary

Key Points:

  1. Alex Murphy, a senior account executive at 30 Minutes to President's Club, uses Claude to book 4-6 outbound meetings weekly alongside his normal AE work.
  2. The workflow centers on an "intelligence layer" — feeding Claude weekly data from HubSpot (accounts, territory, MQLs, engaged contacts, and deals) so it has accurate, current context.
  3. Claude also uses playbooks stored as markdown documents, including outbound processes, ICP definitions, problem-prompter (from Jen's course), and messaging matrices, which are refined over time.
  4. Key signals for targeting include high engagement with 30 MPC content (downloads, course purchases) and supplementary triggers like new hires, M&A, or fundraising.
  5. Alex emphasizes using project files for permanent memory, pruning old data to avoid confusion, and converting PDFs to markdown to save storage.
  6. The process is iterative
  7. The system is designed for efficiency, especially for a full-cycle AE and dad who can't work around the clock.

Summary:

Alex Murphy, a senior account executive at 30 Minutes to President's Club, demonstrates how he uses Claude to consistently book four to six outbound meetings each week without sacrificing his regular AE responsibilities or personal time. His approach is built on a structured "intelligence layer" that ensures Claude has all necessary context about his company, sales process, and book of business before generating recommendations. Each week, Alex exports data from HubSpot—including all company accounts, MQLs, engaged contacts, and historical deals—and uploads these as CSV files into Claude's project files.

He also maintains a set of markdown playbooks that encode his outbound process, ideal customer profile, and messaging strategies, many of which were co-created with Claude and refined over time based on results. The system identifies hot accounts by prioritizing first-party signals like high content engagement or course purchases, supplemented by external triggers such as new leadership or funding rounds. Claude then generates hypotheses about prospect problems and suggests outreach angles, which Alex validates and executes using tools like Sales Nav and Gong.

After each cycle, he feeds outcomes back into Claude, creating a flywheel that continuously improves targeting and messaging. Key practical tips include managing project storage by converting PDFs to markdown, pruning outdated files, and using dedicated projects for consistent memory. Ultimately, Alex emphasizes that this workflow is practical, not perfect, and encourages others to adapt it to their own needs.

FAQs

Alex feeds Claude weekly data from HubSpot, including accounts, MQLs, and deals, along with playbooks, so Claude can identify hot accounts and suggest outreach strategies. He then works those leads manually and feeds results back to improve future suggestions.

The intelligence layer is a set of project files in Claude containing company data, sales playbooks, and historical deal information. It ensures Claude has the context needed to make accurate recommendations instead of guessing from training data.

Alex exports his territory accounts (including unowned ones), MQLs, engaged contacts, and all deals ever worked on the product. This gives Claude a comprehensive view of his book of business.

He exports all accounts to catch unowned accounts with strong signals that might be missed, and to give Claude more information to work with, as it's cheap for Claude to sift through large CSVs.

He regularly prunes old files to avoid confusion and storage issues. He avoids storing PDFs, instead converting them to markdown documents to save space, and keeps only the most recent data exports.

Playbooks are markdown documents that provide instructions for Claude, covering outbound processes, ICPs, messaging matrices, and more. They're created by Alex describing his needs, and are refined over time based on results.

Chat with AI

Loading...

Pro features

Go deeper with this episode

Unlock creator-grade tools that turn any transcript into show notes and subtitle files.