Nov 19, 2025 · 1h 5m · in-depth
How Harness runs 16 “startups within a startup” at scale | Jyoti Bansal (Co-founder and CEO)
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this in-depth conversation, Harness CEO and AppDynamics founder Jyoti Bansal breaks down his systematic playbooks for scaling enterprise software companies, mastering go-to-market alignment, and managing 16 autonomous startups within a single organization.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Brett holds 13.6% of the talking time here. How this is scored →
Brett as informed peer 3.5 Guest teaching 6.3 Guest disagreement 0.9 Brett pushing back 1.2
speaking balance: gold is Brett, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 45:58 Bansal rejects launching parity-only products without binary differentiators
Bansal strongly rejects standard product management practices that merely match competitive checklists, insisting that launching without a sharp binary differentiator is futile.
Hardest push from Brett ▶ 56:09 Berson contrasts Harness's unbundled strategy with Microsoft's bundling modelBerson challenges Bansal on enterprise pricing and bundling dynamics by explicitly noting that refusing to bundle modules is the antithesis of Microsoft's proven enterprise playbook.
Biggest teaching moment ▶ 12:29 Bansal establishes the 25 successful customer benchmarkBansal provides an authoritative masterclass reframing product-market fit away from closed sales and onto measurable post-sale ROI delivered to at least 25 referenceable customers.
Brett holds their own ▶ 3:39 Berson points out the rarity of enterprise-first software companies going down-marketBerson demonstrates clear domain knowledge of SaaS ecosystem patterns, noting how rarely companies that achieve enterprise dominance successfully expand down-market into mid-market tiers.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Brett as informed peer | Guest teaching | Guest disagreement | Brett pushing back | Why |
|---|---|---|---|---|---|---|
| Navigating Enterprise Versus Mid-Market Go-to-Market Motions | 4 | 6 | 1 | 2 | Brett Berson prompts Jyoti Bansal on the structural difficulty of transitioning between mid-market and enterprise go-to-market motions, observing that companies rarely move successfully down-market. Bansal gives a detailed breakdown of organizational muscle memory, pricing models, and product complexity before explaining why Harness explicitly targets both. | |
| Early Customer Acquisition and Difficult Client Trade-Offs | 3 | 6 | 1 | 1 | Berson asks how Harness selected its earliest customers and whether Bansal was opinionated about client profile. Bansal details early traction at Harness and shares a striking story from AppDynamics about firing Netflix as their third customer because supporting them absorbed all engineering capacity. | |
| Day-Zero GTM and Building a Scalable Sales Machine | 3 | 7 | 1 | 1 | Bansal introduces the concept of product-market-sales fit and elaborates on building a predictable sales capacity model rather than managing deals ad hoc. Berson asks open exploratory questions to let Bansal unpack his mathematical approach to go-to-market forecasting. | |
| Diagnosing Product Versus Distribution Bottlenecks | 4 | 7 | 1 | 2 | Berson probes how to diagnose whether early revenue stalls stem from product deficiencies or distribution issues. Bansal provides an explicit diagnostic benchmark: achieving 25 verified customer value deliveries before scaling sales capacity. | |
| Structuring and Hiring the Early Go-to-Market Team | 3 | 6 | 1 | 1 | Berson inquires about the sequencing of early sales hires and when to bring in leadership. Bansal contrasts his first-time founder approach at AppDynamics, where he needed a VP to teach him how to hire reps, with his hands-on founder-led sales approach at Harness. | |
| Core Attributes of Exceptional Sales Leaders | 3 | 7 | 1 | 1 | Asked to identify traits distinguishing elite sales leaders from average ones, Bansal outlines four pillars: recruiting obsession, thorough enablement, structured sales methodology, and extreme ownership of quota metrics. | |
| Interviewing Sales Leaders and Categorizing Enterprise Sales Domains | 4 | 6 | 1 | 1 | Berson asks targeted questions on interview probing and category pedigree. Bansal segments enterprise software sales talent into three non-interchangeable domains: business applications, enterprise infrastructure, and hardware. | |
| Developing Commercial Instincts and Revenue Focus as an Engineer | 3 | 6 | 1 | 1 | Berson explores how an engineering-trained founder developed sharp commercial instincts. Bansal traces his business foundation to his childhood in India helping in his father's agricultural machinery shop, reinforced by navigating the 2008 financial crash right after raising AppDynamics' Series A. | |
| The Founding of AppDynamics and Early Venture Conviction | 3 | 5 | 1 | 1 | Bansal recounts the founding narrative of AppDynamics, facing roughly 30 investor rejections who viewed observability as an insignificant niche. Berson guides the narrative to uncover how Bansal maintained conviction around distributed tracing. | |
| The Cisco Acquisition and Lessons in Board Alignment | 3 | 6 | 1 | 1 | Bansal shares the intense three-day negotiation on the eve of AppDynamics' planned IPO that culminated in the $3.7B acquisition by Cisco. He candidly notes that board misalignment and execution uncertainty were critical factors in opting to sell. | |
| Board Governance and Vetting Venture Investors | 4 | 6 | 1 | 1 | Berson quips that it is easier to get a divorce than remove a board member, prompting Bansal on board vetting. Bansal emphasizes filtering out FOMO-driven investors and establishing clear boundaries between board oversight and executive management decision-making. | |
| Strategic Competition and Designing Binary Differentiators | 4 | 7 | 1 | 1 | Bansal explains his philosophy on competition: achieving baseline parity on critical table-stakes features while creating distinct, measurable binary differentiators, illustrating with examples from Harness CI and cloud cost management. | |
| Operating 16 'Startups Within a Startup' at Harness | 4 | 7 | 1 | 2 | Bansal details Harness's operating model running 16 internal startups with dedicated product managers acting as CEOs. Berson observes this anti-bundling approach represents the opposite of Microsoft's software strategy, which Bansal confirms is vital for maintaining best-of-breed standards in developer tooling. | |
| Selecting Internal Startup Leaders and Strategic Acqui-Hires | 3 | 6 | 1 | 1 | Berson asks how Bansal selects PMs to lead internal startup units. Bansal explains that half are former founders and discusses using strategic acqui-hires to compress two-year engineering cycles into six months while rewriting legacy code onto the shared platform. | |
| Scaled Autonomy, Continuous Improvement, and Amazon's Influence | 4 | 6 | 0 | 1 | In the concluding segment, Bansal highlights scaled autonomy and continuous improvement culture, citing Jeff Bezos and Amazon as his primary architectural inspiration. Berson adds commentary on organizational physics before wrapping up. |
Statements from this episode (33)
Enterprise product flexibility creates user experience weaknesses in the mid-market
“In enterprise, you need a high, you want lots of capabilities and lots of flexibility. Lots of capabilities and lots of flexibility could become a complex user experience in, in mid-market. So what's your strength in enterprise could become your weakness in th…”
Harness aims to shift its 80/20 enterprise revenue split to 70/30
“We are at 8020 right now. I would actually like it to become 7030.”
AppDynamics signed Netflix as its third customer during the streaming transition
“In AppDynamics, our customer number three was Netflix. And Netflix was just transitioning. And it sounds like an old time ago, very long time ago now, they were just starting to transition from DVD company to a streaming company.”
AppDynamics fired Netflix as a customer to protect its engineering capacity
“Some point in time, and that was the hardest decision I probably ever made, we decided to lose them. They were so much bigger than our customer number, you know, like they were customer number one in terms of their size, and our customer number two to 20 combi…”
Software product design must directly align with its intended go-to-market motion
“How you design the product is very dependent on how you're going to sell it. So if you don't think about it, like the product is solving the problems, your go-to market is going to be PLG and you didn't really design for that experience. That's going, not goin…”
Scaled sales organizations plan revenue around capacity rather than individual deals
“Like to me, it's like a scale sales organization is when you're not talking about deals. You're talking about capacity.”
Enterprise product-market fit requires delivering verified ROI to 25 distinct customers
“For most enterprise software products over time, the rule of thumb that I've learned is that if you have delivered value to 25 customers, you have a very strong product market fit.”
Sales struggles without successful customers indicate a product problem, not go-to-market
“If you don't have successful, happy customers, it's most likely a product problem. It's not a go-to-market problem.”
Harness uses structured pre-sale assessments to guarantee quantifiable ROI before closing
“In Harness, we run a very structured sales process where we will create a assessment before we sell, which is, this is the value you're going to get.”
Technical founders should hire a senior sales leader before individual reps
“If you don't have experience managing salespeople, try not to learn as on the job for the first time there. If you can, you know, you have the money and ability to recruit a senior person, do that because you're de-risking your learning and a lot of things tha…”
Jyoti Bansal managed initial sales reps directly until Harness reached $1M
“Actually at harness, I initially didn't hire until the first million, I didn't hire out of sales. And I said, okay, I'll just hire a few reps. I know how to manage them.”
Early startups should avoid hiring sales executives removed from frontline management
“So if you're hiring your first head of sales, I would look for someone who hasn't been too far removed from a first line or a second line manager.
If they come in from a place where it's like a
Three hundred million sales org, and they are like six layers betw…”
Recruiting ability is the primary differentiator of world-class enterprise sales leaders
“I'll tell you that the number one thing that they do extremely well is recruiting.
And it sounds like, you know, you're talking about sales.
Why is recruiting the
Number one thing.
And all I've found is that eventually the sales scaling comes down to ability t…”
Enterprise sales talent rarely crosses successfully between applications, infrastructure, and hardware
“Most enterprise software. This is my learning is you can put into three buckets in terms of sales experience. One is business applications, CRM and all that. And they're like very good salespeople who are very well-trained to do that. Like, you know, they work…”
Category leaders win on superior sales culture rather than pure product
“Normally what happens is like the number one company Is number one, not just because they have great product, but because they also have a very strong sales culture and DNA. No, most of the time you do want someone who is like spent. So they know what good sal…”
Founders cannot effectively scale a sales organization without first selling themselves
“There's no way you'll be able to
Manage and make that sales org work if you don't learn how to sell yourself.”
Jyoti Bansal faced 30 venture capital rejections while raising for AppDynamics
“About three months. And I got a lot of rejections. I got about 30 rejections.”
The first AppDynamics term sheet required hiring an outside business CEO
“Actually my first term sheet I got, it came with a condition that a business co-founder has to join a CEO.”
Cisco escalated its AppDynamics buyout offer to 2.5x the IPO price
“And this conversation started three days before the IPO Cisco came in and said, you know, we'll, we'll pay you more than you. The price you will list. And we said, no, it's, we'll go IPO. Said, no, they came with another price, which is like, we'll pay you mor…”
Board misalignment over long-term vision was a factor in selling AppDynamics
“But if I have investors who don't want to do that for the longterm, we want to do in something different vision for what the company should be. Then you're kind of misaligned and that creates a big execution and uncertainty and risk. Unfortunately, we had thos…”
Jyoti Bansal spent 25% of his time managing a broken AppDynamics board
“I probably spend, like, One to two percent of my time managing the board in Harness or maybe not even that. In AppDynamics, I had like a kind of a broken board dynamic that I had to spend like. 25% of time, my time managing the board and the shielding our team…”
About half of venture capital investors chase hot deals purely from FOMO
“Because, you know, half of them are, of the investors are just, oh, this is a hot deal. I need to chase it. And they don't really fundamentally believe in the problem or you”
Boards should replace incapable executive teams rather than making their decisions
“And I always look at like if the board believes that the executive team is not capable of making the right decisions, they should change the executive team instead of trying to make the decisions.”
Most startup board members lack the domain expertise they believe they possess
“Most of the time, board members have, don't have the domain expertise, or the industry expertise, or the expertise. They think they might, you know, that's a problem.”
Harness requires feature parity and a binary differentiator before launching products
“The rule that, that I like to follow is like, we are not launching a new product until we have, these are the table stake parity capabilities. And these are the binary differentiator that we have at least one strong binary differentiator. Maybe it's two, but a…”
Harness operates 16 internal startups across different development and funding stages
“And we have You know, 16 startups inside harness now, and they're all of these different stages.”
Bundling software modules destroys internal accountability by hiding inferior underlying products
“But we create a model that we don't bundle things. So each of these modules are sold on their own to create accountability. They will never be best of breed in them. If we don't have accountability to the customer, the minute we bundle it, then you can have li…”
Developers will build homegrown tools rather than adopt inferior bundled software
“I feel like in developer tool space, you want to build good products for each of those. Otherwise people won't use them. People will just build their homegrown things and all that.”
Harness compounds its revenue growth by launching three internal startups annually
“And every year we are launching three. So you start layering the waterfall of that. We are growing well now. Actually our business growth has accelerated, which at a certain size the growth starts coming down. Part of it is that compounding that comes from tha…”
Former company founders lead half of Harness's 16 internal product startups
“About actually half of our startup founders are former founders.”
Building a best-of-breed enterprise product from scratch takes at least two years
“To build a world-class best of breed product in each of these, it takes time. It takes about two years. Like by the time you say, okay, let's build this thing and you mature it enough that you can sell to our enterprise customer base. It's normally a two year …”
Harness completely rewrites all acquired startup codebases within six months
“Whenever we acquire a small company, we rewrite the code completely in six months to on our platform. So the, all the old code goes away on our platform.”
Overly complex sales compensation models drive bad behavior and toxic internal competition
“I don't believe in like overly complicated compensation models, because the problem is overly complicated compensation models are they drive, they could drive bad behavior easily. They don't like to do that, you know, it's the, then you start getting like bad …”