GTM 174: The 7% Rule: How AI Is Rewriting Customer Success Budgets (From 10% to 7%) with Abbas Haider Ali

Podcasts

The GTM Podcast is available on any major directory, including:

Abbas Haider Ali is Vice President of Customer Success at GitHub, where he leads a 550+ person post-sales organization supporting a $2B+ ARR business serving over 150 million developers, including more than 90% of the Fortune 100.

Previously, Abbas was VP of Customer & Partner Success at Twilio following its acquisition of Segment, and has held executive roles at xMatters, Opnet Technologies, Managed Objects, and IBM. He is also a General Partner at GTM Operators Network, investing from seed through growth, and is deeply committed to mentorship and advancing underrepresented leaders in tech.

Discussed in this episode

Why “AI-led growth” can hide churn and value erosion
Lagging vs leading indicators for retention endurance
Why the CS investment benchmark is shifting from 10% → ~7%
A simple “waterfall” for allocating post-sales budget: support → onboarding → outcomes
How to “lever up” the envelope with premium support + professional services
Why expansion (not renewals) is the early signal of product-market fit
The rise of AI-powered specialized generalists in post-sales
When forward deployed engineers make sense (and when they’re just a fad)

Episode highlights

00:00 — Why customer expansion is the real signal of product-market fit
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=0

00:50 — Lagging revenue vs. leading indicators of endurance
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=50

04:11 — Why the CS benchmark dropped from 10% to 7%
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=251

08:41 — How AI moved from internal efficiency to customer-facing leverage
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=521

11:41 — Why retention cost matters more than CAC in the AI era
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=701

14:17 — The simplest framework for allocating the 7% CS budget
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=857

18:53 — Founder-led CS, design partners, and early-stage PMF myths
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=1133

23:16 — The rise of AI-powered specialized generalists
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=1396

30:00 — When forward-deployed engineers actually make sense
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=1800

49:31 — The one rule for building a durable SaaS company
Watch: https://www.youtube.com/watch?v=SgQdKjWT8kQ&t=2971

Key Takeaways

1. Expansion is the real PMF proof.
Renewals tell you customers tolerated you; expansion tells you customers depend on you. If your earliest customers aren’t growing usage with you, you’ve sold promise, not value.

2. AI growth can be a mirage.
Record-breaking acquisition can mask that the core product isn’t sticky. When the market shifts fast, retention becomes the governing constraint—not pipeline.

3. Track endurance like a leading indicator, not a postmortem.
NRR and GRR are lagging truth. The winners build instrumentation that detects value drift early, before retention metrics roll over.

4. The CS budget “envelope” is tightening for a reason.
If AI meaningfully automates post-sales work, 10% of revenue becomes hard to defend. The new bar is closer to ~7% unless you’re sustaining exceptional expansion.

5. Allocate spend like a waterfall.
Fund reactive support at the minimum viable level, then prioritize onboarding, then outcomes. If you reverse that order, you create an expensive org that still leaks customers.

6. Monetization is how you stretch the envelope.
Premium support tiers and professional services let customers subsidize cost-to-serve. Done well, it turns post-sales from overhead into leverage.

7. Early-stage CS is “founder-absent readiness.”
Seed-stage isn’t about building a perfect CS function, it’s proving customers can succeed without the founder in every meeting. That’s the real Series A readiness signal.

8. Specialized generalists are the new unicorn.
AI doesn’t remove specialization, it changes where it lives. The specialization becomes “using AI to deliver expert-level help” across support, onboarding, and adoption.

9. Forward deployed engineers are for discovery, not scale.
They’re best when product evolution is rapid or use cases are unclear. Once you’ve mapped the value, the motion should graduate into SE / PS / CS systems.

10. Value is still the boss.
Technology is only durable when it ties cleanly to making money, saving money, or mitigating risk. If you can’t connect those breadcrumbs, you’re building vibes, not outcomes.

Follow Abbas Haider Ali

LinkedIn: https://www.linkedin.com/in/abbashaiderali
X (Twitter): https://x.com/abbashaiderali
GitHub: https://github.com/AbbasHaiderAli

Recommended books

Scaling People by Claire Hughes Johnson (Stripe Press)
The Culture Map by Erin Meyer
Job Moves by Ethan Bernstein, Michael B. Horn, and Bob Moesta
Startup CXO (Techstars) — (as referenced by Abbas)

GTM 174 Episode Transcript

Abbas Haider Ali: 0:00

If your customers are expanding with you, you actually don’t have product market fit.

Sophie Buonassisi: 0:04

This episode explores why the next phase of SaaS growth isn’t about who acquires fastest, it’s about who can endure. Abbas, SVP of Customer Success at GitHub, explains that growth in the AI era can mask fragility unless companies anchor around durable retention, real expansion, and long-term customer outcomes. You’ll learn why SaaS is shifting from a growth era to an endurance era, why the long-standing customer success investment benchmark is shifting from 10% of revenue to 7%, how founders should think about allocating budget across support, onboarding, adoption, and outcomes, and much more. You’ll also hear why the future belongs to AI-powered specialized generalists and which leading and lagging indicators matter most for building durable, expanding revenue in the AI era.

Abbas Haider Ali: 0:50

You have to watch the lagging indicators for revenue, but you also need to think about the leading indicators for endurance in your business.

Sophie Buonassisi: 0:58

All right, let’s get into it. This is all a dream.

Abbas Haider Ali: 1:20

Let’s see how the conversation goes, then we’ll then we’ll make the final judgment on that.

Sophie Buonassisi: 1:24

Deal, deal. Sounds like a plan. And now got a pretty big topic topic to cover today together, because everyone’s really obsessed with AI-led acquisition. And if retention breaks, obviously none of it matters. So let’s start with really the truth about what’s actually changing. Like you’ve written about growth endurance as a problem. How does retention become the governing constraint?

Abbas Haider Ali: 1:48

Well, so if you think about like uh when we see all the record break and gross, there’s always new milestones. First to 100 million, first to 200 million, first of three, first of five, first to a billion. Like we’re breaking records, it feels like on a pretty much a quarterly basis, someone else carries a new banner. But as you dig into any of these businesses and others that are growing, you quickly realize that unconstrained growth is amazing. Like it’s like a being on a rocket ship. But when you’re under the experience, it can actually mask a lot of underlying challenges. And so one of the things I like to talk to companies about is no matter what’s going on, you really need to pay attention to what’s happening on a couple of key levers. From a lagging perspective, you do watch things like gross and net retention for what’s taking place. And this is particularly true in the AI world because things change so quickly that actually your initial value, you might erode it yourself. So let’s think about like a model, for example, right? Like we’ll isolate it down to like one generic model company that a research lab that’s producing producing models. You have an amazing customer set, amazing use cases. Customers are talking about value that they’re realizing. And then you yourself ship a new model. And all of a sudden, the usage for that old model will just like fall off a cliff, right? And you just have to make sure that your new model actually captures those use cases and keeps cruising past what you had before. And so you have to watch the lagging indicators for revenue, but you also need to think about the leading indicators for endurance in your business. And we used to think about those as things like API consumption or user counts and things like that. But there’s more complexity in there underneath this as you swap out wholesale models and foundational pieces where parts of your product, you’re like, this has been great. You don’t build on top of it. You kind of take it and like you throw it to the side and you bring out a new thing. And so that tracking of the underlying association with value for your customers and the endurance of that becomes super important as a leading indicator for the lagging gross and net revenue uh measures that you look at.

Sophie Buonassisi: 4:00

Super interesting where you break it down, leaving lagging. And I know you benchmark also, and there’s a 10% benchmark, but you updated that to 7%. So what changed and why does it matter?

Abbas Haider Ali: 4:11

Yeah. So look, I think uh a lot of times people assume that I’ve been doing this CS thing like for my entire career. That’s actually just not true. Like I’ve done product, engineering, marketing, sales, pre-sales teams. I just happened to sort of be in my customer success era. Um and so as I went into taking on these sorts of roles, one of the things that I looked for was models for investment. Um, and what I found was a little bit of fuzziness. You get scenarios that talk about how many people you need to hire per the number of sales reps you have, or the amount of portfolio revenue you give to people. But I found those to be very difficult to model and apply to different businesses. Like if you apply it to a business that is PE based versus venture-backed versus a lifestyle business or a super top 10 growth uh you know company, those things don’t translate. And so what I look for is a measure that allowed me to think about like uh as a leader, as a CCO, you sort of get like an envelope, right, from your CFO that says, hey, I trust you with this to do the things that we need to make our business successful. And I was like, what is in that envelope? Like, how did we decide what in that envelope? And as I talked to different leaders and looked to extract the content from it, uh, you know, until last year, I used to talk about that number as being like 10%. Like I’d gone through and looked at all the variances or whatever, and that was sort of like the marker. You could go a little bit more, a little less, depending on different business dynamics. But that was about where the biggest part of the curve sat. Um and I was very and I shared that on various stages. I shared it at Saster at TSIA events, and everyone would listen. I was like, this is what goes into the number because people would ask me all the time. After you do it for a little while, people assume you have the answers. And I was like, well, look, man, I spent a lot of work trying to figure it out. So I’m gonna share it with everybody so that it’s very easy for people to run onto the same number. And then things started to change. Um, what I realized was as uh in sort of the fall of last year, I felt like we were still in the experimental chain stages of impact of AI for efficiency and productivity gains for uh CCO scope and CS teams in general. Like it was still like we were dabbling in it a little bit. But the numbers weren’t there to say, do I change what goes into the envelope? Because that’s a big bet, right? Like when you say to your CFO, hey, I’m willing to take less and deliver the same outcome, you’re not gonna get that back. Like, you know, I know a lot of CFOs, they’re amazing people, but when you ask and you tell them I can do with less budget, they’re gonna be super happy and not be willing to give that back to you. Oh, yeah. So ultimately, as I was looking at the efficiency that I was seeing through a lot of conversations with lots of different organizations and you just observing what was happening in the marketplace, it became very clear to me as we got into the spring of this year that actually the experiments were producing consistent and durable results. And if I calibrated all the pieces that were coming together, the net of it was it felt like that 10% number now was too high. Unless you could really justify incredibly high growth and super steep NRR, you just couldn’t really justify it anymore if you were doing the right things in terms of adding AI into the mix of what your team was doing across all of the post-sales functions. And so I wanted the number to be durable for at least a year. I don’t want to change my mind every three months. It makes it look like I’m being super flaky about the whole thing. So I did the math on various things I was seeing, and it looked like 7% was turning into the new number. So seven became the new 10. And that’s what I talked about in uh the articles that that you mentioned earlier that I’ve been talking about is that’s the number, and I believe confidently that that’s what lasts for at least another six months. I’ll I’ll I’ll uh go as far as predicting six months in the future.

Sophie Buonassisi: 8:10

Okay, that was gonna be my next question of where are we where are we gonna sit when we actually have this competition again in one year? But I guess we’ll revisit it when we can sit there.

Abbas Haider Ali: 8:18

I don’t think it’s gonna be as a bigger drop significantly. Like I might say, like, hey, yeah, maybe you can get away with like six or six and a half, but I’m pretty sure unless like we have some amazing breakthrough, which like we just don’t know, I I don’t think it’s gonna be materially different than 7%.

Sophie Buonassisi: 8:35

What were the biggest labors that enable that big drop right now? Ted to seven.

Abbas Haider Ali: 8:41

Yeah. So um I’d say the early experiments where a lot of the AI tooling really focused in on inward-facing productivity. And I think what the biggest change that I saw was the ability to actually have a lot of the AI pieces become customer-facing, because that’s where you get to leverage. Um, so for example, in support, you don’t have to build like an internal ticket analysis and ticket response response sort of system. That the all the systems we have today are becoming confident enough and good enough where you can have them face customers directly. And there’s a whole set of you know companies that are building products in this area, whether you’re you know, um at Intercom with Fin or with Maven AGI or Decagon, like lots of companies are building amazing businesses around these use cases because it’s no longer like kind of a uh let’s be a little careful about this, the experiment. These are working at production scale for individual users all the way to serving the needs of large enterprise customers. And they’re working. Customers will say, I’m happy about this because it’s producing good results for me. So that’s within the support pillar. And you could extend that to all the other teams that exist within the post-sales function as well. Not every company at every spend level is going to be able to afford to pay for professional services from the vendor. Like you’ll never have the scale to do it. But if you capture enough knowledge, and the knowledge capturing is a whole conversation around best practice for doing that, but if you’re doing good work in that, in that area and expose that expert system to customer, it will help them drive adoption and unlock use cases. You can do it for CSMs to help build a plan of I just bought this thing. What’s my 30, 60, 90 day plan look like to unlock value? You can build these systems that now actually work reliably enough where they can be customer-facing, and that’s the biggest scaling lever on unlock.

Sophie Buonassisi: 10:32

Incredible. A quick pause for a company we’re a huge fan of. Because if you run go-to-market, you already know the problem. Your data lives everywhere: spreadsheets, CRMs, sales calls, ad platforms. Yet you’re still guessing what to do next. Hockey Stack is the AI platform for modern go-to-market teams. It unifies all your sales and marketing data into a single system of action. Built-in AI agents help teams prospect the right accounts, improve conversions, close and expand deals, and scale what works. That’s why teams like Ring Central, Outreach, Active Campaign, and Fortune 100 companies rely on Hockey Stack to eliminate wasted spend, take better decisions, and make space to think. Learn more at hockey stack.com. That’s H-O-C-K-E-Y-S-T-A-C-K.com. And now you mentioned a lot of people are experimenting with AI, yourself included, at GitHub. And if you read it, Cassie Young, she’s the GP of Primary Ventures, she actually had a really interesting piece that she published where she expressed that we’re heading towards a gross retention apocalypse where essentially innovation, budget, ARR, collapses and only products that deliver real, undeniable, fast ROI actually survive.

Abbas Haider Ali: 11:42

Yeah.

Sophie Buonassisi: 11:42

What does that actually force founders and operators to change about how they design customer success today?

Abbas Haider Ali: 11:50

Well, look, I think ultimately, um as is always the case, every organization gets enamored with customer acquisition. And so you focus in on, you know, like what’s a primary measure you’re going to go raise money on? It’s going to be things like CAC, right? Like what what does your acquisition cost look like? I don’t think we spend enough time talking about retention cost or growth cost. Because ultimately, when you acquire a customer, if you don’t do a great job of sustaining that customer through durable usage and durable growth, you’re basically creating a problem in the future. Um, it’s what I had to think about for myself, right? Like past the boss really doesn’t like future bus. And if you really like future bus, you’re gonna make sure that you set up for success down the road. And the way you do that is you have to invest in the success of your customers. And that take that starts from the very beginning. You need to have a really strong hypothesis on why someone is buying your product, not just because it’s cool and people want to use it and it sounds great. You really have to think about from a pure founding perspective, what is the value equation that you’re delivering? Um and you know, just as a foundational principle, really, why does anyone buy a new piece of technology? Like why they don’t?

Sophie Buonassisi: 13:02

Yeah.

Abbas Haider Ali: 13:02

There’s really only three reasons, in my view. They’re doing it because it helps them make money, it helps them save money, or it helps them mitigate risk. And any product that you ship out there that you’re out there in the market representing, you have to be able to connect the breadcrumbs to get to one of those three endpoints. And if you can’t, you really haven’t honed in on the true value that your product unlocks. And I can pretty much guarantee that either you’re counting on your customers to figure it out and they’re like, well, they know why they’re buying it. But a really strong founder will build that into the DNA of their company, saying, like, what is this thing that we’re doing? How do customers value it? And then make sure it lands in how they build product, how they sell it, and most importantly, how they build endurance by making sure it’s adopted and grows successfully. Because you have if you haven’t gotten the customer to go from the point where they’ve bought into the vision of what your product promises in terms of value to realizing that value, you’re just adding risk into the business in the future.

Sophie Buonassisi: 14:08

Definitely. Yeah, absolutely makes sense. You want to invest in that endurance, like you said. So let’s say somebody’s got the envelope. It’s now not 10%, 7%. If you were to teach a founder or operator how to actually allocate that 7%, what’s kind of the simplest framework?

Abbas Haider Ali: 14:26

Yeah. So I think about this as sort of like a waterfall. Like, what are the ways you think about investing this envelope? Um, the one thing that you have to do is I always tell people like a very important part of how you manage your finances is you probably should like try to max out your 401k, right? Like there’s like some basic things you kind of want to do. And in this world, in my view, is like you have to handle support. And that’s a really important starting place is how do I make sure that support is adequately funded? Now, the key behind that, of course, is support is when your customers have a problem, right? They’re calling you to say product’s not available, something’s not working, the speecher isn’t right. Like it’s usually like the pain scenario where your customers are calling you. But that doesn’t unlock adoption and growth though. So you have to start investing there, but sort of counter to my initial point, you have to invest as little as possible. Like, what is the minimum amount of investment that you’ve put in to actually go through and uh and go and actually have the customers be successful in the support run? Uh, right. So that’s basically like that particular scenario where you want to solve for that. Now, once you solve for that, you’re like, all right, my customers are taking care of on the reactive support side. What’s the next place you want to uh uh we’ll put uh put money in is going to be maximizing the pool to activate your customers, get them through the most critical gaze through product journey. Well, onboarding is certainly very important, right? They don’t get onboarded, they’re not going anywhere fast. So we got to make sure we we invest on the onboarding side. And then you want to invest in the really the customer outcomes side of the business, which is they bought your product for a reason, for a business outcome, for again making money, saving money, mitigating risk in some series of steps that gets them to that goal. You need a team of people that essentially manage your customer through that journey. Like it’s super important that they’re on that road. Those are sort of like the two big switches, right? Reactive customer success, which is your support side, and proactive customer success. That’s really the first fork in the road you go in the journey. Now, there’s optimizations to this model, right? If our CFI CFO is giving me 7%, I can stay within that 7% envelope, or actually I can leverage that 7%, right? So this goes into like, I don’t know why I’m going to this weird financial model thing, but we’ll keep rolling. Oh, it’s like I’m gonna like use this money and like lever it up. So, how do I lever it up in the world of customer success? The way you do that is by monetizing. So that’s really the the place you do that. And there’s a and you can monetize in both those forks on the reactive support side. Do you monetize is you create offerings that are premium support tiers? And it interestingly, I find this formula to be very easy to repeat. If you go, if you’ve acquired like a large enterprise customer, you tell them, look, you’ve got a choice. We’re gonna give you support. You can always call us, we’ll always be here for you. But we have an offering where you can pay a little bit extra, let’s say five to eight percent of your subscription fee. And in exchange, when you call us, if you have a problem, I’ll give you an SLA, you’ll be front of the line, I’ll guarantee a response, you know, like all the sort of like VIP type of capabilities. And you know what? Every large enterprise customer, if you said the choice between would you like to have someone lift up the velvet rope for you, or do you want to wait outside? Everyone wants to go to go around the velvet rope, right? And so that actually is a pretty easy sale. The great part about that, you’re now actually not having to spend as much money from your 7% envelope. Customers are paying for it. You can agree with your with your with CEO, CFO, like what’s the margin rate you want to run that at, and you’re able to go through and optimize those sorts of situations as well. Um you can do it on the proactive side, it’s professional services, right? Most at large, if you’re selling in the enterprise segment, they’re used to paying for services for no products. And again, you do the same thing, it allows you to stretch your 7% out, and that’s long term. Uh after series A, usually, is when you have the ability to do those sorts of things. You have the ability to stretch this out to maximize your investment dollars.

Sophie Buonassisi: 18:45

Super interesting. And then what about pre-Series A? So series A is when you really have that choice to be able to stretch out that seven percent. We don’t yet have that choice. What about, you know, uh a little bit earlier, like a seat stage company, where are the best places for them to use that several percent?

Abbas Haider Ali: 19:02

Well, so look, I think this is something that um I talked about at Saster in the spring. Um, I get a lot of questions from early stage companies, you know, companies that we’ve worked with all the time, right? Where they’re like founder led, they’re just getting started. They’re like, what is customer success? Right. So that is probably the most uh common question I get from founders is like some flavor of what is customer success? What do I do with customer success? Who do I hire first? Like, how does how do I get this thing going? People are saying it’s important, endurance of my revenue is important, adoption is important as well. How do I get going? And so I think about this as like an evolutionary journey. Um, the very first phase of the company, of course, like at as is true with any given part of the business, is sort of that like you’re really early, right? Like we’re talking pre seed million, you know, million dollars, ARR or less. You’re really in the point where customer success isn’t even really a thing. Even if you have revenue, they’re not real customers. I I tell founders all the time what you have are design partners who are willing to pay to design. I did design with you. They’re not really likely going to be even necessarily customers in the long run. You might be lucky to have them, but the odds are your vision might change in that early stage and you’ll kind of part ways. So I don’t think of those as pay customers. Some people don’t like them like you have paying design partners for you to build your business out in that pre-seed sort of million dollar-ish ARR phase. And then you get to like the next point where you’re in like the low single digits of ARR and you have or at seed stage. Under those conditions, really you’re in the world of founder-led CS. So you’re in a scenario where you’re going out there in the world and you’re proving to customers that there is a pain point that your product solves and that they’re willing to pay for it. These are your actual first like real customers. They’re not design partners, but you’re actually selling them, right? So we talk about founder founder-led sales. That’s kind of the world you’re looking at. At that point, really the split, you don’t really have much of this split between support and customer success teams, proactive, reactive. It’s all kind of like one big amoeba, right? Like you’re kind of in that mode, everything is all of this giant thing, and you just call it customer success, maybe. Um, and the people you hire for that role are not specialized in one way or the other. What you’re looking for are people who can switch back and forth across these conditions. They’re your customer success unicorns that you’re basically looking for. And that gets you to like single digits ARR. Through there, through series A, you do start to see this initial split between like this reactive support people who feel like they’re a proxy between your customer into your organization and into engineering, but like you’re not talking directly to engineers. They help filter out a lot of stuff and then bring in engineering as they need to. And that’s sort of your proto support organization that’s coming into being. And then you have a team that tends to be around, you know, going through and saying, look, we were doing founder-led CS. Now I need to prove that we can do this without our founder, right? So it’s sort of like founder absent customer success is really what you’re trying to build in this world.

Sophie Buonassisi: 22:06

Less catchy name.

Abbas Haider Ali: 22:07

It’s not, yeah, it’s not, it doesn’t sound as as uh snazzy for sure. It kind of makes you feel like you’re pushing a founder out of the way. But in the same way that at some point you have to have sales that can be led without a founder, you have to be able to have customers be successful that a founder. And that’s sort of what needs to happen to really make for a strong Series A, is you’re looking for certain signals of product market fit, as we uh talk about, you know, in these sort of conversations. But you’re in a lot of cases, I think people mistakenly assume, like, oh, we’ve had a few customers, we’re new. Like, that’s not product market fit, right? At that stage, if your customers are not expanding with you, you actually don’t have product market fit. Because you’ve you’ve got if you’ve already sold out a few customers at that early stage, that’s too soon. Um, so the expansion signal is what you’re looking for, and that’s your first split C starts seeing these reactive and proactive setups.

Sophie Buonassisi: 22:58

Got it. So signals sounds like look for expansion, not necessarily renewal. And that’s when you might might want to get proactive reactives, yeah.

Abbas Haider Ali: 23:07

That’s right. Yeah, you have to be able to expand and you have to be able to do it without your founder, right? Like that’s sort of the other thing that’s really important.

Sophie Buonassisi: 23:14

Yeah. And you talked about the CS in generalist. Do you think that’s a durable hire beyond Series A, as things get broken out now with AI? What does that look like?

Abbas Haider Ali: 23:25

Well, so you know, I think there’s um when we talk about using a lot of these AI tools, what powers do they give the the uh you know people who are doing these things? So in historically, we’ve built a lot of these specialized teams. So you have someone who is a super detailed product area support engineer. You’ve got someone who’s really detailed at implementation of your product, you’ve got someone who may have like migration from another tool skill, you’ve got someone who’s like a business consultant type, someone who helps you build your success plan, manage risk, right? Does your training for you. When we think about roles, this applies in customer success, it applies in sales. We have built these specialized roles that serve a very specific part of the customer journey. Now, our organizations reflect that in a lot of cases, especially as you scale, because that’s what customers want. They want to talk to experts in these domains to help them through various stages. It is really hard to find a human being who can do all of it, which is why we talk about early go-to-market roles are unicorns, right? And you know, I’ve seen a few unicorns. I have never seen a herd of unicorns. Like that’s not a thing, right? So, like you’re kind of like, well, you sort of want to be do something that’s more scalable. And so that’s when you look for horses, right? So again, terrible analogies. This will look great in the in our uh in our debrief, but you know, like you’re looking for these people who can run, run at pace, but they can be generalists. And I like this term specialized generalists because they can cover lots of functions, but what does it mean to be specialized? In this world, being specialized means you’re able to utilize the technology of the day, which is AI, to allow you to deliver the specialized skills in that general purpose function. So I can basically uh so you know, I can basically open up a tab for an AI for uh in our customer conversation, and I can answer a support question. I can answer the question that a services engineer can answer. I can help you do a plan for how to drive adoption for products, and I can do it because I have these tools available to me to do to provide those answers. And my specialty is using them to actually serve the needs of my customer. So, what I’m looking for are people who spike in certain skills, spike high in customer empathy, spike high in portfolio management. But the other thing is spike really high on in using AI technology to serve all those functions together. And that’s part of how you get that increasing scale as well. But that’s a skill that is not super common yet, right? Like if you go across the market, you’ll find operating specialized people, but not people who can cross those boundaries because that’s how we have built talent for many years in uh, you know, leading to our current uh organizational designs. But I think that’s gonna change. I think that’s gonna change going forward where any given one of these specialized generalists actually operates more like a cross-functional manager of those teams rather than a specialist on any given function.

Sophie Buonassisi: 26:28

How do you gauge someone’s aptitude to be able to use and leverage tech to be a generalist when it’s such a new thing, AI?

Abbas Haider Ali: 26:38

Yeah, uh, I wish I had a strong test. I I don’t I don’t actually have a great one of these. Really, what you’re looking for is aptitude, desire to learn, and demonstrated skill. Um so someone who like leans into the tap. Um, there’s you know, people in the same way we think about demonstrating portfolios for like an artist or a software engineer or designer, you have these portfolios that talk about the things that you’ve done. I think that’s what you’re looking for is something that shows that you’ve adapted these technologies to help you be successful in your role. That’s really what I look for. Like someone who is comfortable tinkering, comfortable leaning into new tech. They’re like, oh, we used to do things this way, and then I changed it. Like, here’s what we did. We made this switchover, and they didn’t do it because it was a top-down mandate or uh, you know, like someone figured it out for them. They saw the opportunity themselves to disrupt the work they’re doing to make them better at their craft, and they took it. Like those are the skills you’re looking for in these unicorns.

Sophie Buonassisi: 27:36

Yeah, it sounds like proactively identifying areas to approve and then actually making those changes. I have heard somebody express that the question that they ask in the interview process is what are the differences between some of the core LLMs? Because they find if somebody’s truly interested in AI, they understand the nuance, or if they looked into the nuance, I thought that was an interesting one to the layer on top of the behavioral ones.

Abbas Haider Ali: 28:00

Yeah, yeah, I think that’s right. I think like asking for examples of tell me the last time you changed something, like change the process, change how you work. Give me an example of that. That’s always a great question to ask someone like, tell me the last time you changed how you work. Um you’ll get a great, you’ll get some really interesting insight. I’ve had people tell me, you know, things like, oh, I migrated away from, you know, using uh, you know, a particular note-taking product to another one. And I asked them why. And what you learn about the why is how they think about technology. Um, I talked to someone who’s a sales leader who told me they were doing a lot of their notes and markdown. And I’m like, that’s sort of weird. Like it’s not something you expect usually from like someone on the sales side. He’s like, Yeah, like I do it this way because um he was very specific to your earlier comment about how LLMs work. He said he he felt and he’d done and he’d read research that the LLMs uh loved, uh found uh I’m trying to remember the exact words. I could describe it as they found Markdown to be delicious. I think that was roughly paraphrase, like what it was. And it was because there’s so much training data in Markdown that actually the structure is very easily data driven and it’s still in language and in text and easily consumable. And so he had better luck reusing his notes to answer questions afterwards if he took them in Markdown. And it was a fascinating exercise of the you did a thing, why’d you do a thing? Tell me about the research. And I was like, this person, they know what’s up.

Sophie Buonassisi: 29:28

They get it. Yeah. So you talk about how we’re transitioning really from specialists to generalists in great part because of AI technology. Another role that we’re hearing on our side of GTM Fund is really around the for deployed engineering. So a lot of our portfolio companies, a lot of founders just in our ecosystem, have been fielding a lot of questions, or we’ve been fielding a lot of questions from about it. And that’s really, you know, should I be bringing on for deployed engineers? Is so who should they be reporting to? Should it be CS or should it be engineering? And just what does that structure look like for a company that has not yet introduced forward deployed engineering, but is interested in doing so?

Abbas Haider Ali: 30:09

Yeah, well, look, I um I try to be a little wary of fad roles, right? Which is like you don’t want to be like, oh, like everyone else is doing it, so I’m gonna do it too. So the first question when founders ask me is like, do I need forward-deployed engineers? Is I try to understand the why behind that they think they do. And a lot of times they’ll be like, well, you know, Palantir is doing it. I see a lot of AI companies doing it as well. So maybe I should do it too. And what I’m looking for from that conversation is really to get a sense of a couple of different things. Um, and it speaks to what I think a Ford Deployed engineering organization can really deliver. And to me, it’s really the best place where they can provide input is when you want a short circuit path between customer-facing technical people and your product and engineering work. And you want that for one of two reasons. The first reason is you’re actually trying to evolve your product very quickly with deep feedback from a product standpoint from the marketplace. So you’ve kind of built the beginning of a thing, but you’re not really sure what features you’re even missing or in terms of completeness. And that’s a good way to do it. You’re actually going out there in the field, you’re testing out what already exists, but you’re quickly providing feedback back into the system to go through and make those changes in near real time as the product evolves. So it’s in a high product evolution rate environment that short circuiting can be very, very valuable. That’s one lens. The second scenario is where you have technology, but not a use case yet. You may have the beginnings of it. So going back to our value conversation, right? Unless you know, make money, save money, mitigate risk. You might be at the point where you’re like, I know this can deliver value, but I can’t quite connect the breadcrumbs. The use case feels fuzzy to me. Like, does this applause apply in a financial services org? Do I get more value in retail, more in manufacturing? Is this more like government tech? Like you kind of have a sense of like raw tech, but you haven’t really built that strong perspective. Um, you know, one of the most useful assets I think a founder can have, really through any any part of the company journey, is a value map, something that essentially is writing down all the use cases that your product delivers, who cares about them and how they measure that value. And the goal of a Ford Deployed engineering team is to help you draw that map. So when you start off in any given founding stage, you might be able to see none of that map. It’s like all sort of fog of war type of thing. You’re in this map and you don’t know where to navigate. FDEs are a great way to actually figure out this map and navigate it. But every time you’ve discovered a part of the map, it has to go into business as usual mode and goes to like other teams, whether it’s CS or sales, to run through their normal motion. I don’t believe that they are a durable part of any given use case long term because you shouldn’t need it. Once you’ve actually discovered the path, you sort of need this the system to scale. And FTEs are not a scaling lever. They’re a discovery and rapid evolution uh model within the GTM work.

Sophie Buonassisi: 33:14

Would the extension of how that accounts scale will be professional services? And really you’re just moving forward cloud engineering over professional services once you validate it?

Abbas Haider Ali: 33:23

It depends. So uh I think you can actually split in one of two ways. Um, I started my very first go-to-market role that I ever had was as an SE. So I will fully admit to having a forever SE bias in like how I see the world. Um, and so I actually think FDEs can split in one of two paths. The team that you always want at the forefront of discovery of use cases and value are going to be your SEs. So I do think FDEs can move into senior SE type roles. So whether at like the staff level or you know, if you go to the point where you have field CTOs, they can be really strong in that world where they’re technically grounded, go-to-market oriented, very strategic in nature. You don’t want them going on doing, you know, 15 demos a week to like your regular customers or doing POCs that have been done before. You want them at the edge. So I think there are certain SE roles you can put in that world. Other FDEs, I do think, actually become your professional services team. But again, not your regular day-to-day professional services. So if you do, you know, uh onboarding or migration uh professional services and you do it every day of the week, that’s not where you want to put your FDE type skills. You want to put them at the point when we say, we just shipped a new feature. It’s going out to uh, you know, our financial services customers for the first time. We’re going to implement it. They’re willing to pay for us to implement it. Who do we send? There is no prior. So who are you going to send? The FDE turn professional services, right? So I think it’s a little bit org dependent, but those are the two pathways I think they can fall. And I believe the same thing applies for reporting structure when those organizations are young as well. Is you can choose um where to report into. I actually think a lot of cases, you probably don’t really have much of a split between go to market, like it’s all one big blob of go to market anyway in the earlier stages. So it’s not as important within the reporting structure because they’re part of the blob. Um, but as things start to differentiate, you should kind of start to think about where you want it to go. Um, I think there’s I think two schools of thought on this in general, in terms of like, do you need it or not as well. If you are in a place where use cases are well known, if you’re in a market that use cases are well known, you probably don’t need FDEs. Um, it might sound cool to say, you know, all my friends have FDEs, I should have one too, but you probably don’t need it if you’re already going on well-defined pathways, but you’re building like the next wave of tech, right? So in every generation of tech, there are like certain technology patterns that repeat. Like there’s going to be security tools. There’s always security tools. In every era of technology, there always are, there are going to be AI security tools. Do you need FDEs to go out and sell a security tool? I don’t know. I I would be skeptical about that because like there’s well-known pathways, well-known use cases. You may be applying technology in a different way and a different domain, but the pathways are very well known. But if you’re doing something that’s never been done before, there’s really no prior art or use cases or you know, value maps for you to operate on, then I think you have a stronger case for an FDU.

Sophie Buonassisi: 36:35

Mm-hmm. And I mean, one of the biggest things that we hear across our portfolio, because we invest preceded Series A predominantly, is they’re looking for product projected fit. Everybody’s working and iterating and better understanding their customers. It almost feels like what you’re describing is a extended version of an ICP mapping exercise where it’s actually deeper, more intentional. I’m curious your thoughts there. Like, can that replace it? And if so, it sounds like a Ford Deployed Engineer would be a great partner in helping it do so and helping ex find a product market fit.

Abbas Haider Ali: 37:09

I think you’re exactly right. I think like, you know, when we’re looking for when you’re looking for your ideal customer profile, we’re going back to our conversation a little while ago around we focus on customer acquisition. Your ICP tends to be about acquisition, right? Where your value map really thinks through the whole life cycle. It thinks about how you build product, thinks about the types of customers you want to acquire, and lets you actually track on that map like my checking off the boxes to say, yes, we have actually delivered value in this domain. They can measure it in these, in these ways. I think that is really important as a mechanism to do that. So I think these are really part of the same thing, just a different lens that you apply to.

Sophie Buonassisi: 37:45

Abbas, you’ve been on GitHub now for over three years. And we actually had Elizabeth Camerall, the CRO of GitHub on a podcast quite a little little bit ago. And she’s been at GitHub for over 10 years, so over a decade. And one thing that she identified to be clear was that having the longevity of an employee is massively advantageous for growth. And we’ve seen that across time and time again for operators, you know, even yourself who uh X Matters or over 10 years to like that kind of repetition of employees having the cycles and the continuity of a company seems like a really big competitive advantage. How do you find that throughout your experiences has proven to be true?

Abbas Haider Ali: 38:31

Well, look, I think there’s two ways of looking at this. There’s sort of like what’s in it for the company and what’s in it for the employee. I think those are this they’re related, but they’re not always the same. So for a company, having that really steep grounding in all the stories of the company in a go-to-market role, like a lot of what you do is storytelling, right? And having those stories are just in your DNA. You can say, I did this at this stage of company, right? Like your very personal experiences that you can relate for any customer situation, right? Any go-to market. Or like whether you’re an SE, you’re in sales, you’re in services and support, you’ve seen and done so much that the knowledge you have is incredibly valuable to your organization. Whether it’s for the work you do every day, for onboarding people, for acting as sort of a cultural steward, those things are really important as long as a person also grows with the changing needs of the company, right? What you how you think and operate at seed in series A is very different than where it’s series D, or uh, you know, the letters keep going up these days, but you know, uh go play Republican or not, like in all the later stages, the the things you have to evolve along with that as well. I think that can be incredibly valuable for a company. There are limits to that though. I do believe you need injection of new ideas, new energy, new experience as you grow as well, because your organization cannot be made up of. People who are at their max, right? So if you’re like, this is the biggest team I’ve ever run, right? And you look around and everyone else is sort of doing that too. You’re like, this is the most uh experience I’ve had in this function. And you look around and everyone is also in the same boat. That can sort of become challenging. So you need a couple of things. You either need like outside energy coming in, or you need like a strong culture of mentoring and coaching that sort of helps everyone kind of rise as the organization evolves. I think that’s really important, is the growth element of people is really important to do that. And you do that by switching people around. You know, in my run at at X Matters, for example, I had lots of different functions, right, over time. Yeah. And so on a personal level, that was very useful for me because it gave me a lot of variety of operating experiences and different functions. Um, so there’s a lot of stuff in it for the company. What’s in it for me in the cases where I’ve done it is I’ve learned lots of stuff. Um, my learning kept on going in different domains and different disciplines. So it actually felt fine. Like I never really felt like, man, I’ve been at this company forever because you’re constantly learning and developing new skills and doing things. Um, and I think for individuals, it becomes problematic when they’ve been in a place for a long time and they start to feel like they’re not learning. Um, for some people, they’re like, this is actually, I love doing what I do right now at this level, in this place, in this time, and they’re totally fine. Some people will start to chaf against those constraints, right? They will say, you know, I’m not really learning anymore. And that’s when you sort of run into a conflict between what’s good for the company and what may be good for the individual. And I think as long as both parties are intellectually honest about like what the trade-offs are, I think you can definitely do it. And it’s incredibly valuable to maintain that tenure because you do get to retain certain elements that are really important. Um, the converse of it is organizations that are growing so fast that I would be terrified. Like I talked to you know, the companies we talked about earlier. They’re breaking growth rates, they’re hiring like breakneck pace. And it is uh terrifying to me to think about how those organizations maintain culture. You have to be so deliberate about saying, you know, a year ago, three quarters of these people were not here. So the core is like 25%. Like you have to be very deliberate about maintaining culture. And I think anyone who’s any founder who’s going through hypergrowth, that is the other side of this that I would say is you do get to build tenure, but also if you have the flip side of it where you have a lot of new people, you have to be very deliberate about how you steep them in your organizational values and how you actually get them to be part of your organization successful.

Sophie Buonassisi: 42:50

Yeah, excellent, excellent explanation. That we recently interviewed a Jean DeWitt grocer, she’s the CEO of Vercel, and she previously spent essentially a decade at Stripe and Google before that. And she was saying at both of them, she ended up being the 99th percentile of people from the beginning. So it’s like the longevity is just crazy. The turnover is just crazy, but if you stay, you have that, you know, root like deep-rooted knowledge of the company that is so, so valuable if executed on correctly.

Abbas Haider Ali: 43:22

Yeah, Stripe actually is a really good example of a company that, at least from looking from the outside in, and and I and I know June as well, like you know, we uh just talking about maintaining cultural and energy is so important. Above the fold right over here on the next level of shelf, if I was standing, not sitting, you would actually see the book from Stripe Publishing scaling people. Um I actually think that is an excellent read in terms of how you actually think about scaling an organization. One of the best, in my opinion. So like uh very much there’s all the books on my shelf are very purposefully selected. That one is on there like uh in like a selection area specifically around scaling at org at high pace.

Sophie Buonassisi: 44:01

I love it. Well, that sounds like a fantastic read for the incredible things about it. Are there any other books that have made a particular impact on your career? I can see there’s quite a few on the bookshelf behind you.

Abbas Haider Ali: 44:11

Yeah, so you know, I um funny enough, I’m a Kindle reader because I like to be able to read and switch books all the time. Um, the books you see behind me are actually from the collections I tend to refer to the most, to refer to people the most, but they’re sort of like physical bookmark reminders more than books I’ve read. I don’t think I’ve actually read uh any physical copy of any of the books that are behind me. Um but I’ll tell you that uh I’ll tell uh the number one book that I recommend to people as a from a business perspective is a book called The Culture Map. Um I believe that’s Aaron Meyer, if I have the if I have the author correct. Um apologies if I do not, but the book is called The Culture Map, and it is a very scientific view into um how you actually think about managing global teams, interacting with teams, what their relationship building process looks like. You know, do you do business first and have dinner? Like, what’s the order of operations? What’s this uh what’s the direct versus indirect issue of communication styles? I literally have all the cultural pages bookmarked on my Kindle. And before I land in a new country, I refer back to it and say, like, okay, remember this note. Here is the cultural context that you’re stepping into. Now, of course, cultural is not, you know, uh homogenous. Sorry, like everyone is sort of different, but it sort of like grounds you in like a uh a rough sense of how those uh how those relationships might work. Um but I found it to be a super helpful read from that perspective.

Sophie Buonassisi: 45:37

Um very cool.

Abbas Haider Ali: 45:39

The other book that uh that I more recently actually added, it may be the most recent edition on my bookshelf, actually, is a book called Job Moves. Um and there is a lot of conversation, especially founders, early stage employees, people who grown earlier in their career, you know, there’s a lot of conflicting information about how you think about managing your career. Like, how do you choose to stay in a role, to leave a role, to pick a role? Like, what’s the framing by which you operate that? And that is the closest book that I have found that matches my own style of like career planning. And I do talks on it within uh my role at GitHub, at Microsoft, and other companies. I do a talk called Practical Guide to Career Planning. Um it’s sort of like my own flavor of how I like to do things. I’m like, don’t do what I do, but if you’re curious, here’s an approach that works pretty well. Um, and that is the closest book I found that actually goes into it in much more detail and is actually written by a professional. Um, and so like that’s it’s a great conversation starter there. Um, and I’ll give one other book recommendation. Yeah, in particular for founders. Um one of the challenges with founders is you tend to come from a discipline. So you may be a product founder, an engineering founder, maybe a sales founder, but a lot of cases you’ve not really had the experience of running those other functions. And now you’re the person. Like you’re like a person who maybe is an engineer, technical founder, you are now at like in the CEO seat. You’ve never had to hire a CRO before or a CFO. You’re like, I don’t really know how to interact. You have like a hypothetical sense, but practically there’s not a lot of great guidance I’ve found on how you understand what these roles are, what their dynamics are, what the people are. Um, there is a not super well-published book in the sense of like they don’t like, doesn’t it’s not going to show up in a bookstore at an airport, but Techstars publishes this great book called Startup CXO. Um, and it’s one of my most uh often recommended books to founders because it goes into like a lot of detail about each function that they’re eventually going to lead. Um and it’s probably the best source I found for that short sort of like sifting through blobs, right? Like that’s like the best written source that I found. And so it’s a very often uh often recommended book for my shop.

Sophie Buonassisi: 48:06

Incredible recommendations. And that last one will probably start recommending to some founders that we work with because it is such a challenge when you come from a specific functional area, and then you suddenly have to run them all. It’s a lot, and especially when you’re your founder mode for all of them too, and you’re actually hands-on doing a lot. So great recommendations. Any quotes or mottos that you live by?

Abbas Haider Ali: 48:29

So um there is a there’s a uh podcast I listened to a long time ago. I don’t even, it’s gotta be like probably 10 years at this point. Um, and it was a conversation with Reed Hoffman, and they were talking about like superpowers of people. And a thing they talked about there was um the concept of being an infinite learner. Um, and so that is something that on my personal computer is actually part of my background. It just says, be an infinite learner. And there’s like a whole series of things of how you demonstrate them, whether it’s like reading, trying new stuff, doing things all the time. And so that’s like a motto that I live by and I encourage people to think about is how do you sort of optimize your life to be an infinite learner? Um, it is rewarding, it keeps you at the top of your profession all the time. Um, and it’s how you are able to adapt to the ever-increasing pace of change in the world around us.

Sophie Buonassisi: 49:26

Beautiful, beautiful words. That’s a great one. And then what about if we hone back in on go to market or company building and software space? If you had to leave listeners with one rule, I suppose what would it be?

Abbas Haider Ali: 49:40

Yeah, so the the rule actually, I’ll I’ll go back and restate something I said earlier is we get very enamored with technology. Um, and you know, I’m a I I I’m a computer engineer, but that’s what I went to school for, right? So I definitely have a bias towards getting really deep into technology. Um, but it’s very easy to fall in love with the tech. Um and what I always tell founders is if you want to build a durable business, at some point that tech has to translate into value. And so the earlier that you can start to think about the shape of what that value looks like, and again, going back to why do people actually buy technology ultimately? Like what do the what does a C-suite leader really care about? Uh the spend of a technology for, it has to go back to one of those three fundamental things. And so I encourage what if you’re in sales to like ask the question in your sales cycle of do I have clarity on how this company customer is going to make money, save money, or mitigate risk? And if I don’t have the answer to that, I truly do not understand what the economic buyer is looking for, like what is in it for them. And if I’m in CS, where I’m looking to understand is have I demonstrated that value to the customer where it’s obvious to them that they’re at those one of those three destinations or combination of all three destinations, and that’s how I know I’ve secured a renewal and I’m and I’ve earned the right to expand in that account because I’ve delivered on the promise of value in a truly measurable way. If I’m a product leader, it’s I’m going to ship a thing. Which one of these things am I actually delivering ultimately? Is it if and it doesn’t mean that I would change what I ship, but I want to ask questions about how it’s going to be consumed in the marketplace to make sure I’ve got a full, complete set of features. And it’s a really great handoff to product marketing to say, we just built a thing, and here’s ultimately why anyone should care about it. And it makes your conversations really, really easy. And having a shared understanding across an entire organization of what this is, like what is the value we’re delivering it? Who do we deliver it to? How do they see the value? Forget the tech. Just talk about like purely from a value perspective. What are we delivering? If an organization has a shared understanding of that at a startup across all those functions, it is a superpower because it becomes the communication tool by which you talk about what we’re shipping, what we’re building, what we’re fixing, what we’re delivering, what we’re selling, what we’re activating, all of those things are connected into that singular story. And it makes all the interactions anchor in on ultimately the thing that builds enterprise and startup value is the business outcomes you deliver to a customer.

Sophie Buonassisi: 52:32

Incredible words to end on. Abbas says it’s been fantastic. Thank you so much.

Abbas Haider Ali: 52:37

Thank you for having me. It’s been great.

Sophie Buonassisi

Sophie Buonassisi is the SVP of Marketing at media company GTMnow and its venture firm, GTMfund. She oversees all aspects of media, marketing, and community engagement. Sophie leads the GTMnow editorial team, producing content exploring the behind the scenes on the go-to-market strategies responsible for companies’ growth. GTMnow highlights the strategies, along with the stories from the top 1% of GTM executives, VCs, and founders behind the strategies and companies.