We use cookies on this site to enhance your experience. Visit our Privacy Policy for more info.

The Next Stack: Generative AI from an Investor Perspective

For over a decade, Insight has invested in artificial intelligence (AI) applications and the infrastructure underlying AI/ML development and deployment. The firm’s investments have spanned the ecosystem — from leading applied AI companies like dental AI platform Overjet, AI-driven lending software Zest AI, AI-powered care coordination Viz.ai, to generative AI like content writing tools like Jasper and Writer, data generation/mimicking tool Tonic, human avatar generator Hour One, entertainment content localization tool Deepdub, and AI infrastructure Run:AI, Fiddler, and Weights & Biases.

Large language models (LLMs) are on the precipice of accelerating fundamental disruption at both the application and infrastructure levels. It has the power to transform nearly every industry and business.

Read: 8 Tech Investors Share Predictions for 2023

The recent breakthroughs of transformers and few-shot models in unsupervised learning have allowed model parameters and corresponding accuracy to grow exponentially without degradation in loss function. For the first time, AI can
create. This is a market inflection perhaps profoundly bigger than the rise of the internet, mobile, or the public cloud. We are in the early innings of the next generational shift.  

AI as the enabler of transformative software

That said, the rising attention on generative AI and its seemingly limitless potential and applicability means it’s important to consider where in the stack value will ultimately accrue. The rise of the public cloud drove a platform shift, but the public cloud was an enabler, not a product itself. Like the public cloud, generative AI, and AI more broadly, is an enabler of transformative software, especially at the application layer. Some of the most exciting opportunities will emerge where AI is a more efficient and effective way to solve a core business problem and drive durable value long-term.

A step change in AI capabilities

Before 2017, AI was mostly about prediction tasks like classification or recommendation. With the release of Attention is All You Need, the world was introduced to a new form of neural network called the transformer model. Transformer models are created with significantly larger datasets, enabling them to make more accurate predictions and, importantly, enabling them to create.

Over the last five years, we have seen three major developments democratizing access to these models and enabling more complex use cases:

  1. New and free-to-use LLMs have emerged, e.g., OpenAI making GPT-3 open-source to developers.
  2. Computing costs have become cheaper.
  3. Innovations in MLOps technologies and infrastructure have accelerated.

As a result, developers have the infrastructure to build new, disruptive applications on top of cutting-edge, open-source LLMs.  

The rise of GPT-3 and public foundational models

Some of the most important publicly-released models in the past two years include GPT-3 (and the recently released GPT-3.5 family), DALL-E, and Stable Diffusion. These foundational models are among the current super enablers for new AI as they enable text and image generation. ChatGPT, in particular, stands out as producing remarkably accurate and detailed human-like text that is finding its way into a plethora of everyday use cases. The recent launch of GPT-4 will take GPT-3’s ability to contextualize and fill in language using NLP to a new level.  

New large language model capabilities, combined with higher data quality and availability and declining computing costs, create favorable market conditions for the rise of applications that will fundamentally change how we work. Many existing applications will also be retrofitted with generative AI superpowers. And we are excited to see companies emerge that solve the gaps in the infrastructure stack needed to support this growing suite of applications.

The evolving generative AI stack

Generative AI has been picking up steam in VC and tech circles – but where will durable value accrue? Here’s how Insight believes the generative AI stack will evolve and where we see the most immediate and exciting investment opportunities. 

next generative AI stack insight investor perspective


Foundation models

Generative AI models are emerging along a wide range of data types, including text, images, audio, and video. Some examples of popular generative model techniques and the types of data they can generate include GANs, VAEs, Flow-based models, and language models.  While applications based on LLMs have captured our imagination and demonstrated the rapid progress of AI, this is just the start. We think of foundational models as more than LLMs. 

New types of generative models will likely continue to be developed as research in this area advances. The foundation models are progressing at a rapid pace and are already often as good or better than the human generation of content. 

Domain models

As data becomes king and more foundation models proliferate, the demand for AI-first applications will drive the emergence of many highly specialized domain or vertical-specific models. The models may chain foundation models or be trained in specific data or curated styles and tailored for specific industries (e.g., e-commerce, insurance, logistics). Some of these models may become Model as a Service or will likely become full-stack offerings with applications and tooling that sit on top.  


We believe tooling will be a critical layer in the value chain, with different types of tools arising across many dimensions. While some infrastructure needs will become more acute with the rise of LLMs (e.g., the growth in image and video data, higher query and data volume), in other cases new gaps in the stack or net new problems will emerge entirely. Included below are four key segments where we believe new infrastructure tools will evolve: 

  • Playgrounds: allow non-technical individuals to interact with and explore the capabilities of foundation or expert models; this will typically enable prosumer use cases.  
  • Programming Frameworks: streamline and automate AI-specific workflow needs to access and build applications on top of LLMs (i.e., enable the programmability of models). Key workflow needs include but are not limited to state management, instrumentation, and chaining). 
  • Model Lifecycle: support the training, deployment, and performance management of models relying on complex, unstructured data (e.g., MLOPs stack for unstructured data as laid out in our Scale Up presentation here).  
  • Management & Safety: manage the safety, compliance, and security concerns and requirements surrounding LLMs.  


The next generation of software applications will emerge with AI as a first-class citizen. While data-driven applications will enable a more personalized and improved customer experience, like successful traditional software, AI-first applications must solve an acute need and have attributes that suggest long-term durability (e.g., being embedded within a workflow, access to proprietary data, network effects, etc.).  

Investor POV: What we’re excited about 


The rise of foundational models, specifically LLMs exposes many unserved and underserved gaps in the current infrastructure stack. New tools will need to emerge for developers, data scientists, and non-technical users to leverage LLMs within an enterprise. Hence, we are excited about products that improve the accessibility and usability of large language models. There are numerous challenges for new infrastructure technologies to address, including the raw, unstructured quality of image and video data, the static nature of existing LLM interfaces, the fragmentation of different models (which we believe will only continue as models become increasingly industry-specific), or the need for effective governance to ensure unbiased, responsible results.  

Companies are already beginning to develop new frameworks that re-imagine prompt engineering used in the design and deployment of LLM apps and enable LLM applications to be built through composability. Model lifecycle management solutions will take on even greater importance as LLM creation, deployment, and collaboration become more complex. For example, one of Insight’s recent investments, Visual Layer, provides the API building blocks to streamline computer vision data management, a gap in the computer vision stack today. Moreover, emerging solutions in the management and safety of datasets will construct important guardrails and processes against bias in line with new regulations and expectations in the ethical uses of AI.   

We are excited about tooling platforms that make the lives of the “prompt engineer” or LLM-focused data scientist easier. They will form the backbone of next-gen model creation, deployment, and orchestration.  The MLOps toolchain will continue to thrive as LLMs continue to multiply. For instance, Weights & Biases is a significant beneficiary as the hyperparameter tuning and version control solution for most LLM builders. 

Horizontal applications

We believe every function in an organization with repetitive and/or skill-based work will be reshaped by foundation models, whether it is the democratization of coding, generating sales, supporting customers with virtual agents, or creating content for design and marketing. Just follow the Jasper.ai story to understand how disruptive a new entrant can be.  That said, some of these functions will be reshaped more effectively by agile and innovative incumbents. Identifying where large language models are better suited to enable a better feature instead of a standalone platform will be critical when investing in this space. In addition to this, it will be important to find problems that require solutions deeply embedded into business workflows and access to unique data sets.

Two areas where we believe new standalone platforms will arise are developer productivity and security. 

Developer productivity: We see significant opportunities in the creation of coding automation tools that accelerate the delivery, development, and testing of software code. Developers spend hours per day on debugging and test/build cycles which computer-generated code and code-review tools can significantly reduce. An estimated $61B can be saved per year from these tools in reduced inefficiencies and costs for software development. Moreover, democratized access to product and tech development for non-coders holds the potential to open the doors of entrepreneurship and software development for non-developers, sparking innovation and value creation in the economy.  

Security: One of the most exciting applications of generative AI to security comes in auto-remediations where computers can predict and effectively address security vulnerabilities. Moreover, AI/ML models already power fraud detection for major software and fintech companies – however, the lack of quality data is a major issue. With generative AI, businesses can create more complex and widespread fraud simulations using synthetic data (i.e. fake photographs, fake PII) that solve for the lack of real-world data and enable more accurate detection rates. 

Vertical Applications

Vertical-specific applications present some of the most significant opportunities for durable company creation, given the existence of proprietary data sets, tailored GTM motions, and the ability to embed deep into business workflows. We have been excited about sizable verticals with large and complex data that is mission-critical to businesses, in many cases combined with labor shortages and regulatory or compliance requirements. 

Two examples of industries where exciting developments have emerged and where there continues to be significant headroom for innovation are life sciences and supply chain and logistics.

Life sciences: Generative AI offers many interesting use cases, from synthetic data creation for clinical trials to generative protein designs based on protein folding models to accelerate drug discovery to research summarization of academic papers. While adoption is in the early innings, the potential to accelerate drug discovery and approvals, improve patient outcomes and deliver cost savings in healthcare is massive. Generative AI applications building in life sciences will have access to critical and hard-to-access clinical and patient datasets which create barriers to entry and enable better results; as more patient data becomes accessible -> models become better -> results and outputs become better -> flywheel starts again.  

Supply chain and logistics: 3D modeling and blueprint generation offer promising methods to automate product development, including design and component substitutions, leading to new products at lower costs with higher performance and sustainability. Generative AI will also enable document automation and contract generation to allow for better sales and negotiation insights and streamline workflows.  

Questions for founders

Insight is excited to partner with the next generation of AI founders from the earliest stages. And we believe that long-term category leadership will come from a few key ingredients: problem, data, and product. Companies must solve a problem with a clear and material ROI, where there is a tangible data flywheel and where the product is highly embedded into the customer’s day-to-day operations, creating material customer stickiness.  

Here are some key questions Insight investors ask when evaluating a startup.  We hope these are helpful for founders beginning to ideate in the space.  

  • Value Proposition/ROI: Is there a quantifiable ROI with quick time to value being delivered? 
  • Product: Is the product deeply integrated into the customer’s day-to-day operations, and so valuable to the customer that it is very difficult to rip out? 
  • Data Flywheel: Does your product rely on proprietary data and benefit from a data flywheel? E.g., flywheel between user engagement, data collection, and model performance → more engagement = more data = better models = more users and revenue. 
  • Incumbent & Other Threats: Is there a flexible and agile incumbent that can easily launch this product as a feature? 

A Former Credit Suisse and Verizon CIO Reveals Her Top 3 Pieces of Career Advice

This post is a special feature of Insights Distilled, a weekly tech-focused email briefing for busy financial services executives. Learn more and subscribe

Twenty years ago, Radhika Venkatraman received invaluable advice that served as a catalyst for her successful career trajectory and leadership style. At the time, she was working at Verizon, where she became the CIO of its Network and Technology organization. After that, she went on to become CIO at Credit Suisse’s investment bank and a CIO-in-residence here at Insight Partners.

Back then, Venkatraman found herself drawn to only the most challenging problems. She was less interested in ones that didn’t challenge her and did not exert herself to enable co-workers who were focused on those other problems.

“I would almost always be willing to work on only the super hard problems with the super smart people,” she recalled in an interview with Insight IGNITE exec Elizabeth van den Berg.

While her superiors admired her problem-solving skills and passion, her disinterest in other projects often manifested as condescension.

Words of wisdom that changed her outlook

A mentor recognized this and offered some words of wisdom that changed her outlook: “They told me, ‘You cannot reach the mountain and stand there by yourself, Radhika – you’ve got to make sure your team comes with you. You’re not accomplishing anything alone on top of that hill.’”

This advice was a wake-up call to how she could be more successful by embracing a collaborative approach to working across teams and projects. Venkatraman took this on board and quickly realized there was so much she could learn from her colleagues. Uplifting her teammates, she found, could reap far greater rewards than working alone.

“You cannot reach the mountain and stand there by yourself”

This simple advice turned her into a better coworker and leader and has enabled her to succeed in various roles at Verizon and Credit Suisse – and now as a board member at Brightspeed and an advisor at Insight Partners’ portfolio companies Quantum Metric and Writer.

Now, she has cultivated a leadership ethos around broad collaboration, and her genuine interest in her colleagues and teams translates into satisfied workers, more significant opportunities, and more cohesive organizations.

“Today, people always tell me, ‘You’re excellent at selling the value of your team,’” she said. Good leaders “can inspire people to go to places where people wouldn’t normally go,” she said, by showing genuine interest and appreciation. “It cannot be that you provide a vision and ask people to march up some hill: You’ve got to be with them every step of the way.” By being the biggest cheerleader for her team, she can accomplish what she couldn’t have imagined when she first provided her vision.

Her advice for career growth

Venkatraman also shared three pieces of advice for people looking to grow in their careers:

1. Keep pushing your boundaries.

Get comfortable being uncomfortable because all your opportunities will not come from sitting in place. Keep doing things you don’t think you can, and you will become more comfortable doing them.

2. Support others without expectations.

Deposit your network credits. In other words, go and help people unconditionally without expecting anything in return. You will be surprised at how much the same people, or some other people, will return to you (with interest) over your lifetime.

3. Maintain your network.

It’s all about relationships. We’re all building products for humans. We rarely do things for pets and dogs or nonexistent creatures on Mars. Mostly, we’re building products for people, and people buy from people they know and trust. People do business with people they know and trust. People enjoy being with people they know and trust. And so, it’s all about your relationships and networks.

Join Insight Partners’ celebration of powerful female leaders throughout Women’s History Month. For more, tune into our Real Talk from the Top session later this month.

7 Things to Include in Every Board Deck

Why do many executive teams loathe board meetings?

At Insight, we work closely with executive teams to drive growth, scale operations, and achieve better and faster results through Onsite, our team of 130+ dedicated operators in sales/customer success, marketing, talent, and product/tech. As a result, we have attended and prepared for many board meetings.

Board meetings are invaluable. They are the ultimate time and place to evaluate strategic progress, refine or revise goals and timelines, and receive insight from experienced people who are invested in the company’s success. These meetings can be a time to reflect on progress and celebrate wins and milestones.

But often, executive teams dread the quarterly board meeting. They are a lot of work, they necessitate business scrutiny, and they might even reveal issues.

Compounding this dread is that founders and company leaders are rarely told what to include in the board agenda and materials, and what makes a presentation valuable for both management and the board. Crisp, open, thoughtful, and productive board meetings build positive relationships between the board and executive teams. However, board meetings that waste valuable time can undercut the board’s confidence in management. We summarize our best practices for productive board meetings here.

Board meetings are a CEO’s strategic weapon

Effective board meetings clarify priorities; they allow leaders to spotlight results, highlight challenges, and discuss key strategic issues. The best management teams and board members hold each other accountable and truly want to review company progress and improve where possible.

CEOs and management teams get the most out of board meetings when they arm directors with the information required to ask smart, pertinent questions. And the exceptional teams focus on the places where those questions — and the answers — will have the highest impact.

Seven must-haves for your board presentation

While every company is different, the software companies that run the best board meetings have common approaches and provide similar information in their presentations. At a minimum, include these seven things in your board presentation decks.

1. Establish and stick to stated objectives

Too many of the worst board meetings begin with an agenda, but not with objectives. That is a big point of difference. An agenda is merely the data that gets presented. Objectives are what you want to get out of the time together.

Before every board meeting, the CEO should complete this sentence, “This meeting will be a success if we…”  “…get through the agenda on time,” is not an ideal response.  “Agree on a budget,” or “Set baseline goals and metrics” are much better.

2. Include a “State of the Union” from the CEO

Board meetings are most effective when organized around top priorities and issues. An update from the CEO on key accomplishments, challenges, and how the company plans to address those challenges helps orient the board. With a 360-degree view of the business, the board can provide sound advice and guidance.

Avoid the temptation to speak in phrases. “Good” is not a number. Include data. Real measurable outcomes aligned with key performance indicators will help the board – and the management team – make good decisions. It’s also important, as part of the CEO report, to include and review action items, including any from the previous board or leadership meetings.

This is also the time for the CEO to ask for guidance on key decisions. Asking for the board’s point of view on challenging decisions ensures that the board agrees with the decision and is invested in the outcome.

3. Don’t shy away from non-financial numbers

Board members can add the most value when they have access to the rich data that informs management decisions, not just the requisite financial statements. So, while you may summarize, also provide the metrics you use to monitor progress. That way, board members can help you be sure you’re measuring the correct things, in the best way.

Board members can share how your measures stack up against other businesses and help provide greater context to the data. When in doubt, hand it out. If you don’t have data on something, it may be worth discussing that as well.

Data such as financial statements, HR metrics, sales and marketing data (e.g., bookings growth and customer success metrics), and product management and development metrics (such as product roadmap milestones) are all useful for board members.

Board meetings are more effective when the materials – especially data information – are sent 48 hours in advance. To assist with productive board meetings, Insight Onsite has developed a template board package that includes the metrics that are important for effective board discussions. Your investor should be able to provide this sort of guidance.

4. Functional summaries matter

A one-page summary by function with key highlights from the quarter, near-term priorities, and current challenges lets the board quickly see what’s happening by department.

The executive team leaders in a software company (product, engineering, marketing, sales, customer success, HR, and finance) should present a dashboard of key metrics, current priorities, and progress against previously discussed priorities. Good CEOs have leadership team meetings where functional heads know the constraints and priorities of their colleagues. Where this doesn’t occur, the board meeting is a good forum to disseminate information so everyone may understand the situation. When the team has a complete view, priorities can change, cooperation can grow, and teams can be more effective. By getting visibility into these functional priorities, your board may be able to help the process along.

5. Review strategy

A board’s role is governance, results, and strategy. Too often strategy gets lost amidst the approval of board minutes and the dissection of business metrics.

CEOs should discuss market dynamics, competitive moves, environmental factors, new relevant regulation, talent retention, M&A, and company direction. The board meeting is an opportunity to get a broader perspective and review industry dynamics that may impact the business. Discussing the probabilities of different scenarios is a core responsibility of the board.

6. Spotlight your team

People are the most important asset of any business. As any good manager knows, recruiting, hiring, training, and developing the best talent is what separates great teams from the rest.

Leadership and execution are hard work for company leaders. By giving functional leaders a chance to display their potential and be acknowledged by the board for their accomplishments, the CEO ensures that leaders are motivated and aligned.

Understanding and displaying their senior leaders’ potential and performance in a board deck calls out key team members and helps keep the organization focused on talent development.

7. Seek out direct feedback

Good board meetings include time alone with the CEO for the board to provide the CEO with confidential feedback, and for the group to collectively review executive team composition, highlight capability gaps, and discuss succession plans. The board is able to provide observations and proffer help.

The CEO job is lonely. The board meeting time represents an opportunity to discuss concerns behind closed doors and obtain input. The CEO should view this time as one of the key benefits of board meetings. The collective capability of the board is focused on improving the company. Take the chance to tap into this experience and knowledge.

Board meetings done right

When prepared and delivered well, board materials help leadership teams focus on what matters and allow board members to prove their value. As you prepare your presentation and run your board meetings, follow the rules. Be honest, support your plans and presentations with data, and most importantly, seek and solicit feedback from board members. Rather than dreading the work board meeting preparation can take and seeking perfunctory sign-off, view this time as an opportunity to get the advice and investment every company needs to deliver outstanding performance. Our list of seven must-haves will get you most of the way there; the quality of the dialog will do the rest. 

Break Through with These 5 Lean Principles from Unicorn Companies

In 1988, John Krafcik coined the term “lean” in his graduate work at MIT’s International Motor Vehicle Program. His paper, “Triumph of the Lean Production System,” challenged that it was not the location, the culture, or even the technology that determined car manufacturing plant performance. Indeed, the plants that operated with a “lean” production mindset were highly productive, while maintaining high quality. Lean was his way to express what he came to believe in his previous role as a Toyota manufacturing engineer to be the essence of the game-changing Toyota Production System: “an absence of slack in the system, aka waste.” Krafcik famously went on to become CEO of Waymo, the self-driving car company spinout of Google’s parent company, Alphabet.  

Watch: What a Unicorn Knows: 5 Principles for Growth in 2023

A key part of Insight Partners’ approach to helping portfolio companies scale has centered on helping leaders eliminate these kinds of organizational impediments, applying the principles of  lean thinking in a rather unconventional way: to the operations of software ScaleUps. That work with highly successful “unicorn” companies has led to the development of five foundational principles any company can use to create rapid and lasting growth. 

Five lean ScaleUp principles

The term “lean” became popularized as a management philosophy with the 1996 bestseller Lean Thinking by James Womack and Daniel Jones, who led the MIT study during John Krafcik’s graduate work in the late 80s. In 2011, lean gained a resurgence in the tech world with Eric Ries’ The Lean Startup, which focused on helping entrepreneurs test ideas and iterate quickly.

Over the years, lean has evolved and grown to become an organizing principle that engages people in adding the highest possible value for customers across all operations. What makes lean compelling and different as a management philosophy is how that value is created. The lean process is one of addition by subtraction — reducing or removing anything that impedes the free flow of customer-defined value. Amazon calls it “working backwards.”

We have discovered that applying a broader interpretation of lean can be a powerful stance for battling the momentum-stealing effects of drag, inertia, friction, and waste. By keeping Krafcik’s original idea of “zero slack” front and center in efforts to help tech firms scale for growth, we have seen certain themes repeat themselves across various successful scaling companies. Those patterns evolved into a set of guiding principles, which, when adopted, make success more likely. The handy mnemonic to remember is SCALE:

  1. Strategic Speed
  2. Constant Experimentation
  3. Accelerated Value
  4. Lean Process
  5. Espirit de Corps

Learn more about using these lean principles for rapid, lasting growth. What a Unicorn Knows is out now!

Principle 1: Strategic Speed

Fighter pilots, professional cyclists, and race car drivers know what geese flying in a V formation know: You can travel faster and farther with half the effort by “drafting” in the slipstreams created by those in front of you. The faster you go, the more energy you save. It’s a virtuous cycle. And the more people in alignment, the bigger the slipstream, so you go even faster. This is the simple physics of momentum, the equation for which is velocity (speed with direction) times mass.

You can apply the concept to your company’s strategies. We call it strategic speed, defined as the optimal speed for swift strategy deployment and decision-making.

To produce a similar effect and create the organizational equivalent of slipstreams requires strategies, priorities, and objectives to be simultaneously linked vertically and horizontally. Mechanisms like Japan’s Hoshin Kanri (“strategy deployment” or “policy management”) and the younger but more well-known Western version, OKRs (objectives and key results), implemented with tools and practices like the lean alignment practice of catchball — essentially the business equivalent of the children’s game of tossing a ball back and forth — help boost strategic speed.

We have observed that ScaleUps achieving company-wide alignment are able to accelerate their growth over 30% more than their peers.

Principle 2: Constant Experimentation

Continuous innovation is a survival need and a competitive must. Without that capability, inertia will act as a speed governor. But innovation cannot be relegated to department status or reserved for the next-level killer app that may never materialize. Doing so is an inertia-producing temptation, but one that can be avoided by making simple, fast, and frugal experimentation an operating norm.

One of the big misperceptions about lean is that it’s all about quality and cost. Those who have spent time embedded in the Toyota culture will delight in correcting you, letting you in on the little-known fact that the Toyota Production System was developed to shorten the time from order to delivery and create a “dash to cash” method without requiring the deep resources of the big U.S. automotive companies. The entire system was evolved through a series of desperate experiments to scale up and grow revenue faster with less.

For high-velocity ScaleUps, creating a steady stream of innovative new product and process concepts that consistently make it to market requires an equally fast, lightweight, high-impact method for carrying out constant experimentation, one that is, unfortunately, missing in most.

Experimentation also isn’t just about product development. Applying agile principles to rolling out a new sales process in one market allows you to test and improve before rolling out globally to your entire organization. As Netflix’s co-founder and first CEO Marc Randolph writes in his 2019 book, That Will Never Work:

I’ve realized that the key to being successful is not how good your ideas are, it’s how good you are at being able to find quick, cheap, and easy ways to try your ideas.

Principle 3: Accelerated Value

A failure to understand and align with customers on their desired business outcomes can produce enough downstream friction to produce what every recurring revenue business dreads: churn.

The tendency is to equate the concept of a customer journey with a sales funnel coupled with a monolithic view of the customer, which is wrong. In other words, customer = account is a key source of friction that can ultimately lead to head-scratching when seemingly satisfied customers churn.

At the root of the issue is the difficulty of thinking and operating horizontally in a structurally vertical world. Customers are organized vertically, as are most company support functions, but the customer experience is horizontal. Rather than think like a star quarterback leading a team with set plays being sent in from the sideline (vertical thinking), think like a Formula One pit crew. A horizontally-oriented Formula One team has over 20 people with specific roles so tightly synchronized that they can stabilize the car, change the tires, adjust the aerodynamics, and safely release the car to get back in the race in under two seconds.

Enabling customers to realize value quickly promotes product adoption and positively impacts community spread, customer retention, renewal, and expansion.  Ensuring that everyone in your company is aware of how to enable that value quickly, and in a unified fashion, only helps to accelerate your growth through improved customer satisfaction.

Principle 4: Lean Process

Lean as a concept encourages simplicity as the path to speed. It holds that less is best, and that to make more room for what truly matters, eliminate what doesn’t. It’s a subtractive approach to continuously improving and simplifying even the most complicated workflows. It starts with a clearly defined value, then systematically removing everything blocking the path to delivering it. It’s a relentless endeavor, a different way of thinking, and requires a mind shift.

Targeting waste involves using a methodology over 80 years old developed by the U.S. War Department in 1940, who coined the term continuous improvement. The concept was aimed at the effort to convert the American manufacturing base to the war effort. It was then utilized to stabilize war-torn Japan under the leadership of General Douglas MacArthur during the seven-year U.S. occupation. Japan, having scarce resources other than human creative capital, termed it kaizen, meaning “change for better.”

With fast-moving tech ScaleUps, we use an adapted method of traditional continuous improvement called a kaizen blitz, which works best, as it is both faster and more effective.

When applying lean principles within Insight’s portfolio companies, we have been able to achieve a 20-30% improvement in time to value.

Principle 5: Espirit de Corps

You can’t build a Formula One car by yourself, or for that matter, a company. It takes a team and leaders of and within that team to create the kind of environment that enables the first four principles to come to life.

Enter the notion of esprit de corps. French for “group spirit,” esprit de corps figures centrally in military and paramilitary organizations, which are notorious for favoring results-oriented leadership. Mission first, people always is the mantra. But social research suggests that for a high-velocity organization like a ScaleUp, a cohesive culture of “people first, mission always” may just be a better approach.

As UCLA social psychologist Matthew Lieberman reveals in his bestselling book, Social: Why Our Brains are Wired to Connect, those viewed as having predominantly strong results focus have only a one-in-seven chance of being viewed as a great leader, while those viewed as having a predominantly social or empathic focus have about the same or slightly less chance. But for those strong in both results and social skills, the likelihood of being seen as a great leader is five times greater.

Leaders of this ilk understand that a people/culture fit is every bit as important as a product/market fit when it comes to scaling for growth. Your star product requires a team of star players to advance it to market and capture maximum value…so much so that Netflix is happy to advertise to all job seekers that they will pay an ill-fitting employee an industry-leading severance of four months’ pay while they search for a star replacement.

What a unicorn knows

A cursory glance at each of the individual principles in the S.C.A.L.E. framework might lead you to ask whether there is anything really new here. That’s fair. What is unique is the lean interpretation of the principle: Well-worn terms like strategy and experimentation take on entirely new meanings when viewed through the lens of lean. What is unique is the synergy created from integrating any one of the individual principles with the other four and pointing the collective model toward the goal of scaling up by leveraging a lean, zero-slack mindset.

Learn more about applying lean principles to scale by reading our book, out now: What A Unicorn Knows.

8 Tech Investors Share Predictions for 2023

2022 was a busy year for the team at Insight. As hype started to build around the use of AI in our everyday lives, Insight held its first ScaleUp:AI conference, featuring top industry speakers and hosting over 1,700 attendees. The firm also grew the Onsite team — Insight’s dedicated ScaleUp engine of Sales, CS, Product, Marketing, and Talent experts — to over 120 operators to better support portfolio companies, help them focus on metrics that move the needle, and prepare them for whatever comes next.

As we wind down the year, eight of Insight’s Managing Directors share some thoughts about what’s top of mind for tech investors going into 2023.

insight partners investors
From left to right:
Ryan Hinkle, AJ Malhotra, Rebecca Liu-Doyle, Thomas Krane
Michael Yamnitsky, Nikhil Sachdev, Lonne Jaffe, George Mathew

We’re going to hear a lot more about AI.

If 2022 was the year of crypto, 2023 will be the year of AI truly breaking into the general population’s awareness.

The shift from analytical AI to generative AI

Lonne Jaffe: “Many had been operating under the assumption that manual labor and simpler knowledge work would be most disrupted by AI and automation, but with large foundation models like GPT-3 and DALL-E, we’re seeing AI systems make enormous progress in highly creative tasks like design, programming, music, and creative writing. This will likely continue in 2023 with the release of systems like GPT-4. At the moment, the reliability of these models is still a major challenge — they often hallucinate answers that are false but still ‘speak’ confidently. This kind of unreliability could be problematic for a lot of use cases, like customer service, education, and healthcare. If you don’t already know the answer, it can be hard to tell whether some AI-generated responses are correct.”

Nikhil Sachdev: “We’re moving from analytical AI (analyzing/parsing data and identifying trends and patterns) to generative AI (creating new content or interactions based on patterns). Applications we’re seeing now are benefiting from powerful (often open source) large language models, cheaper computing costs, and established MLOps platforms. These AI applications are starting to overtake human functions and have the potential to augment and disrupt existing entrenched software apps.”

George Mathew: “More of us should be talking about explainability and bias detection as more large language models (LLMs) get to scale and production. We should all be preparing for what opportunities will emerge with a multi-trillion parameter large language model like GPT-4 being released.”

Lonne Jaffe: “It will be very interesting to watch where the value will accrue and where economic moats will be the deepest. Some believe that the economic moats will accrue to the companies building the large foundation models because they require so much time, skill, and infrastructure spend. Others think that the moats will be with the companies fine-tuning the models for specific use cases because of the feedback data demand-side economies of scale. Still others believe that the value will be in the non-AI software that allows the models to integrate with real-world systems. There may even be a layer of value in between the foundation model creation and fine-tuning, requiring a new set of MLOps tools and skills that focus the foundation model for a specific domain, but in a way that is more involved in modifying the internals of the foundation model than needed during the fine-tuning process.”

Moving from AI in infrastructure to AI in applied real-life situations

lonne jaffe quote

Lonne Jaffe: “One area where we’re likely going to see continued huge progress in 2023 is in applied computer vision AI in healthcare. The tech is already approaching human ability in domains as varied as polyp detection in colonoscopies, diagnosing gum disease in dentistry, breast cancer screening in a mammogram, etc. This can improve diagnostic accuracy, save physician time, surface candidates who would benefit from clinical trials, and even reshape how the industry works.”

The metrics investors care about in 2023 will shift to retention and efficiency.

“More nailing it, less scaling it.”

Ryan Hinkle, Managing Director: “2023 is about more nailing it, less scaling it. 2023 should be a year where it’s efficiency first, additional costs second. It is really difficult to focus on efficiency when you are adding costs. That is the fundamental pendulum shift: it has abruptly shifted from ‘if you believe it, it will come’ to ‘if you can’t see it, it doesn’t exist.’”


Metrics that matter

Nikhil Sachdev: “Customer NPS is always important, even more so in this environment. (Are you nice to have? Or, I can’t live without you?) NPS flows through all the relevant financial metrics in a business. The more customer value/love you generate, the better your logo growth, pricing power, retention, and efficiency. And goes without saying in this market, it’s no longer growth at all costs. Companies and investors are focused on durable, efficient growth.”

George Mathew: “Gross retention — more than ever, you have to be able to retain customers to stabilize your 2023 growth plans.”


Thomas Krane: “Path to breakeven based on current balance sheet, cash burn as a multiple of net-new ARR.”

AJ Malhotra: “It’s all about how you’re investing to drive efficient growth. My key metrics are about the same: previously, it was all about net-new ARR, and now gross profit matters more. Your true gross (and net) retention becomes very, very important as well — this separates strong companies from weak ones. Cash burn also becomes imperative in this environment.”

Rebecca Liu-Doyle: “In this environment, two things investors are watching especially closely are gross margin and gross retention, both of which are prime leading indicators for steady-state free cash flow potential. In steady state, will this be a 15%+, 25%+, or 50%+ FCF business?”

DevOps will prioritize simplicity.

Michael Yamnitsky comments on the developer perspective: “The great vibe shift of 2023 is a return to simplicity! Back in 2017, it was cool to tinker with the nuts and bolts of Kubernetes, but as of 2022, we’ve reached peak complexity and specialization in cloud infrastructure, and the pendulum is swinging back. Developers want to simplify their stack and ship code faster. To this tune, we’ll see a resurgence of PaaS and other developer-friendly services that eliminate the toil while retaining all the benefits of 10+ years of advances in cloud technology.”

Thomas Krane, Managing Director: “In DevOps, cost pressure will put new pressure on public cloud workload adoptions and reinforce the need to have interoperability between on-premises IT and cloud services. This creates opportunities for new vendors in the space.”

Rust will be all the rage

Additionally, Michael adds: “Rust is all the rage and demand for rust programmers is growing. The performative nature of this programming language makes it a fit for backend-heavy development, particularly in the infrastructure and developer tooling space where performance can be a key differentiator.”

The overall economic environment will be uncertain for a while, but it’s not all bad news.

Ryan Hinkle: “None of us are used to inflation. Inflation hasn’t been a consideration for literally 30 years. Because of inflation, if you aren’t growing 8%, you are shrinking on a real basis. We enter 2023 with a great deal of known issues — inflation being front and center — but no real ability to forecast what comes next. In 2023, we will need to re-evaluate on a quarterly basis or even more frequently, as a year will feel like an eternity. Years make sense as forecast building blocks when things are well-behaved. These are not well-behaved times.”

Nikhil Sachdev: “Market sentiment is as negative as it has been since the Great Recession. We are seeing a combo of inflation, rising rates, cratering multiples, geopolitical turmoil, and de-globalization, which is impacting our supply chains. On top of that, the demand curve is being whipsawed – first as we lap a period of strong pull forward in digital growth driven by the pandemic period, and now budgets and spend tightening. It’s time to go back to basics — focusing on durable growth and building/scaling efficiently are the fundamentals that will enable companies to succeed regardless of the macro. Just remember that things are never as bad as they seem at the bottom and never as good as they seem at the top.”

Thomas Krane: “Companies that largely sell into tech companies with products linked to headcount will see a significant medium-term downdraft in revenue, but there will be a strong recovery on the other side for those that survive.”


Survival of the strongest will drive consolidation

Nikhil Sachdev: “So much of the bad news is out and now baked in the cake that on balance I think equity markets will be more constructive over the next year. I think we’ll see more private tech dealmaking. Growth-stage companies will still need to raise money, maybe at different multiples than before. We are also going to see much more consolidation as companies that can’t or don’t want to continue down the standalone path look to partner with strategics.”

AJ Malhotra: “We’ll see consolidation — lots of companies have raised lots of money, with unsustainable burn rates, cost structures that may not be efficient, and that means some will not be able to raise follow-on rounds and will need to sell. The velocity of fundraising that happened in the tailwinds of Covid from 2021-2022 was a unique moment in time.”

Ryan Hinkle: “Whatever this recession will be, it will really test what is ‘needs to have’ vs. ‘nice to have’ and inform what gross and net retention looks like. We have not had a meaningful downturn since SaaS emerged as a dominant trend in digital transformation.”

There are opportunities in uncertainty


Lonne Jaffe offers several examples of how tech, and AI specifically, could help to alleviate inflationary pressure: “The go-to reaction to inflation is to have Federal Reserve Bank raise interest rates and to slow the economy and raise unemployment. But this comes at a huge cost. Despite the anxiety around robots and automation taking jobs, there can be an opportunity for tech to help alleviate inflationary pressure by increasing efficiencies and making us all more productive. In a similar way as collaboration software helped the economy cope with isolation from the pandemic, this kind of AI-powered efficiency improvement, in a way, could become the unsung hero of this inflationary crisis period.”

George Mathew: “Backoffice sectors like supply chain, procurement, and business process outsourcing all have fundamental opportunities to be transformed by generative AI.”

Thomas Krane: “The cost of cloud services will create opportunities to preserve and even expand on-premises IT.”

Michael Yamnitsky: “One of the positives of continued economic uncertainty going into the new year: the spotlight shifts away from the hype-chasers and storytellers and towards the humble entrepreneur who has been quietly owning their craft.”

Rebecca Liu-Doyle: “Certain categories — like beauty in consumer and automation in enterprise SaaS — have counter-cyclical tailwinds, and this may be their moment to shine.”

AJ Malhotra: “New company formation will increase because of layoffs, and lots of talented folks will have new time on their hands to build something new.”


Hiring might get easier.

George Mathew: “There will be a much more available labor market as hundreds of thousands of tech workers are being laid off at the ‘Big Tech’ firms.”

AJ Malhotra agrees: “Hiring is a big opportunity right now! A lot of good people are in the job market because of layoffs. Hiring may become easier given the talent out there. We have dueling realities — giant tech companies are doing layoffs and hiring freezes, but unemployment is low. We’re still seeing hiring in many industries.”

There’s still a lot to be excited about in tech.


Nikhil Sachdev: “While I acknowledge we are in a peak hype cycle for AI, I think the secular trend is real and feels like we are on the verge of an explosion here. AI will impact horizontal and vertical segments within software.”

Michael Yamnitsky: “I’m excited about WebAssembly. It has the potential to bring unparalleled levels of efficiency and security to computing and transform the way developers organize and collaborate around code. But most importantly, it’s portable — making it a unique fit for the next wave of distributed applications.”

Thomas Krane: “Threat intel will finally get recognition as a critical baseline/foundational priority for a strong cybersecurity stack.”

AJ Malhotra: “It’s easy to be a pessimist but there are a lot of good things happening right now: hybrid work environments are better overall and have provided more flexibility to people, there’s low unemployment. There’s tons of opportunity to do things more productively and more efficiently. Tech dealing with carbon emissions and clean energy transitions, enterprise software selling into financial services, software for the build environment, and tech dedicated to improving healthcare delivery are all exciting areas right now.”

This post was compiled and edited for conciseness and clarity by Jen Jordan.

7 Habits of Effective Data Leaders

CxOs are realizing every executive in the organization is a data leader in the age of digital transformation. Whether their background is in data analytics or not, successful CxOs are navigating this transition by actively engaging with data departments to fill gaps in knowledge. They proactively build a set of practices and habits that drive useful insights. In doing so, leaders have pivoted from a defensive to an offensive data strategy and culture.

Analytics teams previously collected data, analyzed it and offered insights directly to leadership. Instead, modern data teams are much less siloed. They work with, and in support of, multiple executive officers and stakeholders in different departments, in a decentralized way, throughout the company’s digital transformation.

Data Results | By Company Size

How the leadership navigates this spectrum varies from company to company. However, highly effective data leaders have established a series of best practices to guide them through the growth curve, and continue to follow these practices until they become habits. As Pulitzer Prize–winning writer and productivity expert Charles Duhigg wrote in The Power of Habit: Why We Do What We Do in Life and Business, “there’s nothing you can’t do if you get the habits right.”

1. Shift the culture and strategy

The role of data executives can be transformative. Their ability to collect and analyze data, then create real operational value from the business insights that data reveals, makes them a transformational force.  It’s in any executive’s interest, therefore, to embrace data technologies. Every department stands to gain, from finance (maximizing revenues and minimizing costs) through enhancing sales and marketing practices (buyer intent, lead generation and opportunity conversion to order), to the C-suite (organizational transformation, long-term growth, customer churn and competitive strength).

Data leaders rely on data to actively predict customer behavior, and refine their own engagement practices and pipelines. How sophisticated they are often depends on where the organization is in its digital journey.

Data-driven organizations are maturing as they move beyond relying on low-level automation tools such as chatbots, to measure successful customer experiences, using this insight to predict future customer behavior. A 2021 Gartner report confirmed that customer service departments get most value from technologies that analyze customer data.

Objective | Pivot from a defensive data strategy to an offensive, democratized strategy. Approach > Key Objective > Core Activity > Data Elasticity

Data Culture & Strategy | Approach and Focus

2. Refine the business context

The business context in which data is collected is changing too. Increasingly, organizations are focusing on customer-centric research to enrich their analytics strategies. A large part of data leadership has focused on identifying short-, medium- and long-term metrics to better understand and predict customer behavior. Although some customer trends stay the same, others fluctuate, year by year, as the ways customers interact with technology slowly evolve.

shows 7 criteria meant to drive customer adoption.
Data Strategy | Customer-First Principles

In response, business leaders are learning to modify their organization’s operations, sales and analytics strategies and to adopt more customer-focused data solutions. Data pipeline company Rudderstack provides a data platform to help organizations implement this data-driven strategy for improved customer support. The platform allows enterprises not only to track customer data, but to directly engage with the customer.

3. Grow the data pipeline with the company

As mentioned above, data strategies shift as companies grow. Early-stage companies often have a very different set of tactics — and see very different results — than multi-million dollar scaleups. As they scale, though, data leaders realize that the range and level of data-driven insight must scale with them.

Leaders often progress from planning a data strategy that merely harvests data, to one that focuses on action by using data to shape organizational decision-making at every level. This is a crucial transition, as it requires data departments to undergo a fundamental intellectual shift, from being passive collectors of data to active advisors. In this way, data streams and enterprise systems are integrated so that predictive data can actively inform decisions.

Data analytics innovator Kubit, for instance, provides a self-service behavioral analytics platform. It harnesses behavioral insight that allows organizations to optimize sales and marketing conversion.

4. De-silo the architecture

The shift towards more proactive data strategies also involves rethinking the data architecture. Previously, organizations might have organized their data by team or purpose. Now that companies need greater volumes of better quality data, siloed datasets can hinder proper access to the data.

Many data leaders are migrating from a data warehouse-centric architecture to a data lake-centric one. Data lakes provide much greater reporting capability and analytical flexibility thanks to their accommodation of unstructured data. Conversely, data warehouses require structured data while data lakes can accept almost anything: Structured data, unstructured data, media and more.

Diagram showing Data Warehouse (Late 1980s), Data Lake (2011), Data Lakehouse (2020)

Data Architecture Evolution | Source: Databricks

However, the structured data of a warehouse works well for analytics, but is too rigid for advanced AI or machine learning (ML) models to work with. Data lakes are not without their challenges either: They provide great flexibility for AI-driven processes but are unsuitable for reporting dashboards. As a way to overcome the gaps in functionality for both alternatives, the data engineering company Databricks offers data lakehouses for both advanced reporting and agile AI.

Regardless of the implementation, executives are leading the move to a more unified, de-siloed architecture where data from multiple streams and teams can be integrated to inform better decision-making. A new trend is to master customers in the data lake versus the CRM systems, allowing for a complete 360 view of the customer available for real-time AI.

5. Build trust in data

Earlier, we described the growth trajectory of early-stage startups as they scale, in terms of their shift in focus in data strategy. This doesn’t mean, however, that DataOps principles such as governance and security take a back seat. Privacy, compliance and security remain critical to maintaining customer trust in any organization. Data leaders are always sensitive to these concerns and are constantly working to enhance trust in their company’s data policies.

Proper governance is a huge pain point for startups —  Gartner’s 2021 outlook predicted that, through 2025, 80% of organizations will see efforts to scale their business fail because of a lack of modern data governance. To make matters worse, some businesses don’t even have visibility into how effective their governance policies are. Over 40% of executives surveyed had no metrics to measure whether their data governance policies were actually working.

Fixing this problem is an ongoing challenge as executives struggle with perception as much as reality. Every time a data breach becomes public knowledge, company leaders confront a tide of general suspicion and distrust from customers, even if the breach happened to a different company.

Credit bureau Experian, for instance, saw its brand tarnished by the infamous Equifax breach that exposed the financial data of 143 million Americans.

Displays different tiers (lines of defense) for data governance

Data Governance | Prioritization & Lines of Defense

To deal with much bigger threat vectors, executives have begun instituting strict policies at every organizational level: From the adoption of tools like multi-factor authentication (MFA), to precise cloud-access control policies and updated engineering best practices for sensitive data.

Many companies prefer to buy — rather than build — their governance and security infrastructure. Rather than rolling their own authentication and security suites, enterprises opt for third-party offerings like Privacera. Privacera provides a unified suite of access control, data governance and security solutions for data warehouses. This type of third-party, all-in-one solution reduces the workload of a company’s IT and engineering teams by removing the need to maintain and constantly update security code. It also leaves security in the hands of the experts, promoting greater reliability.

6. Focus on people

Building a scalable data department in today’s dynamic, post-pandemic workplaces requires leaders to have an innate ability to collaborate with stakeholders from across the enterprise. Data teams are more decentralized, as previously mentioned, with data scientists embedded within teams from other departments such as engineering, product sales or marketing and finance.

Mission | Establish a Center of Excellence with Standing Teams for ML & Data. Somain Business SME; Product Manager/ Data Analyst; Sata Scientist; Data Engineer

Data Standing Teams | Framework

Just as the remit of the data department has expanded, so has the role of its leaders. Previously siloed chief data officers are now chief data and analytics officers (CDAO), forming skilled teams in an environment where data-driven insight is viewed as a key competitive advantage. And for both growing and large businesses practicing new habits, “Small wins are a steady application of a small advantage,” Duhigg wrote. This collaborative mindset often becomes a culture through the entire department, enabling data teams to work more smoothly with other stakeholders.

7. Drive value

The final responsibility modern data leaders carry is the job of driving consistent and tangible value for an organization. In bringing modern analysis techniques to company teams, data executives have an opportunity to influence genuine change. Evidence shows that by being responsive to customer behavior companies can improve satisfaction, loyalty and drive new revenues.

It’s often the case that one successful project will breed curiosity in other parts of the business. Where successful proofs of concept are adopted by other departments, overall efficiencies are amplified. Data-driven optimization initiatives that help companies transition from insight to action have led to real gains. Driving this kind of optimization helps establish the importance of the rise of data leaders and CDAOs as an engine room for company credibility and growth.

New data habits enable successful digital transformation

The emergence of every CxO as a data executive has cemented the importance of data to a company’s digital transformation. As analysis methodologies become more sophisticated, and the use of data continues to evolve, the ability of leaders to form new habits will be crucial. Storytelling using data and visuals can change a company’s trajectory from being a follower to becoming a leader — as well as a disruptor — in an industry segment.

Note: Insight is an investor in Rudderstack, Kubit, Databricks and Privacera

Selling to the CIO: Practical Tips from Insight Partners’ CIO-in-Residence Daniele Tonella

Founders & CEOs: Here’s Why You Need to Get Enterprise Architecture Right

People Matter: Prioritizing Human Capital Along with Investment Capital

The 1-Day Strategy Sprint: Framework, Process, and Tools