Andy Toung on Scaling Strategic Planning

How Andy Led Gusto

Gusto is 11 years old, has grown to 3,000 employees, and serves a customer base roughly equal in size to its employee count. Andy joined Gusto when it was a 27-person company doing about $1M in ARR and serving 3,000 customers. Over 11 years, he helped scale the business into a category leader in HR and payroll, growing from a scrappy startup into a platform serving hundreds of thousands of businesses nationwide. Gusto helps companies manage payroll, benefits, and people operations through a single, integrated software platform. Today, Andy has taken on a new role as CFO at Lovable.

1. The AI Era

• Here is the Lovable slide deck Andy used during the session

The Physics of AI Business are Different

• Andy believes that annual planning in the age of AI makes long-term plans obsolete almost as soon as they’re written. Traditional annual planning cycles no longer work, even at large companies. The velocity of AI means conditions change faster than any static plan can capture.
  
  • Multi-year plans (three- or five-year horizons) are not as relevant in the AI era. Even five weeks out may look radically different.
  • Instead of projecting years ahead, founders should optimize for adaptability and treat plans as flexible instruments for short-cycle learning.
    
    • Not all plans serve the same purpose: Boards need a “plan of record” to evaluate progress and confidence while internal managers and execs need dynamic plans that guide weekly and monthly execution.
    • The sophistication comes from layering plans appropriately and keeping multiple synchronized but distinct views of the business. Gusto framed managers as “people empowerers” whose goal was to align employees rather than keep them adhered to a specific plan. 
• Finance teams need to evolve from plan creators to real-time operators, integrating live metrics and cost signals into decision-making. Finance and BizOps teams must operate in tighter feedback loops with leadership to adapt continuously.
• At Gusto, Andy observed both extremes — highly structured planning that wasted resources and chaotic planning that lacked accountability. Both miss the mark if not continuously adjusted.
• In an AI-native company, planning needs to be living and iterative, with “half-lives” shorter than the plan itself. Plans should be revisited constantly as assumptions expire faster.
• Revenue models have shifted from predictable, seat-based SaaS models to usage-based, nonlinear revenue in AI products.
  
  • Growth is unpredictable: it’s difficult to discern what’s durable trend vs. temporary “tourist” demand.
  • AI-native leaders must learn to manage through this volatility without overfitting to short-term spikes.
• Cost structures have inverted. In SaaS, people were the largest cost center (60–80% of total costs).
• In AI, compute becomes the dominant variable cost, while headcount requirements are leaner and more specialized.
  
  • Andy cautions that frameworks from the SaaS era can be actively misleading when applied to AI companies.

Evolving Profitability and Margin Structures in the AI Era

• Traditional LTV:CAC and nine-year customer lifetime assumptions no longer hold. Founders should focus on fast CAC payback instead.
• Gross margins for AI companies differ dramatically from SaaS, as compute intensity lowers margins but may still offer high operating leverage.
  
  • Andy notes Lovable has been profitable for two months, but also adds that profitability isn’t the goal early on. Early-stage profitability can indicate under-investment in growth.
• The new finance challenge is defining healthy profitability when the physics of cost and growth are changing in real time.
• Growth and margin volatility are defining features of this stage of the market.
  
  • Growth remains unpredictable, while margins can fluctuate widely month to month due to changes in compute costs, usage variability, and shifting demand.
• Reframing margin expectations for AI-native businesses:
  
  • Traditional SaaS benchmarks (80–90% gross margins) may not apply.
  • Many AI infrastructure and platform companies could stabilize at lower gross margins (e.g., 50–60%) but still achieve 25–30% EBITDA margins through operational leverage and scale.
  • The key question is whether the composition of spend (e.g., heavier compute vs. sales and marketing) matters if the overall cash flow profile converges to similar profitability.
• Historical precedent — Twilio:
  
  • Twilio maintained ~55% gross margins but projected 25–30% EBITDA margins over time.
  • Despite sticky customers and strong NDR, investors discounted the stock because it took too long to reach profitability and margins remained unpredictable.
  • It’s one thing to sell the promise of margin expansion. It’s another to operationalize efficiency and actually deliver it.
• Investor perspective is that predictability matters more than mix:
  
  • Public markets will tolerate thinner gross margins if earnings are stable and defensible.
  • The most critical factor is moat and predictability, not necessarily the margin composition itself.
    
    • Example: Infosys vs. Cognizant: Infosys allowed margins to fluctuate between 23–26%, while Cognizant held steady at 20% and reinvested excess into growth.
    • Result: Cognizant consistently earned a premium multiple for maintaining predictable, reinvestment-driven margins.

Set a North Star 

• Andy shared that at Lovable, traditional forecasting is nearly impossible given AI market volatility. At Lovable, clarity of direction replaces precision of planning.
• His framework for alignment starts with four grounding questions:
  
  • Who are we building for and selling to?
    
    • Lovable targets the “99% who don’t code”: the vast pool of creative, non-technical professionals who want to create software rather than consume it.
  • What are we building for them?
    
    • Tools that make software creation as intuitive as building a spreadsheet or narrating an idea, lowering the friction from concept to creation.
  • How are we selling and serving them?
    
    • Even for product-led growth, service and community responsiveness matter.
    • Meet users wherever they are: parents experimenting with code, professionals building weekend projects, or analysts augmenting workflows.
  • Why will we win?
    
    • Lovable aims to be a consumer-grade product that converts individual passion into workplace adoption through a B2C to B2B flywheel.
• For Andy, this clarity acts as the company’s compass. Even without accurate forecasts, everyone can “look up at the same North Star” and move in the right direction.
  
  • As teams and product surfaces multiply, multiple North Stars may emerge, but having coherence over the mission is still necessary. For example, as you grow to 1000 people, you may have 3 of these North Stars within your org.

Building an AI-Native Team

• Companies are beginning to employ a new class of AI-native talent. These are individuals whose first exposure to work coincided with tools like ChatGPT.
  
  • Andy’s most AI-native team member is a high school dropout who began working with AI tools at 15.
• The generational distinction between talent that is and isn’t AI-native is that AI-native team members operate intuitively within probabilistic systems and automation-first workflows. “Immigrants” to AI (experienced operators from SaaS-era companies) must consciously re-train their mental models and learn to embrace more agile learning.
  
  • For finance and ops leaders, the challenge is integrating these two work cultures to blend the speed and intuition of the AI-natives with the structure and rigor of experienced operators.

 Designing Scalable Economics

• Andy encourages founders to reverse-engineer their business model from outcomes back to controllable inputs. He calls this fractal thinking.
• His “pyramid” connects four levels of the business:
  
  • Outcome Metrics: what investors see (revenue, customer count, margin).
  • Unit Economics: how those outcomes are generated (CAC, revenue per customer, gross margin).
  • Productivity Metrics: the efficiency ratios that drive those unit economics (e.g., customers per rep, feature launches per PM).
  • Activity Metrics: the atomic actions and inputs that power productivity (calls made, queries run, compute cycles, user actions).
• Andy believes that true leverage exists at the bottom of the pyramid. Founders often over-index on the top metrics, but sustainable advantage comes from understanding and optimizing the micro-drivers.
• Classical SaaS levers like “customers per rep” no longer map cleanly to AI companies:
  
  • Reps may be partially replaced by agents; compute, automation, and model efficiency become new productivity metrics.
  • Similarly, old ratios (like PM:Engineer) blur as EPD roles merge in AI-native teams.
• Founders should constantly deconstruct their metrics to discover where leverage now lives.
  
  • In the AI era, the drivers of growth and margin shift quickly. This exercise is less about modeling accuracy and more about understanding the machinery of value creation.

Example: How Gusto Identifies Productivity Drivers

• Andy illustrated how understanding the factor tree behind a single metric (like customers per rep) helps identify the real levers that move the business.
• When he joined Gusto, the company had ~3,000 customers and relatively few reps. This was an unusually high ratio.
• The goal was 1,000 customers per rep, while peers averaged 150–200 per rep. This signaled an order-of-magnitude efficiency gap.
• The only way to achieve that was through tight product and process loops:
  
  • Make the product self-serve.
  • Build automation to reduce rep touchpoints.
  • Close the feedback loop between customer issues and product improvements.
• This metric becomes a factor tree of sub-metrics — each project or OKR tweaks one input:
  
  • Deflection rate: How many support cases can be prevented?
  • Time per case: How quickly can each be resolved?
  • Specialization: Are reps matched to issue types (e.g., tax vs. insurance)?
  • Tooling: Can internal systems reduce friction or automate workflows?
• Their key belief was that the true levers of business are hidden within these micro-metrics.
  
  • It’s easy to say “improve CAC or productivity by 10%,” but moving those numbers requires mapping out and directly changing the inputs that drive them.
• Andy believes founders should continuously decompose high-level metrics until they reach controllable, observable drivers.

Model the Future Business You’re Building

• Andy advised founders to build a forward model of their business:
  
  • Assume every key metric (salary, compute, efficiency) works as intended.
  • Push those assumptions through a factor tree to see what your scaled business could look like in 12–24 months.
• This helps define what “good” looks like even if reality deviates later. The modeling exercise isn’t meant to predict perfectly but it helps reveal your system’s constraints and levers.
• This takes about a week to sketch out but becomes a roadmap for how the business should behave if everything works.

Scaling Teams in AI

• Andy posed open questions about how team composition will evolve as AI absorbs human tasks:
  
  • What will be the ratio of customer reps to customers when 99% of cases are handled by AI agents?
  • Will we still need 5-person finance teams in 5,000-person orgs? Or will companies themselves stay smaller?
  • Which functions will merge (e.g., Engineering–Product–Design) as tools collapse role boundaries?
• He believes no one has the answer yet and everyone is living the transition in real time.
• He finds hope in AI-native talent (like the high school dropout on his team) who approach problems without inherited assumptions. These fresh perspectives will likely define how org design evolves under AI leverage.

Velocity and Volume Leverage

• Andy drew from his Gusto experience to explain how to conceptually structure headcount around three buckets. Gusto still uses this “Volume–Velocity–Leverage” framework for staffing and planning, nearly a decade later:
  
  • Velocity: Throughput per unit time: the flow of activity (e.g., sales per month, tickets closed per week).
  • Volume: Accumulated output over time (e.g., active customers, total revenue).
  • Leverage: Work not directly tied to throughput, but to building systems that scale both velocity and volume (e.g., tooling, automation, infra).
    
    • Most headcount growth happens in leverage teams that build internal scaffolding for others to move faster.
• This framework helps clarify why certain roles scale linearly while others scale sublinearly and keeps hiring grounded in business mechanics, not intuition.
• The relationships between these layers will shift as AI matures, but the conceptual model still holds: each company must decide how much of its org invests in throughput versus leverage.

Planning in the AI Era Is About Reflexes

• Andy distinguished between alternatives (choices you make) and scenarios (circumstances that happen to you). The bridge between them, as Annie Duke calls it, is luck.
• Planning is less about predicting outcomes and more about training to respond and developing reflexes for when scenarios change.
• In the SaaS era, annual plans created false precision. They acted as static roadmaps that ignored uncertainty.
  
  • Companies locked into “base, mid, high” cases and executed blindly even when conditions changed.
• In the AI era, plans are no longer valuable as predictions but invaluable as preparation.
  
  • Thinking through multiple futures helps teams recognize and respond faster when one materializes.
  • It’s the same logic as athletic training: repetition builds muscle memory so decisions become reflexive under pressure.
• The exercise of forecasting forces the team to clarify principles, rehearse contingencies, and move with speed when new scenarios emerge.
  
  • This also allows teams to move fast by giving them direction. Teams that move quickly while maintaining alignment on the North Star will outperform those chasing perfect forecasts.
  • He urged founders to periodically pause and recalibrate: moving fast only helps if everyone is pointed toward the same North Star.
  • Precision and predictability, he noted, are temporarily “out the window” but alignment is non-negotiable.
• Optimizing for adaptability over accuracy allows teams to adjust faster. Forecasts will always be wrong, but if your speed of adjustment is faster than your competitors, you can turn velocity into your team’s advantage

Each Team’s Output Is Another Team’s Input

• Andy described a principle often forgotten in startups: every team’s output becomes another team’s input.
  
  • When strategy is unclear (“sell to everyone, everywhere”), it creates chaos downstream in sales, marketing, and operations spin without focus.
  • A real strategy involves sequencing and selectivity in knowing who you serve first, and in what order.
• He noted that founders often fall into “founder syndrome”: trying to be everything, everywhere, for everyone.
  
  • This dilutes focus, confuses teams, and erodes execution clarity.
  • Sequencing markets and ICPs is the only way to avoid strategic entropy.
• Andy explained that misalignment between product and GTM teams creates cascading dysfunction:
  
  • Product says, “We can’t give you a roadmap; things change too fast.”
  • Sales says, “I need to know when things launch and who they’re for.”
  • Marketing says, “I need the ICP defined to plan campaigns.”
• The missing link is translation between product and go-to-market:
  
  • Roadmaps inform sales readiness and regional strategy.
    
    • Example: launching a non-GDPR-compliant feature means Europe is off-limits: that affects hiring, marketing, and resourcing.
  • Every technical choice becomes an operational constraint somewhere else in the chain. One team’s output influences someone else’s roadmap
    
    • Example: if you plan to hire 10 people in Boston instead of SF, finance and ops need to know for budgeting and office space

Activity-Based Planning vs Budget-Based Planning

• As companies grow, budget planning alone stops working. Founders must shift toward activity-based planning to truly understand growth mechanics.
  
  • Most teams at this stage still plan in spend terms (“We’ll burn $20M and double or triple revenue”)
• The real question is: what specific activities will generate that revenue?
• Activity-based planning forces clarity on where growth will come from:
  
  • Existing vs. new customers: What share of next year’s revenue comes from each?
  • If new customers: Have you mapped the pipeline, conversion funnel, and capacity to support it?
  • If outbound-heavy: How many more reps or marketing programs are needed, and when?
• This transition typically hits founders around $20M in revenue (when the company can no longer grow by instinct or momentum alone).
  
  • If you don’t start planning for it nine months in advance, you won’t hit your targets because the operational components simply won’t exist in time.

Scaling Requires Operational Lead Time

• Founders often underestimate the lag between planning and execution. Hiring, onboarding, tooling, and pipeline development take months to ramp.
• You can’t set an aggressive revenue goal in January and start building the team in February, it’s already too late.
  
  • A company aiming to grow from $40M to $200M ARR must plan not just for revenue, but for the organizational infrastructure to support it.
  • That could mean 5× headcount growth in a single year, which is a major operational and cultural transformation.
• Activity-based planning exposes feasibility gaps early and forces teams to plan around real inputs, not aspirational outputs.

Framework: Andy’s Framework for Scaling Headcount

• Andy warned that hypergrowth comes with diminishing enjoyment and rising chaos. As you scale, there are certain sizes that naturally lead to “people breakpoints”. 
  
  • His general assessment begins with the assumption that you need 1 manager to manage every layer of ~8 people
  • The 20–80 person range (roughly under Dunbar’s number, the threshold of how many relationships one can maintain) is the most manageable and fun stage for most founders.
  • Beyond that, complexity compounds faster than output: more people, more problems.
• He referenced Reid Hoffman’s Blitzscaling as a framework for understanding management breakpoints:
  
  • Village (~<150 people): Communication is direct; decisions stay intuitive.
  • Big Village / Small City (~150–500): Managers of managers appear; coordination replaces intuition.
  • City (>500): Systems, process, and policy start to dominate.
  • Nation (>10,000): Scale turns management into bureaucracy
• Andy shared that at Lovable, he modeled headcount milestones to anticipate when each breakpoint would hit:
  
  • The company would reach Dunbar’s number (~150) by Q1 next year and hit manager-of-managers scale (~500) by mid-2027 if efficient.
• He suggests mapping these breakpoints in advance so you can hire, structure, and operationalize before you fall off the next cliff.

Framework: How Andy Uses OODA Loops as an Operating Rhythm

• At Lovable, Andy is reintroducing the classic OODA Loop (Observe, Orient, Decide, Act) as the new backbone of agile planning.
  
  • Originally coined by fighter pilot John Boyd, the OODA loop helps teams move faster and stay aligned.
• Observe:
  
  • Prioritize high-velocity data (what Andy calls “spidey sense”)
  • Spot anomalies and patterns before they show up in dashboards. This is especially important given how fast things are now moving.
• Orient:
  
  • Compare performance to plan (PvP) and budget-to-actual (BvA) data.
  • Reassess whether the direction still makes sense; don’t cling to outdated plans.
• Decide:
  
  • Choose whether to continue, pivot, or replan. Recognize that forecasts describe where you’re going; plans describe where you wanted to go.
  • For example: If you were driving to LA and end up heading toward Vegas, decide if you actually want to go to Vegas.
• Act:
  
  • Implement, learn, and iterate.
  • In fast-changing environments, speed of iteration matters more than correctness.
• OODA is typically modeled on a 30 day cadence, given that’s often how financial close works. Andy described the cadence he’s experimenting with at Lovable:
  
  • Weekly: Metrics meetings: quick feedback loops on what’s working.
  • Biweekly: Headcount and resourcing reviews: to avoid employees running around with ad-hoc hiring.
  • Monthly: Financial closes and cohort analysis. These are still valuable for visibility, but a month is a tenth of your life in AI time. Andy believes monthly is too slow relative to how quickly decisions need to be made for AI companies
  • Quarterly: Scenario refreshes: update models based on new data, market shifts, or performance.
    
    • Example: if pacing 2× faster than plan, increase hiring. If the opposite, cut burn rate quickly.
  • Andy believes probabilistic thinking is also important when you do these reviews to keep yourself from adhering too strictly to an old plan
    
    • For example, Lovable’s controller was worrying that they didn’t have enough hands on deck to reach a certain milestone, and wasn’t hiring given the assumption that they’d be automating most of the work of that function eventually. Andy told her to hire in the interim because the job still needed to get done. These kinds of decisions are made on a weekly cadence – you can’t wait a month to discuss them. 
    • Similarly, if you find yourself outperforming plan, you should adjust to reflect the new needs and priorities of your org. If that means hiring 2x the number of salespeople as you originally planned for, that’s okay. 

2. Finance Team

The Finance Team’s Role

• Andy believes a modern finance team doesn’t just report, it operates.
• The job has three pillars:
  
  • Define the business model: Build clarity on how the company makes and spends money.
  • Defend the business model: Control costs, but more importantly, defend focus.
    
    • Ensure teams spend energy on the right priorities, not just on reducing expenses.
  • Accelerate growth:
    
    • Unblock execution.
    • Build systems, tooling, and financial frameworks that allow teams to move faster. In cases where a new function doesn’t yet exist, they should be able to go in and help bring it to life.
• He called this philosophy “fly high and dive deep”: the ability to zoom out to strategy and zoom in to execution with equal fluency.

Hiring and Profile Fit

• For early-stage companies scaling to or beyond Dunbar’s number, finance must evolve from accounting to acceleration.
  
  • Controllers are useful, but founders should prioritize forward-looking operators who can help make numbers happen, not just report them
  • There’s no specific Dunbar’s number after which you need a head of finance. It could be at the 300-700 people mark: you can feel this out as you know your team best. 
• When building finance functions, Andy looks for operators with banking, consulting, or investing backgrounds, but who’ve moved beyond the “broker mindset.” Andy wants finance hires to feel as though they have skin in the game.
  
  • In those fields, “you are the product.” Inside startups, the product is the business.
  • The best people combine technical fluency, systems thinking, and empathy.
• Key traits he screens for:
  
  • Technically fluent: Can reason about data, tools, and algorithms.
  • Systems thinker: Designs processes that scale beyond themselves.
  • People person: Understands that running finance is as much about influence as analysis
• Andy also sees the CFO job as being a service provider. The ability to offer a service to the company and leadership team is another important trait to assess in a future Head of Finance. 

3. Dealing with Investors

Managing Forecast Expectations

• Andy emphasized that while founders should be truthful and transparent with their boards, it’s important to understand the different expectations of various stakeholders:
  
  • Boards want realism and accurate visibility into how the business is performing
  • Public-market or Wall-Street–minded investors expect a familiar “beat and raise” rhythm of setting a conservative forecast, beating it slightly, and lifting the guidance each quarter to create a sense of predictable growth.
• You may want to have a version of your plan with cushion baked in for investor expectations. But founders shouldn’t start playing “forecasting games” too early. At the growth stage, the market rewards predictability, but precision at the expense of truth can backfire.

Setting Expectations with Growth Stage Investors

• As companies move into Series C and beyond, investor behavior changes: they start to track your plans long after the round closes.
  
  • Even investors who don’t invest in your round will save what you shared with them and revisit it 12–18 months later to assess how well you’ve executed.
  • If you missed your stated plan, even slightly, it will shape their perception of your credibility in future raises. 
    
    • Missing by 1% on extreme growth isn’t material, but what matters is that you establish a consistent beat-and-raise rhythm over time.
    • That predictability creates confidence and compresses perceived risk.
• Never share your stretch plan externally. Only share base-case projections that you have a high probability of meeting or exceeding.
  
  • Keep ambitious targets internal, where hitting 70% of goal may already be strong performance.
• Growth-stage investors expect consistency over aggressiveness.
  
  • They will compare your actuals to prior models
  • This discipline doesn’t matter much at Seed or Series A, but from Series C onward it becomes one of the main ways you’ll be diligenced.

Aligning Plans Across Partners and Stakeholders

• Founders often forget that planning audiences extend beyond investors: partners, vendors, and product collaborators depend on your projections to plan their own resourcing and launches.
  
  • Misaligned planning across companies (e.g., GTM partnerships or integrations) often derails progress more than internal misses.
• Andy noted that coordinating across organizations is exponentially harder than within one, especially past ~500 people or with multi-party partnerships.
  
  • He suggests treating partner planning as part of your operating cadence, not an afterthought.

4. Valuation

Founders Must Control Their Valuation

• Founders must own the valuation conversation. Too many companies allow investors to set prices based on hype or scarcity, rather than fundamentals.
  
  • In the last 18 months, growth rounds have decoupled from diligence and valuations are often set by momentum, not intrinsic value.
  • While this feels flattering (“everyone’s a billion-dollar company”), it creates long-term pain: harder fundraising, talent turnover, and retention issues when the music stops.
• The YC data is sobering:
  
  • 39 of the top 40 YC companies — including DoorDash, Instacart, Gusto, and Coinbase — struggled to raise one growth round.
  • The problem almost always occurred between $2B–$5B valuations, when expectations got ahead of performance.
  • Only Stripe avoided this trap because John Collison deliberately set every valuation himself, even when higher offers came in.
• Why the “overpriced” round becomes the trap:
  
  • Early rounds (A/B) are ownership-driven: investors target 10% ownership and underwrite 5–10× returns.
  • But at $1B–$2B valuations, the next round must imagine a $10B+ outcome, which only a handful of companies achieve.
  • Once that growth investor says, “I can’t see the 2× from here,” they’ll quietly pass — citing “market conditions” or other excuses.
    
    • Founders then face a frozen cap table and a difficult re-entry point for new capital.
• Valuation discipline is a strategic advantage.
  
  • Founders should proactively reduce valuations if the next-year plan doesn’t justify it.
    
    • Example: a Batch 4 company had 17 term sheets at a $2.5B valuation but chose to price at $1.5B after modeling forward revenue.
      
      • Current ARR: $40M.
      • Next year forecast: $120M–130M.
      • At $1.5B, the company would “grow into” the valuation comfortably.
      • At $2.5B, they risked being stuck for 2+ years.
      • The founder’s decision to opt for a realistic multiple preserved flexibility and long-term credibility.
• Use public comps as your anchor, not private hype.
  
  • Long-term value is constrained by public market comparables, which don’t reward inflated multiples.
  • Example benchmarks:
    
    • Marketplaces: ~$1B in gross profit needed for a $10B valuation.
    • B2B software (70–80% GM): ~$1B in revenue required for a $10B valuation.
    • Median public multiple: 7–9× gross profit, regardless of cycle.
  • Private investors may offer premiums, but public markets set the ceiling and eventually, every company faces that gravity.
• This matters beyond fundraising. Overvaluation impacts employee morale, retention, and recruiting.
  
  • Example: Dropbox was valued at $10B pre-IPO (Series C/D), went public at $8B, and never recovered.
  • Employees who joined during the inflated round saw their equity underwater for years — forcing the company to add cash comp and reissue options just to retain talent.
• Founders who calibrate valuation risk early avoid painful resets later.
  
  • Andy shared Gusto’s approach: even in 2021’s froth, they priced rounds below market multiples, targeting durability over optics.
  • If publics trade at 40–50×, they priced at 20×, ensuring Gusto could grow into each mark instead of chasing reversion.
  • When the market corrected, they avoided down-rounds and retained top talent, unlike peers who pushed valuations to the max.

5. Compensation

• Here is the avra founder compensation doc with benchmarks

Navigating CEO Equity Refreshes 

• Most founders are fully vested within 4–5 years, yet remain deeply involved in scaling the business.
• After vesting, founders often delay asking for new equity, either from hesitation or fear of pushback from the board.
  
  • The board’s typical reaction may be: “You already own 15%+, why do you need more?” but refreshes are standard practice and essential for long-term alignment.
• Step 1: You must make the ask.
  
  • Boards rarely offer refresh grants proactively. They’re disincentivized because it creates dilution.
  • Every founder who asks, gets one; those who don’t, don’t.
  • Benchmarking and preparation are critical. Go in with data and precedent, not a soft request.
• Step 2: Understand what type of grant fits your stage.
  
  • Series B/C stage ($10–30M revenue):
  • Typically eligible for a time-based grant only.
    
    • Standard size: 1–2% of fully diluted shares.
  • Harder to secure large grants before breakout growth or profitability.
    
    • Later-stage ($80–100M+ revenue, e.g., Lovable-level growth):
  • Can request a large, performance-based grant (usually 5–6% total across founders).
    
    • Benchmark: if a board would give 5–6% to a replacement CEO, that’s the same justification you can use.
• Step 3: Structure the grant around performance, not time.
  
  • 80% performance-based, 20% time-based is the typical structure.
  • Tranches vest when valuation or liquidity milestones are met (e.g., $3B, $6B, $9B valuation).
  • Usually double-triggered:
    
    • Company must achieve the milestone and become liquid (IPO or acquisition).
    • This approach aligns founders and investors. Investors are happy to grant 6% if it implies a 10×+ company.
• Step 4: Handle co-founder splits thoughtfully.
  
  • The 6% grant covers all founders combined, not per founder.
  • Boards typically leave allocation decisions to the CEO.
  • If split equally, be ready to justify that decision — boards often probe whether roles and contributions merit parity.
• Step 5: Go in prepared with comps and frameworks.
  
  • Bring data on peer CEO grants (from avra cohorts or alumni).
  • Don’t rely on the board to figure it out: 70–80% of boards lack compensation expertise.
  • If you’re unprepared, the board will likely:
    
    • Hire an outside firm like Compensia,
    • Spend 6–9 months benchmarking,
    • Return with an underwhelming offer (e.g., 0.2–1%),
  • Doing the homework yourself can compress the process to <2 months and make it collaborative
• Your CFO will often be the one negotiating with the board on your behalf 

Leveraging Your CFO

• Founder equity refresh negotiations are often run through the CFO rather than directly by the CEO.
• Founders should collect market comps (from peers, alumni, and recent deals) themselves. CFOs typically can’t access these informally.
  
  • Once gathered, share the data with your CFO, who can present it objectively to the board and compensation committee.
  • This approach helps depersonalize the negotiation and maintain professionalism while still ensuring your position is well-supported.
• Handle internal visibility carefully:
  
  • Founder comp decisions are not private. Even in private companies, equity refreshes become visible through:
    
    • Cap tables,
    • Form 409A filings, and
    • Secondary disclosure documents.
  • Once a company goes public, all compensation details are fully disclosed.
• Because of this transparency, it’s crucial that the grant appears reasonable and defensible to employees and investors alike.
• Be prepared to explain why it’s fair and market-aligned, especially since equity scales exponentially — small percentage differences can represent life-changing dollar outcomes.
• Limit awareness to essential participants — typically the CEO, CFO, GC, and board comp committee.
  
  • The fewer people involved in discussions, the lower the risk of misinterpretation or internal resentment.
  • Maintain discretion until the grant is finalized and approved.
• The best CEOs use founder refresh negotiations as an opportunity to reward and retain other critical leaders.
• For long-tenured or high-performing C-level execs (3–4 years+), offer smaller performance-based refreshes (e.g., 0.5–1%), often tied to the same company milestones as the founder grant.
  
  • Boards tend to be highly supportive of this, viewing it as a signal of strong leadership and team stewardship.
• This reinforces alignment across the leadership team and reduces optics of founder-only upside.

Secondary Sales

• Secondary liquidity is common for later-stage founders particularly around Series D/E or once the company approaches $100M+ in revenue.
  
  • A small secondary to achieve financial security or remove personal distraction is generally accepted.
  • However, large secondaries (e.g., $50–100M) can create deep morale and trust issues internally, especially when employees are illiquid and stuck at a high valuation.
• These moments quickly surface on internal forums and can trigger retention and cultural fallout.
• Setting the right tone as a founder:
  
  • Great founders understand that personal liquidity sends a signal to the company.
  • Examples of restraint and long-term alignment:
    
    • Ali Ghodsi (Databricks): has never sold a single share, even as Databricks facilitated employee tenders reinforcing his conviction and credibility.
    • Tony Xu (DoorDash): similarly held his equity throughout, signaling belief in long-term value creation.
  • This approach fosters trust and loyalty among employees who see founders “in it for the long game.”
• The balance founders must strike:
  
  • There’s a healthy middle ground between financial anxiety and excessive cash-out.
  • Founders shouldn’t have to worry about rent or basic comfort. A modest secondary that provides mental security allows full focus and long-term commitment.
  • But once liquidity becomes extravagant or habitual, it raises existential questions about focus and intent.
• Board and investor optics:
  
  • Investors are typically supportive of limited, structured secondaries that help founders stay all-in.
  • The “reasonableness test” applies: Does this amount improve founder focus, or reduce it? Does it align incentives with the long-term mission?
• Secondary programs that also include senior employees (even at smaller scales) help maintain fairness and signal shared upside.

Compensating Employees

• High net revenue retention (NRR) makes planning difficult since most growth comes from customer expansion, not new deals.
  
  • Annual revenue targets can over-reward past work since expansion often occurs 12–24 months after the initial sale.
• Anchor comp to controllable actions:
  
  • Pay sales teams for landing, not passive expansion.
  • Tie commission to booked or first-year realized revenue, not multi-year projections.
  • Limit exposure to long-tail deal growth
• Introduce role specialization:
  
  • Add Account Managers (AMs) or Customer Success teams to own post-sale expansion.
  • Reps focus on acquisition, AMs focus on retention and expansion.
  • Use tapered residuals (smaller trailing commissions) for long-ramp usage accounts.
• Benchmark and model for your motion:
  
  • Compare structures across infrastructure and enterprise SaaS peers.
• Common payout methods:
  
  • Pay at booking for committed revenue.
  • Use quarterly true-ups for realized usage.
  • Apply clawbacks or adjustments for non-deploying accounts.
• Incentives drive behavior:
  
  • Overweighting expansion leads to pipeline stagnation. Reps chase easy growth instead of new logos.
    
    • Example: A company with 160% NRR underperformed because no one was comped on net-new ARR.
  • Rebalance comp plans to 70/30 or 80/20 (new vs. expansion).
• Keep windows short:
  
  • Comp plans should have ≤2-year payout windows.
  • Anything longer becomes administratively unwieldy and behaviorally detached from current goals.

‍

‍