Tom Wentworth is the CMO at incident.io, and the most AI-pilled marketing leader I’ve talked to. Before incident.io, Tom was CMO at Recorded Future, RapidMiner, Acquia and Optimizely. You can call that quite a track record.
In this episode, I have Tom show and explain how he runs marketing from inside Claude Code, his “agency agent” that builds Linkedin campaigns autonomously, why engineers are taking over GTM, and his top advice to CMOs in 2026.
Listen on: YouTube, Apple Podcast & Spotify
We discuss:
How he runs his day-to-day as CMO now vs 3 years ago
Why 30% of his time goes into bug fixing
How he builds agents & Claude skills
Claude Code vs Cowork vs Clay
incident’s “Applied AI” team and why the next GTM engineer is a real engineer sitting inside the CTO’s org
The most impactful agents/skills/workflows his team built and ACTUALLY use
The one - and only - thing Tom would still hire an agency for
Don Jeter’s monster trucks vs Tom’s agents: comparing two playbooks
His top 10 pieces of advice for CMOs in 2026
Why the CMO role might actually go away
Connect with Tom:
Tom’s LinkedIn: https://www.linkedin.com/in/twentworth/
Tom’s blog: https://tomwentworth.com/
incident.io: https://incident.io/
Connect with Finn:
LinkedIn: https://www.linkedin.com/in/finnthormeier/
Project 33 - LinkedIn Agency for CEOs: https://www.project33.io/
Mentioned in the episode:
Clay: https://www.clay.com/
Granola: https://www.granola.ai/
Linear: https://linear.app/
Fibbler: https://www.fibbler.co/
Torq: https://torq.io/
Blog post from 2016: https://tomwentworth.com/2016/02/18/some-practical-advice-for-aspiring-tech-cmos-c4f2a69a09f2/
Hope you enjoy!!
Full Transcript:
Finn Thormeier: all right my guest today is Tom Wentworth he is the cmo at incident.io incident.io is a b2b SaaS company it’s an incident management platform so icp is enterprise engineering teams specifically SRE site reliability engineers. Customers include Netflix, Etsy, Ramp, Lovable, Navan and many others. They raised the $62 million Series B last year, have about 200 employees according to LinkedIn.
Tom is probably the most AI-pilled CMO in B2B SaaS right now, lives and breathes Claude Code, builds his own marketing OS and GitHub, self-updating battle cards, a ton of different agents to handle different tasks, custom plugins, vibe coded apps and games for example for one-to-one ABM campaigns and even a personal workout coach. an OpenClaw. Before joining Incident in 2024, he already had a long-standing successful career in tech marketing specifically CMO at Recorded Future which was acquired by Mastercard, CMO it Altair, CMO the Acquia, Cmo Optimizely.
He has a background in product marketing and actually sales engineering among other things he was field cto at Interwoven when they were acquired by HP in 2000. my god what a list i’m just saying all of these things because it probably means that Tom knows a thing or two and you should probably pay attention so Tom thank you for joining and anything to correct or amend
Tom Wentworth: i will say that uh on the record if i were to be using OpenClaw which i will not either confirm or deny on this podcast it would not be at all connected to anything to do with my job at incident.io if i were to be using OpenClaw and it were to be doing things like helping me run my day-to-day workout activities it would be in its own sandbox that touches nothing other than things i’ve given explicit access to but then again i’m not confirming or denying that i use OpenClaw
Finn Thormeier: love it um all right so i know you went from cmo at Recorded Future which is a fairly large company and you had a fairly large team reporting into you back to startup life incident was a series a startup when you joined them Because you wanted to build more, be more in the weeds, manage less from what I understand. How is your day-to-day different now versus then?
Tom Wentworth: Yeah, it’s vastly different. The job of a late stage CMO has nothing to do with marketing. It has everything to do with being a middleman. You’ve got to keep the CFO happy first. We’re going to get the CEO happy which is obviously the hardest job you’ve got to keep the CFO happy because you’ve got a big budget and the CFO is watching your big budget closely you got to keep the CRO happy because we’ve got big numbers to hit you got to keep the chief product officer happy because you’re supposed to be out there representing their product in the marketplace you have all these stakeholders and teams that your job is to keep happy so the job is essentially slide creation and shaking hands and kissing babies and evangelizing what marketing is working on it has almost zero to do with actual marketing
and it’s why i think there are a class of cmos that can do things that i could never ever do that are exceptionally good at that role and then on the other side i think there are a lot of cmos now who as the world has shifted and there are a lot less of those types of jobs aren’t as well equipped to do what i’m doing today which is essentially doing marketing again which i think is much more fun but you know if you’re if you’re cmo of a huge fortune 500 company and you’re good at it like that’s a that’s a really hard job but i think in b2b SaaS i think the trend is very much moved towards cmos as individual contributor player coach doing the job not just managing the slides
Finn Thormeier: so you join incident series a you know series b you’re much more in the weeds with marketing you talked about about it before that you basically want to replace yourself with the ai cmo i know that’s a little bit clickbait but do you think as let’s say incident moves to series d series f bigger company the job of the cmo at incident will be again just meetings all day
Tom Wentworth: and at this point we’ve we’ve the advantage of doing the job of marketing at an early stage company is that you get to build the org structure of the future i think it’s really hard if you’re if you’re like 200 million dollar series e f SaaS company you’re doing well but not well enough to go public you’re not Anthropic but you’ve got this huge market this huge candidly bloated marketing team you’ve built up over the years a lot of times cmos value themselves on the size of their budget and size of their teams so like i’m not good at my job unless i’m running a team of 50 people at 200 million so i think for a lot of those companies the transition is going to be really hard
because how do you how do you build a sort of modern ai native marketing team you know you’re changing the tires while the car is moving i didn’t have to deal with that here so when we hire people here it’s for specific roles that we think that ai is not well suited towards like ai isn’t gonna pick wine for an event we’re going to run in San Francisco ai is not going to figure out how to design a great experience you know we just were in London at LeadDev conference and it was raining and we had made these umbrellas that we handed out to everybody coming to the conference ai is not going to stand out in the rain and hand out umbrellas so i think they’re
like so there are roles now in marketing that are incredibly well suited to people and of course there are roles in marketing now that are incredibly well suited to agents so i don’t have to hire those roles with people anymore but I do get to hire people where I think that people can make a difference. And there are still tons of roles where I feel like you’re not going to create this sort of mythical AI VP of marketing and they can do everything.
Finn Thormeier: So how is the marketing org structure of incidents going to look like in five years or 10 years when you’re a Series F company?
Tom Wentworth: Well, if I’m here 10 years I love this company If I’m Here 10 Years something’s gone wrong for me personally I would hopefully, hopefully I’ll be doing something in 10 years. I’ve been in this. I’ve been in tech since 1994 you can do the math on how old that makes me so i have been doing this for a long time but maybe who knows?
I look at it like the org chart is going to be comprised of people managing agents Like I look At agents at this point almost like I look At interns and so agents require a tremendous amount Of daily coaching like you can’t just deploy something so i have an agent that is a what i call digital agency agent and it’s not as good as the best digital agencies it’s better than an entry-level marketer it’s been coached on it right now manages primarily LinkedIn i’m expanding it to work on Google AdWords as well but every single day it looks it looks at budgets it looks at what’s working what’s not working it looks at intent signals on our website and it can literally create a campaign from scratch
so if it sees Goldman Sachs coming to our site like hitting our pricing page and we don’t have a LinkedIn campaign set up for Goldman Sachs it will literally create a LinkedIn campaign create creative for Goldman Sachs create a landing page publish all that stuff autonomously like that is super cool so that’s a role that i don’t have to hire anymore and i think there are a lot of roles on the org chart that are going to look like that but it’s a ton of maintenance so somebody has to manage that agent if you deploy an agent and just let it go it will fail miserably so i look at the job in the future of you’re going to have people doing people things and i don’t think that’s going to change and then i think we’re gonna have a lot of people who are marketing generalists who are good at understanding the business problem and we’re going to sit there and train agents to do marketing for them
Finn Thormeier: what’s the biggest difference in skill sets between a people manager and an agent manager
Tom Wentworth: yeah i mean and i don’t know it’s a good question and i don’t know that the pure people manager role has a lot of legs especially in b2b SaaS like i manage a pretty i have three direct reports probably all i’ll ever have at this like it’s a pretty tiny number um i i think that like there are people managers who are exceptional they have empathy they are very good at building you know coaching their employees and building career paths and all those things are still important and i still do those things people on my team but i don’t think the pure people manager is going to make it very far these days i think the person who understands the problem very deeply
and the person whose passion is to build software that to fix that problem is what’s going to win it’s sort of the same in engineering like the future of engineers we used to have front end engineers and back end engineers and product managers and we had all these roles involved in building software and now you look at someone in Anthropic and it’s like now we have one role member of technical staff they build they do everything from build to test to ship to do feature requirements they just do everything end to end i think that’s kind of where marketing is going to end up being the generalist who really understands the problem and who can get passionate about solving the problem with tech is the one that’s going to win pure people manager they’re going to stick with fortune 500 and i think they’re not going to have it’s going to be hard for them to succeed in SaaS
Finn Thormeier: so the generalist skill set back in vogue
Tom Wentworth: yeah yeah for sure
Finn Thormeier: so where does how where does your work happen these days like are you inside of Claude Code or how can i imagine the i don’t know eight hours that Tom Wentworth sits in front of his computer
Tom Wentworth: a lot of it is i can share my screen if you want and give you uh give you a look at like we can look at my maybe we can well i’ll show you some of the stuff all right so let’s go ahead and see if i can share that
Finn Thormeier: where do you start your day what’s the first thing do you do you have a dashboard like
Tom Wentworth: and so Slack is still the sort of our our product is about empowering engineers working inside of Slack to resolve incidents so our company culture is very much Slack so i start in Slack what Slack does is gives me clues that i need to go into specific processes to check things is sort of so my day is uh what what sales reps yelled at me because something got defined incorrectly it’s essentially triaging who’s yelled at me when i wake up and it’s
Finn Thormeier: you get bug reports you wake up to bug reports
Tom Wentworth: bug reports is a great way to put it um all right so here’s a good example of a few of the agents that i check every single morning so i use a few different a few different uh ways to run agents these are called Claude Code routines which i like because they run on Anthropic infrastructure and they just sort of run automatically so like one of them that i look at every single morning is called demo request sla this just goes and tells me of all the of all the qualified trials and demo requests that we’ve created in the past 24 hours have we seen at least one human touch on all of those mqls if not we should post a summary of that into Slack so every single morning Claude Code runs goes
and does all of this analysis um it use it goes into Salesforce using my favorite uh thing in the world all of a sudden Salesforce command line tool so if you’re paying attention to Salesforce You hear Mark Benioff talking about like, we are going to have complete coverage. Anything you can do in the UI, you can use the command line tool and he’s right.
is entirely right the Salesforce cli command line interface if you plug that into Claude Code you never have to really log into Salesforce anymore it’s honestly incredible so my skill logs into Salesforce using the command line tools goes and does a bunch of lookup stuff finds a bunch of relevant information and goes and posts a message out to Slack which is this message here so i check that out
Finn Thormeier: was this thing something that you previously had someone do manually or is this just a net new thing
Tom Wentworth: we didn’t do it all so i was asked by our cro like hey Tom you know you tell me that that we need to get better at following up with with demo requests but like do we have any visibility to it and the answer was no so now we do another problem we have which is a problem that every single company on the planet has is duplicate accounts and Salesforce it is incredibly difficult even in 2026 to prevent duplicate accounts from getting created in Salesforce so Claude Code routine that runs every time a new account is created in Salesforce and this simply just goes and does a bunch of research to make sure that the account we created isn’t a subsidiary or a misspelling or uh or a different url
like go and exhaustively match this account to our entire account list in Salesforce couple hundred thousand accounts to make sure it’s not a duplicate and if it is a duplicate um automatically merge it so this automatically reduces the amount of duplicates that we get created in Salesforce
Finn Thormeier: how much time would you say per day do you spend building new routines and skills
Tom Wentworth: um it’s it’s probably a fourth of my day that could be a little bit it could be a little bit it’s it’s somewhere like 20 ish percent is going through all the routines that automate things and trying to make sure they’re either working as expected or they’re um or figure out how i can improve them
Finn Thormeier: okay so so the the 25 also includes kind of troubleshooting and and and and and improving existing ones
Tom Wentworth: yes mm-hmm so for example like we had uh we use Clay to do a lot of our automations as well Clay is fantastic we found this morning a Clay process was broken because uh the process used uh Anthropic Sonnet 4.0 Clay stopped supporting 4.0 didn’t tell us about it so all the processes stopped working and it happened to be a process that set some important data on the account and Salesforce that we needed like that was just like really important data So like it’s diagnosing, that’s back to the point of you’ve got to manage these agents. Like every day I’m looking at all these things to make sure they’re behaving the way I want them to behave.
Just like if you had an intern on your team you would audit the quality of the interns work every day or every week. It’s a similar thing to me. They’re providing tremendous value to me but I still have to look at them.
Finn Thormeier: i mean it almost sounds to me that your next highest leverage hire is some sort of marketing engineer who can help work through the bug report
Tom Wentworth: i gotta i got a strong point of view on that one so um so you had marketing ops forever right marketing ops and they were tool-centric people they grew up learning HubSpot and Marketo and Eloqua and like all the market automation tools so because they were reasonably technical they became the people that owned the deployment of marketing technology right then Clay again who i love popularized the kind of gtm engineer which is like these are the most technical of the marketing ops people they know a little bit more they got they can think boolean logic they have like a little bit more technical brain but their background is still more in business than they’re more on the gtm side
and less on the engineer side what we’ve seen at my company is the other side happening we have what we call an applied ai team and these are real engineers who don’t know anything about gtm like they don’t come from gtm but a really good engineer can solve any problem you give them a problem they know how to solve problems so i actually think the future of of the whatever we call it marketing engineer gtm engineer whatever are real engineers coming into marketing as an example what we’ve set up at incident.io that is honestly incredible is a series of tools that we’ve built internally through our gtm engineering team and these are all the tools i’ve listed here
and these tools let us do everything from gathering data from our data warehouse gathering transcripts like anything that has context that a gtm person needs everything you get from Salesforce right all these tools our gtm engineering team has built a very sophisticated and a very robust set of tooling that’s rolled out across the entire go-to-market organization at incident.io and with this tooling you can do all sorts of really cool stuff like let me go show you all right so here are all the skills so skill is a package set of instructions and connectors to solve a problem so if you’re a one of our aes of this company
and you want to know what customer references that we can name drop great you wanna be able to get a summary of a deal you’re working on super simple you wanna get prepped on a demo you’re about to deliver there it is built in for you you wanna get let’s see you want some ideas on what to post on LinkedIn we got that for you you want to like you want to put together some materials on how to migrate from our competitor PagerDuty we’ve automated that for you so the set of of capabilities our applied engineering team has built have been packaged up into all these skills that are that are made available across our entire gtm team
so now they can do all sorts of really amazing stuff this is how has been where so you asked me what my favorite tooling is yeah what what we do the pattern i think that’s my favorite i think is the best is to build these skills inside of Claude Code because Claude Code is designed for writing code so Claude Code is predominantly building these skills where it needs something that can’t be done natively it builds things in python to do stuff so Claude Code to build the thing and then Claude Cowork to hard to be the harness for the thing to distribute it out to the mere mortal users So, and it’s very easy to do this.
So you have this idea of organizational plugins in Claude Cowork which is what these are And as a person who has the right level of access You can deploy the thing that you build As an organizational plugin For example Here’s my marketing ad agency skill I talked about And here it says coming next Google ads That’s cool This is a skill that now is deployed across anyone in marketing at incident.io and they can, with one click, they can run this agency. Super cool.
Finn Thormeier: How has adoption been for all of these things across the team?
Tom Wentworth: Near 100% adoption daily.
Finn Thormeier: And is there any driver behind it or you just hired the right people who are interested in this anyway
Tom Wentworth: answer? The drivers stuff is valuable. Like this is the thing about sellers, Is like you can send them unlimited emails Slack messages hey use the thing if the thing is valuable and helps them hit their number they’re going to use it if the thing is not valuable it doesn’t help them hit their number they’re not going to use it so if you deliver something valuable you will get adoption and and the team that has put all this together is delivering things that are very valuable
Finn Thormeier: what have you learned there about documentation because one thing it’s useful it’s another thing i mean this is a lot of skills do they all know every skill and how how are you communicating
Tom Wentworth: again this is where having engineers if you have me build something like i’m i’m i cosplay as an engineer like i know a lot of the basics but real engineers they think about things like observability and error handling and documentation and all the stuff that to me is like i don’t care about that who needs documentation um so they do and every single one of these skills when you look at how it’s documented and put together it was incredibly well thought out even like the like just the description of things is much better than i would have done so here’s actually one of my favorite ones this one is an incredibly badass skill this one will turn anything that you create
and co-work into a branded asset at incident.io so i go into i go into here and say use brand design to create a hello world uh document for my customer whatever tomwentworth.com and the thing will that gets produced will be applied to our design system and will look like it was built bespoke by one of our designers our designers this is a good question i get a lot is like i think design has become one of the jobs that’s most defensible against ai if you think about AI in general it produces sort of a regression to the mean like it’s going to produce the average of stuff a really great designer is going to sort of break new ground
and I’m a huge fan of the role of design now in this new world and we’ve got great designers at incident.io it’s one of our superpowers so this is not going to be as great as our designer building it bespoke from scratch but making sure that anything we send to a customer looks good and is on brand is now just automated
Finn Thormeier: yeah but i mean that’s only true for the what top top one percent of really great designers most designers was just kind of average by definition and so
Tom Wentworth: yeah well don’t hire don’t hire them um don’t hire any of my designers either but don’t hire average designers um but the great designers are honestly one of the most it’s a defensible place to be in marketing and it is a weapon and our brand as a result of our visual identity our brand is spectacular in my opinion
Finn Thormeier: how do you decide who to make the applied ai engineer
Tom Wentworth: uh so in our case it was people on the engineering team who had a passion for building and solving new sets of problems and who sort of saw the writing on the wall there are a bunch of people on our team i don’t want to single out any individual one because they’re all awesome um but they all sort of raised their hand when we introduced the concept and this team by the way lives under our CTO and co-founder organization so it’s not a marketing thing it is a company thing they happen to spend a lot of their time in GTM because that’s where the most leverage is but they raised their hand signed up for this charter because this charter was so important to the organization and because they wanted to be a part of it
Finn Thormeier: yeah and then
Tom Wentworth: all right so this by the way this did work right so this is obviously not the world’s most uh complicated sophisticated but uh but the point is uses our font uses our icon uses our are all the attributes of our style guide were correctly enforced and you know this is a so imagine now if you’re a rep and you’ve got to send a one pager over to a prospect about a specific feature or competitor it is trivial for you to do that now using this tooling
Finn Thormeier: yeah and then how do you guys how do you personally and and how does the applied ai team decide what to automate what to turn into a skill what to prioritize and in which order
Tom Wentworth: yeah they talk to people this is why also i’m not a great applied ai engineer i don’t like talking to people as much as they do but they sit down with the stakeholders. They watch them work. They sit down on the BDR team, they sit down with the AEs and then they document it. They use Granola to take notes and then turn that into an incredibly detailed set of specs that turns into stuff they implement and build.
Finn Thormeier: What’s your take on what I read Zuckerberg did, which is basically to install like a screen recorder on every employee so you can record how they do things so that you can then turn those things into skills and agents?
Tom Wentworth: yeah i think that i think our team that’s it there’s a carrot and stick side to it i think our team our team does that by choice i don’t know that i’d feel super good about Zuckerberg putting spyware on my computer and like you know um but almost everybody so we use Granola almost everybody we have a corporate Granola license almost everyone on the team has connected Granola Granola is one of the connectors that are all of our skills know how to access so if you’ve got Granola connected it will connect to your Granola your transcripts so it has the context from you in all the stuff so the context when we do any sort of analysis it’s looking at the sort of corporate sources, Salesforce Gong Slack channels.
Like for every customer at incident.io we have a Slack channel. So it’s going to look at the customer Slack channel, it’s gonna like Gong call transcripts. It’s going To look at Salesforce and it’s going to Look at your Granola and try to pull all those things together when it’s doing things like giving you a summary
Finn Thormeier: You said before that about 25% Of your day is building these skills And routines and fixing them and optimizing them slash managing them. managing what’s the remaining 75
Tom Wentworth: uh helping other people on the team do their job so part of part of my you know part of my role as cmo is i’m an ic and like my ic charter on this team i’m not super creative so my background was product marketing but i was i was more tech i was good at product marketing because having spent a bunch of time as a sales engineer i was used to talking to customers i understood the markets like i was i liked that side of it but i also was pretty technical and and my role here is more to be the i guess the marketing engineer we talked about earlier for my team so helping them implement the things that they need to do their job more effectively to provide coaching and guidance
so that’s that’s another big chunk of the day um i spent a lot of my time in demand gen like the number one thing that that we’re trying to do here and said i always grow faster so i spent a lot of my time thinking about the things that we can do to to grow faster as an ic so that’s a big chunk of my time those are the big three it’s how do we grow this company faster how do we how do we leverage ai better and then how do i just sort of be there for the team to help them when they need me
Finn Thormeier: how many meetings do you have at this point
Tom Wentworth: not a ton we do a daily standup that you know we use Linear to which is generally a bug tracking system slash issue management system but it’s really great honestly for uh for teams so we use Linear we have a daily stand up we go through what we’re working on Linear we try to ship things weekly that’s kind of the cadence we operate on i don’t believe and that i don’t know that i should say this out loud but i will i don’t believe in quarterly plans i certainly don’t believe in annual plans i believe at this point honestly in weekly plans like no like i can i can tell you what we’re gonna do this month and probably next month but i can’t tell you what we’re gonna do in q4
like i know our event is in q4 but i don’t know i don’t know what’s gonna bubble the top in q4 i hardly know what i’m gonna like what’s gonna happen next week honestly so i try to operate the team on a quarterly cadence where we can ship things weekly sorry operate the team in a weekly cadence so we can ship things weekly so that’s sort of what we try to optimize for is what value can we deliver to the company today tomorrow this week and we’re pretty good
Finn Thormeier: so what’s so far the most useful skill or routine you you build you think
Tom Wentworth: um For go-to-market as a whole, it’s been all the skills that just simplify all the research that they need to do. I think for marketing, I think the thing that we’re most excited about is the agency skill because agencies are expensive. Agencies can be slow and this is just reality. The agency doesn’t work at your company so they don’t have the same motivation, but like my Claude, a co-work agency. can work 24 seven. And, you know, how being able to wake up and say, Hey, I stopped some budget here. I increased some budget here. I spun up these three campaigns. It’s pretty incredible.
Finn Thormeier: What thing do you think you’re most likely to still hire an agency for?
Tom Wentworth: Um, if we get big enough, PR is one where you need, it’s, it’s relationships. You need relationships with journalists. And I think PR is one that, that I we had an incredible PR program at Recorded Future. I think the problem with PR for most companies is they have nothing interesting to say. It makes no sense to do PR if like all you’re going to talk about are product releases, but no one cares.
At Recorded Future, we were sort of central to the nation state cyber story so what Iraq North Korea Iran etc China all the countries were up to and we had visibility into that so we could go to New York Times Wall Street Journal and get stories placed very regularly most companies just don’t have anything to say pr is just an absolute waste um but but pr is one like you need external help to make pr work
Finn Thormeier: You went quite all in on the Anthropic slash Claude ecosystem. Are you worried at all? I’ve seen some engineers, I think talk about that Codex is now better for certain skills or for certain things. Are you worry that that’s going to happen and are we stuck with it?
Tom Wentworth: No, I think a more interesting conversation is back to OpenClaw. If I were to be using OpenClaw, I would be using it with one of the open source Chinese models purely because of cost. So, the cost delta between some of the frontier models and the open source models is so dramatic. And I think... So, I wonder what’s going to happen there over time? At what point does even do the open-source models get to be good enough to do most of the stuff?
and and i i mean i don’t think we’re there yet for but so i think what Anthropic’s done that’s most interesting is all the tooling like code and co-work is so good and they’re just so smart about building for their stakeholder it’s such a differentiator for them i mean the models are great but i think i think the differentiation for them is more on the tooling and and i think they’re way ahead of OpenAI there um i think open is catching up but i think that’s where their differentiation is The cool thing is, it doesn’t matter. We’ll pick whatever the best tooling is. We don’t have skin in the game necessarily.
We get to benefit from the best tool just like our customers get to choose the best incident management software which makes us better. You know, the fact that it’s so competitive right now just means that these tools are going to get better so quickly and it’s a win for everybody.
Finn Thormeier: how do you decide whether you map a workflow inside of Clay or with Claude Code i’ve seen some people who said they moved off of Clay completely
Tom Wentworth: Clay is good for more deterministic things like Clay Clay as a as a as a sword it’s a bridge it’s an interesting bridge between the old world and new world like in the new like in the new world of i just have a Claude Code routine that in a non-deterministic way just says hey go figure out if it’s a duplicate account or not or merge it if it is like there are very little rules there with Clay i could do the same thing but it would be much more deterministic if this then this then this look here look here so there are certain deterministic workflows where Clay is great like enriching accounts and contacts Clay is great we use Clay to score all of our trials and demo requests which is a pretty deterministic process um so i think where it’s deterministic more deterministic Clay is great where it’s like i don’t care how you get to the answer i think Claude Code and co-worker great
Finn Thormeier: yeah what tools used to be in your toolbox six months ago that have completely disappeared since then
Tom Wentworth: Um, I think it’s, I don’t know if there’s any that’s complete. I think what’s going to disappear. I think the BI sector is really hard. We’re playing around a lot right now with doing data in Claude Cowork which has some new data tools. So we’ve implemented something we call the Data Brain which is how do we connect all the data sources at incident.io? And it’s not just connecting to the data sources.
it’s giving the data the brain intelligence to know what the data sources mean what the columns describe and all of that so when you run a query it can actually do something meaningful for you so in the past you used to have to go to a business analyst to write a bunch of SQL to come up with a report and now almost anyone on my team can go into Claude Code work run a run a query know that they’re getting approved data because it’s coming through data brain and then Claude has really good ability to visualize it so like i think that space is is interesting um i think there are a lot of point solution gtm tools that aren’t going to have a long shelf life because i can rebuild them from scratch so i don’t know that i’ve killed anything yet yeah i have my eyes on a lot of things that i think if i if i if i had the time i could recreate it just hasn’t been a big enough priority for me yet
Finn Thormeier: sure any new favorite new tools gadgets i mean obviously everything happens in Claude but connectors you know anything that you’re currently excited about
Tom Wentworth: we do a lot of there’s a little little company i’ll give a shout out to them called Fibbler that we like a lot
Finn Thormeier: oh yeah million millionish i think i saw them say they had a million in ARR yeah i’ll have to add them on the show
Tom Wentworth: Fibbler the simple problem they solve is you spend a lot of money on LinkedIn you don’t get a lot of direct conversions on LinkedIn because no one like there’s not it’s not really a direct response medium you want to know how LinkedIn is influencing your go-to-market Fibbler is a cool way to bridge that gap in our digital agency the thing i told you about it uses Fibbler to analyze if a campaign is working or not so when when it know when it goes and looks to see if a campaign is performing it’s querying Fibbler’s mcp to get the performance data. And then it’s like, oh, this is doing terrible. We should cut the budget or pause it. But Fibbler is a cool little thing.
I like, you know, it sounds like table stakes, but like if it’s got an MCP or an API and it has proprietary data that’s meaningful to me and I can plug it into processes, I like tooling like that. I don’t like tooling that drives me into your UI and that doesn’t integrate with, you know, Claude is my center of gravity.
Finn Thormeier: i had uh Don Jeter on the show the cmo at Torq i feel like he’s almost the exact opposite to you
Tom Wentworth: totally
Finn Thormeier: i asked him i could not do don’t you know Don goes into this RSA Conference and brings in a monster truck like his his whole his whole like the brand he’s built around Torq which Torq isn’t a SOC company security operation center like it’s pretty like pretty pretty like classic cyber security he’s turned that thing he has this like fake reporter, Trevor from Torq that goes like has like mullet and like goes around in the conferences with a mic and interviews people awkwardly. Like how does he think about stuff like that?
Tom Wentworth: So that’s it, That’s a good example of like there’s not a one size fits all you can build a great company by being Don I think you can build a great Company by being me too. but i think you need both i think like Don probably has people at Torq that do my job and i certainly have people at my company who think the way that Don thinks
Finn Thormeier: well interestingly asked me if he has any kind of hard rules with his team on how they’re allowed to use ai or not and you showed me just your design skill he was like not a single piece of design that goes us from our brand is allowed to be created by ai everything needs to go through our design team like the people
Tom Wentworth: I agree. So, I actually agree with that and With one exception I think there are things that can be Templated like in my agency We have a template For LinkedIn ads that when we put out there But within that template I can change the headline the sub Headline like there are certain things I can Change so everything gets produced by AI Looks the same But the thing it looks like was designed by a person Um, I think Like would I let AI design a trade show booth absolutely not he said LinkedIn post not a single person is allowed to put a LinkedIn post with a graphic that was not created by our design team and i think honestly we’re not too far from that either with some exceptions i think but i think
like when it gets down to the last mile of a seller wants to put together something for a customer like wants to send a summary of the next steps based on a demo like if i could get that to be 95 as good as a designer we’re probably good with that um i think when you get to one-on-one communication i think ai design is okay i think when you get to what we put out into the world as our brand i’m with Don
Finn Thormeier: i feel like it also has to do with knowing where your moat lies and i mean you guys have a strong brand i don’t know if people recognize you for your brand identity and the design but for Torq Torq is absolutely one of the core things that makes them them
Tom Wentworth: what they recognize us for is they our design is is super strong and i think super differentiated we also care a lot about copywriting and i think that’s where we get a lot of recognition for copywriting is hard and what’s hardest about it is applying it consistently everywhere and i think that’s something we’re very good at if you go to our website like you can tell websites you can go there
Finn Thormeier: I saw it on your rain ponchos. What was it saying?
Tom Wentworth: This product market wrote this page, the language is different, the tone is different like the style’s different where you go to Apple like you know everything there was produced by Apple I think we try to stick closer to the Apple model like there’s literally one person in incident.io who is responsible for sort of making sure that everything we put out meets our copywriting standard so I think what Don has done with visual identity we’ve done with copywriting
Finn Thormeier: yeah i saw it you you posted on LinkedIn the way you handed out i think umbrellas and the rain ponchos on the back said we’ve got you covered and i was just like that’s such a small thing but that’s great copywriting
Tom Wentworth: ridiculously good that was the brainchild of our brand designer Chloe who’s fantastic don’t steal her um i can’t afford her stuff that’s her but like but she knows she knows the tone and language she knows how the incident.io way of how we talk about things and i think that’s what’s cool and that we are very serious and very consistent about
Finn Thormeier: your agents will go after people is it kind of like you have to be fully in or fully out with ai i almost see that dichotomy there’s some people who are just fully ai built like you and then there’s people who are fully just like Don who are almost like i don’t i don’t i don’t even use ai at all i think
Tom Wentworth: yeah well so i’m sure Don does and i’m sure Torq as a company does so i look at it like my the formula in my head they put this like in a slide yeah a year ago it’s like everybody’s sort of converging to the same place with ai like we’re all gonna everyone’s all the set the skills i showed you everybody’s building the similar types of skills right we’re All Building The Same Stuff so at some point ai will converge and we’re not there yet you can differentiate on ai right now but at some point we’re going have the same shit just It’s like... 10 years ago, 15 years ago, before anyone had marketing automation, it was an advantage to have Marketo Eloqua because you can email people better.
And then everyone got the tools and it was like, oh, we all have the same shit emails again. So I think AI is in that stage now. And we’ll, we’ll at some point get to a point where, all right, we all have the same set of skills. We’re back to how do you really stand out? And that’s with brands.
so i think you it’s not either i think it’s not either or i think you have to be best at both i just think what’s going to scale better long term isn’t going to be my side of the world it’s going to be Don’s side of the world like the ai like like i was at another i was on a panel last week with Nate Burke who’s the cmo of seven x seven ai if i imagine it start other cyber security startup and the comp he made was my job as CMO is to make everything we do a three Michelin star experience which I thought was a really cool way to say it like everything that we do has to feel like and I’ve never been to a three Michelin star restaurant so I wouldn’t know what that’s like but that sounded cool to me and I think it’s that mentality like you need to do what I’m doing but if you do it at the expense of what people like Don and Nate are doing I think you’re going to fail as well you need both
Finn Thormeier: yeah you said uh i think marketing output is uh ai proficiency times taste squared
Tom Wentworth: yeah
Finn Thormeier: given that taste in that formula obviously matters way more than ai proficiency you build ai proficiency through building things being inside of Claude Code building a skill getting these tools live how do you build taste
Tom Wentworth: uh taste is one of those where you don’t you’re you’re you’re you’re born with it you develop it like i just think to think to Steve Jobs who’s my idol and the person i’ve always admired the most in tech and Steve Jobs you know the the famous example was when he was growing up his dad taught him he had to paint both sides of the fence even the side that no one would ever look at um i think you i think you i think you’re born with taste i don’t think you can develop it so i think you have to find people who have taste and and do your best to hire them recruit them and keep them um i don’t have taste so if you relied on me to be that like i’m like dude that looks good to me bro ship it looks awesome and then the people that have taste on my team will say yeah but it could do this and this and this and this and could say this and this and this um so
Finn Thormeier: but then how do you hire the people with taste because you got to hire them
Tom Wentworth: no only i do Don doesn’t have
Finn Thormeier: but how do you how do you identify them how do you identify them if you don’t have it obviously you have it you’re a little humble here but uh you
Tom Wentworth: well one of the ways is you know we have our our founders one of the things that our founders brought to the table at this company um was was they had always believed there are three engineers who had been working on building incident management tooling but they’d always been designed they’d always had a high regard for design and always had good instincts for design so honestly for me there’s one of our founders named Chris who’s sort of the the brand guy among our founders so Chris has a good eye so I would just leverage Chris and he’d say yeah that that person’s taste is spectacular we should hire her like literally how we hired Chloe we interviewed her
and she put together some stuff in the interview process and Chris and Jessica who runs product marketing and brand our team like she’s awesome Tom we need to hire her and of course we went and tried to hire her um so i think if you if you have someone in an organization who has that taste they become the benchmark for it and in our case it was Chris our co-founder
Finn Thormeier: that’s Chris Evans i had him on the show i think a year and a half
Tom Wentworth: yeah Chris Evans let’s not give him too much credit like i’m already but now field cto but yeah he’s good at many things but he is good at taste
Finn Thormeier: okay uh last thing i wanted to do a rapid fire with you you wrote a blog post in 2016 so basically exactly 10 years ago at the time you were cmo at Altair RapidMiner
Tom Wentworth: yeah
Finn Thormeier: with your biggest advice for tech cmos at the time I’m gonna read each advice to you that you wrote in that blog post from 2016 and you’re just gonna say is it still true and if it’s not true how you would amend it or fix it or change it for today’s world
Tom Wentworth: okay
Finn Thormeier: first one was your most important job is to recruit and retain exceptional people
Tom Wentworth: uh still true still true there’s less of them so it’s even more important today than it was 10 years ago because if you only got a fraction of the hires every single one matters in a way maybe it didn’t 10 years ago
Finn Thormeier: find the right balance between potential passion and experience in hiring
Tom Wentworth: uh still true although i weigh potential more now than i weigh experience just because things change so fast these days that like the playbook from three years ago is irrelevant
Finn Thormeier: don’t fall into the trap of focusing too much on the top of the funnel just because it’s easier to measure often the most difficult challenges happen well after a lead has become a qualified opportunity
Tom Wentworth: yes i think actually this has been my biggest failure and maybe realization of a current company like 10 years ago there’s a lot of pressure on because of HubSpot like we can blame them there’s a lot of pressure on marketing jobs to generate leads and leads are the engine of go to market and more leads is always better the reality is more leads isn’t always better what’s better is the people that you care about caring about you and sometimes that means handing out umbrellas at an event when it’s rainy sometimes it means showing up at your competitors events and handing out plushies like things that don’t result in leads
but i i have great confidence ultimately will result in the company growing faster so i i’ve the number one thing i’ve learned as a cmo it goes back to the Don Jeter example like if you just focus your world on how many leads you pass the sales this week you will never succeed long term as a cmo
Finn Thormeier: messaging is much more important than you think
Tom Wentworth: a hundred percent true we already talked about that like and put copywriting in there or two a hundred percent still true man i’m pretty so far i’m pretty good
Finn Thormeier: write more read more you were basically saying build a personal brand
Tom Wentworth: yeah still still i mean still obviously important
Finn Thormeier: don’t propose a rebrand when you started a new company
Tom Wentworth: a hundred percent the dumbest thing i hate i hate cmos when they like they just go and do the same shit in every company like oh i’ve got to do the branding i’ve got to read you the logo like don’t do that Don Don first big thing that Don did was rebrand the whole company Don is allowed Don is the one person get away with that because like Don took it from a boring thing to what he built today so he gets to do that
Finn Thormeier: don’t chase shiny new martech toys
Tom Wentworth: yes 100 true it’s the it’s the that’s another dumb thing to do the software the technology is almost never the solution to the problem
Finn Thormeier: operate on a monthly marketing plan
Tom Wentworth: uh no now it’s weekly i already i already changed that one but it’s certainly not annual that’s number one true but now i’d say it’s weekly
Finn Thormeier: say no more often
Tom Wentworth: yes that’s still very true hardest thing about being a cmo is getting all these requests from everybody who thinks they’re cosplaying as a marketer and like when it’s from whether it’s from a founder or from a brand new bdr sorry like that’s just not going to be a priority for me
Finn Thormeier: don’t trust your budgeting process it’s broken
Tom Wentworth: yeah budget budgeting is budgeting is broken everywhere i’ve never been in a company where like the marketing budget process was some formulaic scientific thing
Finn Thormeier: listen to your board but trust your own experience
Tom Wentworth: yeah another one is good to board board i used to be scared of board meetings and you all you learn pretty quickly that like these board guys are super impressive and super smart but they’re they’ve generally speaking they don’t not in your business they don’t operate your business and they’re not accountable for your numbers so like if you just listen to your board you might keep them happy but you’ll probably fail as a cmo
Finn Thormeier: it’s okay to feel like you’re in over your head
Tom Wentworth: yeah we’re all i don’t mean to do that we’re all in over i’m in this podcast interview i’m in over my head obviously yeah every day that’s uh i’m i’m yeah every day i feel like a failure
Finn Thormeier: and then obviously the last one you disagree with it’s seek advice from others
Tom Wentworth: well i don’t talk to i don’t have any i don’t talk to him anymore i mean i there are people on LinkedIn i really trust like Don is a good example when they say something i read it i take it seriously there are people in the ai sphere that i’m starting to build that level of trust with that like when they’re doing something uh Dan Slagen at Zapier or somebody that dude says he’s doing something i’m 100 gonna like study it learn from it and try to do it myself so like
Finn Thormeier: any other names come to mind
Tom Wentworth: um i’ll have to think about it Dan was up the top of my list i just was with him last week um let me get let me think about it
Finn Thormeier: i actually interestingly i saw in that blog post or in a different blog post from also 2017 you were referencing Jason Lemkin and i saw you firing some shots at him recently
Tom Wentworth: i’m really shocked i actually agree with the thing about Jason so he wrote this post about the ai vp of marketing and like first of all his job is to sell tickets to his conference so he’s it’s an incredible headline to say what the way that he like it’s he wants to drive sponsor like it’s it’s what he’s doing what he’s supposed to do but the thing that’s made his business successful isn’t that he built an AI VP of marketing it’s that he painstakingly over 10 years wrote five pieces of content every single week that are great like that he’s built a brand what what his marketing is about brand building not about an AI an AI VP of marketing so my only point about him is like, yes, you should automate.
It’s the taste squared conversation. Yes, a hundred percent automate marketing that can be automated. But if you do that without having taste and having built a brand, the AI doesn’t matter. The AI is irrelevant. If you just are great at AI, you’re not going to be successful. And SaaStr wouldn’t be successful without the brand that Jason built.
Finn Thormeier: I just read all of the things from 2016. Anything you would add to that list today?
Tom Wentworth: I mean we talked about a whole bunch of it now like just the how the org structure is going to change. I think it really comes back to people now have to want to be an individual contributor. I think the CMO role might go away at some point, honestly. I think you’re either going to change or go away like it doesn’t. The chief marketing officer was honestly historically the figurehead for marketing at an organization that would travel around, go to conferences, represent the brand do all of the sort of shaking hands type activities I think in SaaS, especially early stage founders don’t care about any of that. They’re like, yeah, it’s pretty cool. What’d you ship today?
Like what are you going to ship me at the end of the week? Like your 30 60 90 day plan is irrelevant. Like, what do you do? How about a 30 60, 90 minute plan? Like, what do You do one? So I think like the drive to, I think there’s a bunch of CMOs that won’t make it because they don’t want, they don’T want to go back to having to do the work again and they’ll, they’ll go fractional. They’ll do, they’ll find another way to make money. but you’re going to definitely see a shake up there and i could see the role going away in favor of either something brand new or of just like a bunch of more senior people reporting to different parts of the business
Finn Thormeier: and that’s a wrap thank you Tom











