PLG Doesn’t Monetize at Scale, Sales Does
The myth that product-led growth eliminates the need for enterprise sales has cost companies tens of millions in unrealized revenue. Apollo crossed $150M ARR primarily through self-serve before layering in a dedicated sales organization. That trajectory sounds impressive until CRO Adam Carr shares the hard truth: waiting that long was the exception, not the blueprint. Most companies that delay building sales infrastructure past $30M-$50M ARR leave massive enterprise expansion on the table.
The fundamental misunderstanding centers on what PLG actually delivers. Product-led growth excels at two specific functions: generating high-quality demand signals and driving efficient user acquisition. A single developer testing Apollo’s contact enrichment API provides more buying intent data than a hundred cold emails. When that developer invites three colleagues, expands usage across multiple workflows, and starts hitting rate limits, the signal clarity becomes extraordinary. This is where PLG creates asymmetric advantage.
But signal quality and monetization velocity are different problems requiring different solutions. Self-serve motions convert individual users and small teams effectively. The friction is low, the buying cycle is short, and product value is immediately apparent. Enterprise buyers operate in a completely different context. A Fortune 500 company evaluating Apollo for 5,000 sales reps across eight regions isn’t going to self-serve their way through procurement, security reviews, data residency requirements, and executive sign-off. The deal structure alone, multi-year commitments, volume discounts, custom terms, requires human negotiation.
Companies that scaled PLG to $100M+ before building sales teams universally report the same pattern: they left 2-3x revenue growth on the table during those years. The self-serve motion maxed out at team-level purchases while enterprise buyers either cobbled together insufficient solutions or went to competitors with mature sales organizations. Miro experienced this exact dynamic before Adam joined to build their global sales function. The product had extraordinary adoption and love from design teams, but enterprise procurement couldn’t navigate a purchase path for company-wide deployment.
The correction isn’t abandoning PLG, it’s understanding that product-led growth creates the demand foundation while sales provides the enterprise conversion layer. At Apollo, this means treating every self-serve signup as potential enterprise pipeline. When product usage signals indicate department-level adoption or multiple power users within a single company, the sales team has permission to engage. Not to replace the self-serve experience, but to accelerate the path to company-wide value realization.
Data from Apollo’s own GTM transformation shows this layered approach drives 3-4x higher average contract values compared to pure self-serve, while maintaining the efficiency metrics that made PLG attractive initially. The self-serve motion still converts small teams at high velocity and low CAC. Sales focuses exclusively on accounts showing enterprise buying signals, multiple active users, cross-departmental usage, or direct outreach from economic buyers. This signal-based routing prevents the most common failure mode: sales reps spamming every free trial signup and destroying the product-led experience.
The One-Team Model: Preventing PLG and Sales Cannibalization
The internal warfare between growth teams and sales organizations has killed more revenue potential than any external competitor. The pattern plays out predictably: PLG teams view sales as parasitic, taking credit for deals that would have closed through self-serve anyway. Sales teams see growth as naive, celebrating vanity metrics while leaving enterprise revenue uncaptured. Both perspectives contain truth, which makes the conflict particularly destructive.
Apollo’s solution, what Adam calls the “one-team model”, eliminates the structural conditions that create this tension. The framework starts with unified data infrastructure. Growth and sales teams operate from identical customer journey maps, usage analytics, and account scoring models. When both teams see the same signals simultaneously, there’s no information asymmetry to exploit. A spike in API usage from a Fortune 500 domain doesn’t belong to growth or sales, it belongs to the revenue team, and the question becomes which motion best serves that customer’s buying context.
Account ownership rules provide the second structural element. Rather than arbitrary divisions like “sales owns companies over 1,000 employees” or “growth owns all inbound,” Apollo routes based on behavioral signals. A 10-person startup using Apollo across their entire revenue team and bumping into product limits gets sales attention. A 5,000-person enterprise with three individual users testing the free tier stays in the PLG motion until product signals indicate broader evaluation. The routing logic is transparent, automated, and tied to customer behavior rather than team politics.
Compensation alignment provides the third critical component. This is where most companies fail, they create separate quota structures that incentivize growth and sales to fight over the same accounts. Apollo’s approach ties both teams to shared revenue metrics with specific carve-outs that acknowledge different contribution models. Sales reps get full credit for enterprise deals they source and close, but they also get expansion credit when self-serve accounts they’ve never touched grow into their territory. Growth teams get credit for all self-serve revenue plus a percentage of enterprise deals that originated from PLG signals. The exact percentages matter less than the principle: both teams win when total revenue grows, regardless of which motion drove the initial acquisition.
The operational cadence reinforces this one-team mentality. Weekly revenue meetings include both growth and sales leaders reviewing the same pipeline data, discussing the same account strategies, and making joint decisions about resource allocation. When a high-potential PLG account shows enterprise buying signals, the discussion isn’t about which team “deserves” it, it’s about optimal conversion strategy. Should sales engage now or wait for additional product milestones? Does this account need technical implementation support or procurement navigation? The conversation focuses on customer success and revenue maximization rather than internal scorekeeping.
Companies implementing this model report 40-60% reduction in internal friction and 25-35% improvement in enterprise conversion rates from PLG-sourced pipeline. The efficiency gains come from eliminating duplicate work, growth teams no longer build parallel account research that sales already completed, sales teams stop cold outreaching accounts already engaged in product evaluation. More importantly, customers experience coherent buying journeys instead of conflicting messages from different parts of the same company.
Talent Density Over Headcount Velocity: The Hiring Discipline That Compounds
Adam missed hiring targets by 40% during his first six months at Apollo. Deliberately. The pressure to triple the sales team, adding nearly 100 reps in a single year, came from reasonable growth projections. More pipeline coverage, more customer conversations, more closed deals. The math made sense on a spreadsheet. But the implementation would have been catastrophic.
The existing hiring process lacked the rigor required for the talent density Apollo needed at their stage. Interview guides asked generic questions that any mediocre candidate could rehearse answers for. Evaluation criteria were subjective and inconsistent across interviewers. The definition of “good” varied wildly depending on who was in the room. Rushing to fill 100 seats with this broken process would have embedded performance debt that would take 18-24 months to unwind.
The concept of performance debt doesn’t get discussed enough in sales leadership circles. Bad hires don’t just underperform, they actively slow down high performers through cultural drag, create management burden that prevents coaching time with top talent, and generate customer experiences that damage brand reputation. A sales team of 50 reps where 35 are excellent and 15 are wrong-fit underperforms a team of 35 excellent reps, even though the larger team has more theoretical capacity. The wrong 15 consume manager time, create customer issues that the top 35 have to fix, and establish behavioral norms that pull down overall execution quality.
Adam’s decision to slow hiring focused on building the evaluation infrastructure first. This meant documenting must-have characteristics for the specific stage Apollo was at, not generic “good seller” traits, but the precise capabilities required to convert PLG signals into enterprise revenue over the next 12-18 months. The team identified natural curiosity as non-negotiable. Sellers needed to genuinely want to understand why customers behaved certain ways, what obstacles prevented adoption, how different personas used the product. This curiosity couldn’t be faked in interviews because it showed up in how candidates asked questions, not just answered them.
Coachability emerged as the second critical filter. Apollo’s GTM motion was evolving rapidly, new product releases, new competitive dynamics, new enterprise objections. Sellers who needed perfect playbooks before executing would constantly lag behind market reality. The interview process tested coachability by presenting unfamiliar scenarios and watching how candidates incorporated feedback in real-time. Did they get defensive when their approach was challenged? Did they adapt their thinking or dig into their original position? The best candidates treated interview feedback as coaching and visibly adjusted their responses.
Ownership and accountability formed the third pillar. Apollo needed sellers who operated like entrepreneurs within the structure of an established company, people who would identify problems, propose solutions, and drive execution without waiting for perfect direction. The interview process evaluated this through past behavior questions focused on ambiguous situations. How did candidates respond when they saw broken processes? Did they escalate and wait, or did they fix things directly? The strongest hires had track records of seeing organizational gaps and filling them proactively.
Team-first execution provided the final filter. Lone wolf sellers who hoarded knowledge and competed internally were explicitly screened out. Apollo’s model required constant information sharing, which product signals indicated real buying intent, which enterprise objections needed new positioning, which implementation patterns drove fastest time-to-value. Sellers who viewed this transparency as giving away competitive advantage wouldn’t thrive. The interview process included scenarios where helping a peer cost the candidate personal advantage, watching whether they optimized for team success or individual glory.
Building this hiring discipline took four months. Documenting must-haves, creating scenario-based interview guides, training interviewers on consistent evaluation, establishing calibration sessions where hiring decisions were reviewed collectively. The process felt painfully slow while Apollo’s competitors were scaling headcount aggressively. But the results validated the approach, the 60 reps Apollo hired in the back half of the year ramped 45% faster than the previous cohort, achieved quota attainment rates 30% higher, and showed 12-month retention above 90%. The opportunity cost of bad hires would have been far more expensive than the temporary hiring slowdown.
Hiring Architects: Why Modern Sellers Need Systems Thinking
The enterprise seller profile that worked for the past decade is obsolete. Relationship-focused reps who could navigate executive politics and negotiate contracts still have value, but they’re insufficient for the complexity of modern GTM motions. Companies running sophisticated product-led growth engines, multi-channel demand generation, and data-driven customer journeys need sellers who understand how the entire system operates, not just their individual role within it.
Adam calls this the “architect” profile, sellers who think in systems and workflows rather than just relationships and deals. An architect-minded rep at Apollo doesn’t just see a product-qualified lead in their queue. They understand how that PQL was generated: which product behaviors triggered the scoring model, what enrichment data confirmed enterprise fit, why this account was routed to sales instead of staying in self-serve. This systems understanding changes how they approach the conversation. Instead of generic discovery, they reference specific product usage patterns the buyer has already demonstrated, connecting self-serve experience to enterprise use cases.
This architectural thinking extends across the entire revenue process. Traditional sellers viewed pipeline creation as someone else’s problem, marketing generates leads, SDRs book meetings, AEs close deals. Architects understand pipeline creation mechanics intimately. They know which outbound sequences generate response rates worth the effort investment, how intent data gets incorporated into targeting models, what content offers drive genuine engagement versus vanity downloads. This knowledge makes them better collaborators with demand gen teams and more effective at creating their own pipeline when necessary.
The same pattern applies to deal execution. Architect sellers don’t just follow a sales methodology, they understand why specific discovery questions matter, how the information gathered connects to later business case development, which buying committee members need what information at which stages. They see the customer journey as an interconnected system where early-stage decisions impact late-stage conversion probability. This systems view prevents common execution failures like skipping technical validation because the economic buyer seems ready to buy, only to have the deal stall when IT raises concerns during security review.
Apollo’s hiring process specifically tests for this systems thinking capability. Interview scenarios present candidates with GTM workflow challenges that require understanding how different functions connect. For example: “We’re seeing strong product adoption in mid-market accounts but low conversion to paid enterprise plans. Walk through how you’d diagnose this.” Mediocre candidates jump to tactical fixes, better sales outreach, improved pricing packages. Architect-minded candidates map the entire conversion system: What product signals indicate enterprise evaluation versus casual usage? How does the self-serve experience prepare buyers for enterprise pricing? What happens in the handoff from PLG to sales? Where might friction be entering the workflow?
The business impact of hiring architects versus traditional sellers shows up across multiple metrics. Architect sellers ramp 35-40% faster because they understand how their role fits into broader GTM systems, they’re not just learning a pitch, they’re understanding how pipeline gets created, how deals progress, how implementations succeed. Their win rates run 20-25% higher because they orchestrate complex buying processes more effectively, anticipating downstream obstacles and addressing them proactively. Most significantly, their deals have 45% better net retention because they sell in a way that sets up long-term customer success, not just contract signatures.
This doesn’t mean relationship skills and executive presence don’t matter, they absolutely do. But those capabilities are table stakes now. The differentiator is systems thinking layered on top of traditional sales excellence. Companies still hiring purely on charisma, communication skills, and past quota attainment are building sales teams optimized for a GTM motion that no longer exists.
Celebrating Value Realized, Not Contracts Signed
Most sales organizations celebrate closed-won. The gong sounds, the team cheers, the rep gets congratulated. Then the account moves to customer success, and sales shifts attention to the next deal. This handoff model made sense when contract signature aligned closely with value delivery, the customer bought software, implemented it, and started getting value in a relatively linear path. That assumption has collapsed in modern B2B, especially for product-led companies.
Apollo’s GTM motion creates a fundamental misalignment between contract timing and value realization. Many enterprise deals close before the customer has fully deployed the product across their organization. Sales signs a 500-seat agreement, but only 50 users are active at implementation. The contract is closed-won, but 90% of the purchased value sits unrealized. If those additional 450 seats never activate, the renewal conversation 12 months later becomes a negotiation about right-sizing the contract downward. The initial “win” was actually the first step toward churn.
Adam’s team shifted the celebration model to focus on customer milestones rather than contract signatures. Sales still gets compensated for closed deals, but additional variable compensation is tied to usage metrics, specifically, credit consumption in Apollo’s case. When a customer activates more seats, uses more product features, and consumes more of their purchased credits, sales compensation increases. When usage lags, even if the contract is large, compensation reflects that gap between purchased and realized value.
This compensation structure changes seller behavior in profound ways. Reps can’t just focus on getting the largest possible contract signed and moving on. They’re incentivized to ensure successful implementation, rapid user adoption, and deep feature engagement. This doesn’t mean sales becomes responsible for customer success activities, the organizational structure still separates sales and post-sales functions. But it means sales has skin in the game for outcomes beyond signature, which changes how they sell.
Practically, this shows up in deal construction. Architect sellers at Apollo don’t push for maximum seat counts in initial contracts if usage patterns suggest the customer isn’t ready for that scale. A 2,000-person company might have budget for 500 seats, but if only one department is currently using the product, the seller recommends starting with 100 seats and planning for expansion as additional teams onboard. This feels counterintuitive, leaving money on the table in the initial deal. But the data shows these right-sized initial contracts have 60% higher net retention because they set realistic deployment expectations and create expansion opportunities driven by genuine value realization rather than contract obligations.
The cultural shift required to implement this model is substantial. Sales leaders have to stop celebrating big contract values and start celebrating activation milestones. Team meetings review usage data alongside pipeline data. Rep performance discussions include implementation success rates, not just closed deals. The top performer isn’t necessarily the rep with the largest bookings, it’s the rep whose customers activate fastest and expand most reliably.
Customer conversations change as well. Traditional enterprise sales often involved selling a future-state vision, here’s what your organization could accomplish if you deploy this solution comprehensively. That vision selling still has a place, but it’s balanced with realistic deployment planning. Sellers walk prospects through the actual implementation journey: which teams will onboard first, what usage patterns indicate readiness for expansion, how long typical deployments take to reach full value realization. This transparency builds trust and sets up partnerships rather than vendor relationships.
The financial impact of this shift is measurable. Apollo’s enterprise deals structured around realistic deployment timelines and milestone-based expansion show 40% higher gross retention and 80% higher net retention compared to large upfront contracts. The initial deal sizes are smaller, average contract value in year one is about 30% lower. But by year two, the milestone-based deals surpass the big upfront contracts in total value, and by year three, they’re 2-3x larger because they’re built on actual usage growth rather than hoped-for deployment.
Replacing CSMs With GTM Engineers: The Post-Sales Reinvention
Customer success as a function is broken in most B2B companies. The traditional CSM model, assign an account manager to a book of customers, have them run quarterly business reviews and check in periodically, generates activity without impact. CSMs become glorified account administrators, answering support questions that should be handled by documentation and running meetings that customers attend out of obligation rather than genuine need. Apollo is systematically dismantling this model and replacing it with something fundamentally different.
The new structure eliminates traditional CSM roles in favor of what Adam calls “GTM Engineers”, technical operators who don’t own account relationships but instead run intervention-based playbooks triggered by product signals. This is a profound structural change, not just a title swap. GTM Engineers aren’t assigned to accounts. They’re assigned to customer journey stages and specific intervention types. When product data indicates a customer has hit day-7 activation milestones, a specific playbook fires. When usage drops below expected thresholds at day 28, a different intervention triggers. When expansion signals appear, multiple users hitting product limits or exploring enterprise features, another playbook activates.
The intervention model is entirely signal-driven. GTM Engineers don’t decide when to engage customers based on calendar schedules or gut feel. The customer journey map defines specific behavioral milestones that indicate need for human intervention. Some interventions are proactive, a customer expanding usage rapidly might need help with implementation planning before they hit obstacles. Other interventions are reactive, declining usage triggers diagnostic conversations to identify and remove blockers. But all interventions are triggered by data, not arbitrary touch schedules.
This model requires technical depth that traditional CSMs often lack. GTM Engineers need to understand product architecture, API functionality, integration patterns, and data workflows. When they engage a customer, it’s not for relationship maintenance, it’s to solve specific technical or strategic problems preventing value realization. A customer struggling with API rate limits needs someone who can diagnose whether the issue is inefficient code, inappropriate use case, or genuine need for higher tier access. A customer with low adoption needs someone who can analyze usage patterns, identify which features would drive more value, and implement technical solutions to reduce friction.
The organizational implications are significant. Traditional CS teams scale linearly, more customers require more CSMs. Apollo’s GTM Engineering model scales more efficiently because interventions are triggered by need rather than relationship maintenance. A GTM Engineer can handle 3-4x the account volume of a traditional CSM because they’re not running regular check-ins with every account. They’re working exclusively on accounts where product signals indicate intervention will drive measurable impact. This doesn’t mean customers get less attention, it means they get the right attention at the right time rather than generic attention on a schedule.
The hiring profile for GTM Engineers differs completely from traditional CSM hiring. Apollo looks for technical backgrounds, former solutions engineers, implementation consultants, or technical account managers. People who can read API documentation, understand database queries, and troubleshoot integration issues. Relationship skills still matter, but they’re secondary to technical problem-solving capability. A GTM Engineer who can diagnose and fix a customer’s data sync issue in 30 minutes creates more value than a relationship-focused CSM who schedules a meeting to discuss the problem.
Early results from this model show 50% reduction in churn risk and 35% increase in expansion revenue compared to the traditional CSM model. The efficiency gains are even more striking, Apollo’s GTM Engineering team handles 200% more accounts per person than their previous CS structure while driving better outcomes. Customers report higher satisfaction because they get expert help when they actually need it rather than obligatory check-ins when everything is fine.
Product Signals as GTM Operating System: Turning Usage Data Into Revenue Actions
Most companies collect mountains of product usage data and use almost none of it for GTM decisions. Marketing might reference aggregate adoption metrics in content. Sales might check if a prospect is an active user before calling. Customer success might review usage before quarterly business reviews. But product data doesn’t drive systematic, automated GTM workflows. This is the single largest missed opportunity in modern B2B revenue operations.
Apollo treats product signals as the core operating system for all GTM motions. Every customer interaction, whether acquisition, expansion, or retention, is triggered by specific usage behaviors mapped to the customer journey. This requires infrastructure investment that most companies haven’t made: data pipelines that move product analytics into GTM systems in near-real-time, scoring models that translate usage patterns into intervention priorities, workflow automation that routes signals to appropriate teams.
The customer journey map provides the foundation. Apollo tracks specific milestones at day 7, day 14, and day 28 of the customer lifecycle. These aren’t arbitrary timeframes, they’re based on historical analysis of usage patterns that correlate with long-term retention and expansion. Day 7 milestones focus on initial activation: Has the customer completed core setup steps? Have they achieved their first successful workflow? Have multiple users logged in? Customers who hit all day-7 milestones have 4x higher retention than those who don’t, which makes this the highest-leverage intervention point.
Day 14 milestones shift to depth of engagement. Is the customer using multiple product features or just one? Are they integrating Apollo with other tools in their stack? Are usage patterns consistent or sporadic? These signals indicate whether the product is becoming embedded in customer workflows or remains a nice-to-have tool that could be abandoned. Customers showing shallow engagement at day 14 trigger intervention playbooks focused on use case expansion and integration support.
Day 28 milestones identify expansion signals and churn risks. Customers hitting product limits, adding new users, or exploring enterprise features get routed to sales for expansion conversations. Customers with declining usage, incomplete workflows, or single-user dependency get flagged for retention interventions. The signal clarity at day 28 is remarkably high, Apollo can predict with 80%+ accuracy which accounts will expand, which will churn, and which will maintain steady-state based on usage patterns at this milestone.
The routing logic connects these signals to specific GTM actions. Not all signals trigger human intervention, many are handled through automated playbooks. A customer exploring enterprise features might receive automated content about advanced use cases and implementation best practices before a sales rep ever engages. A customer with declining usage might get targeted in-product prompts and email campaigns before a GTM Engineer intervenes directly. Human attention is reserved for high-value situations where automation isn’t sufficient.
The signal quality dramatically improves GTM efficiency. Traditional sales teams work primarily from demographic and firmographic data, company size, industry, technology stack. These inputs provide directional guidance but miss behavioral intent. Apollo’s reps work from behavioral signals that indicate exactly where customers are in their journey and what interventions drive value. A rep calling a customer who just hit product limits in their API usage can open with “I noticed you’re processing 5M enrichment requests per month now, let’s talk about how enterprise pricing gives you better economics at that scale” rather than generic “I’m checking in to see how things are going.”
Building this signal-driven GTM system requires cross-functional collaboration that many companies struggle with. Product teams need to instrument usage tracking with GTM requirements in mind, not just product analytics. Data engineering teams need to build pipelines that move product data into sales and marketing systems with appropriate frequency. RevOps teams need to design scoring models and routing logic that translate signals into actions. Sales and marketing teams need to build playbooks that execute on signals consistently. Most companies have these functions working in silos, which is why product signals remain underutilized despite obvious value.
When to Layer Sales Into PLG: The $100M Mistake
Apollo waited until $100M+ ARR to build a dedicated sales organization. This timing is frequently cited as validation for pure PLG strategies, but Adam is explicit: this was too late. The company left tens of millions in enterprise revenue uncaptured during the years when product-led growth was the only motion. The right time to start layering in sales is far earlier than most PLG companies believe.
The signal to start building sales infrastructure isn’t a specific ARR threshold, it’s when product usage data shows enterprise buying patterns that self-serve can’t capture. This typically happens between $10M-$30M ARR for most PLG companies. At this stage, usage analytics reveal a pattern: multiple users from the same company are signing up individually, usage is expanding across different teams or departments, power users are hitting product limits that indicate enterprise-scale needs. These signals mean enterprise buyers are trying to evaluate the product, but the self-serve motion isn’t designed to support their buying process.
The mistake most companies make is waiting for self-serve to plateau before investing in sales. By that point, the market has been trained that the company doesn’t have enterprise sales support, so large buyers have either found workarounds or chosen competitors. Building enterprise credibility from scratch is far harder than growing it systematically from early enterprise signals. Companies that layer in sales at $10M-$30M ARR grow into their enterprise motion gradually, building processes and reputation while the PLG engine continues driving efficient small-team acquisition.
The initial sales investment doesn’t require massive teams. The first enterprise hires should be senior AEs who can operate with high autonomy, people who’ve closed 7-figure deals before and can build their own playbooks. These reps focus exclusively on accounts showing the strongest enterprise signals: multiple active users, cross-departmental usage, or direct outreach from economic buyers. The goal isn’t to build a scalable sales machine immediately, it’s to learn what enterprise buying patterns look like for the specific product and market, then codify those learnings into repeatable processes.
The timing sequence typically follows this pattern: At $10M-$20M ARR, hire 2-3 senior enterprise AEs to start engaging high-signal accounts. At $20M-$40M ARR, add sales leadership and build the first structured playbooks based on learnings from early enterprise deals. At $40M-$60M ARR, scale the team systematically while maintaining the talent density discipline described earlier. At $60M-$100M ARR, add specialized roles, SEs, sales development, vertical-focused teams. This gradual layering prevents the chaos of trying to build an enterprise sales organization from scratch at scale.
The financial analysis supports earlier investment. Companies that wait until $100M+ ARR to build sales typically see 18-24 month lag before enterprise revenue becomes material, time needed to hire teams, build processes, and establish market credibility. Companies that start layering sales at $20M-$30M ARR have functioning enterprise motions by the time they hit $60M-$80M, which accelerates growth through the $100M milestone rather than creating a revenue plateau that triggers panic hiring.
The common objection to early sales investment is that it destroys unit economics. PLG motions have remarkably low CAC, and adding expensive enterprise sales reps increases overall cost structure. This math is correct but incomplete. The relevant comparison isn’t PLG CAC versus enterprise sales CAC, it’s total revenue growth with and without enterprise motion. Companies that maintain pure PLG typically see growth rates decline as they exhaust the addressable market of teams willing to self-serve. Companies that layer in sales maintain growth rates by opening up enterprise segments that were never going to self-serve regardless of product quality.
Box and Dropbox provide the cautionary tale. Both built remarkable PLG engines with tens of millions of users. Dropbox waited longer to build enterprise sales infrastructure and grew more slowly as a result. Box invested earlier in enterprise sales and became the clear leader in enterprise file sharing despite having fewer total users. The difference wasn’t product quality, it was GTM strategy and timing. Enterprise buyers needed sales support that Dropbox wasn’t providing when it mattered most for market positioning.
Preventing Sales Noise: The PQL Routing System That Protects Self-Serve
The fastest way to destroy a PLG motion is letting sales reps spam every free trial signup. This failure mode is so common it’s almost predictable: company builds successful self-serve product, decides to add sales team, gives sales access to all signup data, sales starts cold calling everyone, product-qualified leads get terrible experience, conversion rates drop, company debates whether sales was a mistake. The problem isn’t sales, it’s routing logic that doesn’t distinguish between self-serve buyers and enterprise prospects.
Apollo’s PQL (product-qualified lead) routing system protects the self-serve experience while ensuring enterprise buyers get sales attention. The system operates on multi-factor scoring that goes far beyond simple firmographic filters. Company size matters, but it’s insufficient, a 5,000-person enterprise with one individual user testing the free product shouldn’t get sales outreach. A 500-person company with 15 active users across three departments absolutely should. The routing logic evaluates usage intensity, user count, cross-departmental adoption, feature engagement, and buying signals like exploring enterprise documentation or requesting security information.
The scoring model assigns points across multiple dimensions. Usage intensity tracks daily active users, feature breadth, and workflow completion rates. A customer who logs in daily and uses 8-10 different features scores higher than one who logs in weekly and uses only basic functionality. User count within a single company domain provides another signal, multiple users indicate team-level evaluation rather than individual exploration. Cross-departmental adoption (detected through user roles and team structures) suggests enterprise buying process rather than single-team purchase.
Behavioral intent signals provide the highest-value routing data. Customers who view enterprise pricing pages, download security documentation, or engage with implementation guides are explicitly researching enterprise purchase. These behaviors trigger immediate sales routing regardless of other factors. Customers who remain in basic free tier functionality and never explore enterprise features stay in the self-serve motion even if they work for large companies. The routing logic follows actual buying behavior, not assumed buying behavior based on company size.
The threshold for sales routing is deliberately high. Apollo would rather have sales ignore a potential enterprise account temporarily than have sales spam self-serve buyers prematurely. The routing system includes time-based logic, a company showing moderate enterprise signals gets flagged for monitoring, but sales doesn’t engage until signals strengthen or persist for 14-21 days. This patience prevents the common mistake of sales reaching out the moment someone from a large company signs up, before that person has experienced enough product value to have an informed enterprise conversation.
Sales reps don’t have access to the full signup database. They see only accounts that have cleared the PQL routing threshold, and they see those accounts with full context, usage data, feature engagement, user count, specific behaviors that triggered routing. This context transforms the sales conversation. Instead of “I see you signed up for Apollo, would you like to talk about enterprise features?” the conversation becomes “I see your team is running 50K enrichment requests per month across 8 users in sales and marketing, let’s talk about how enterprise pricing and features support that usage pattern.”
The system includes feedback loops that improve routing accuracy over time. When sales engages an account that wasn’t actually in enterprise evaluation, that data point adjusts the scoring model. When an account that wasn’t routed to sales later submits an enterprise inquiry, that miss gets analyzed to understand which signals were overlooked. After 6-12 months of operation, these feedback loops typically improve routing precision by 40-50%, meaning sales spends more time on genuine enterprise opportunities and less time on false positives.
The impact on conversion metrics validates the approach. Apollo’s self-serve conversion rates remained stable even as they scaled the sales team, because self-serve buyers weren’t getting interrupted by sales outreach. Enterprise conversion rates improved 30-40% because sales was engaging accounts at the right moment with relevant context. The total conversion rate across both motions increased, which is the only metric that actually matters, not whether self-serve or sales gets credit, but whether the company is capturing all available revenue.
Building Sales Teams for the Next 12-18 Months, Not Just Today
Hiring for current needs is how companies build organizational debt. The rep who’s perfect for the $30M ARR stage, comfortable with ambiguity, willing to build playbooks from scratch, effective without sophisticated tooling, often struggles at the $100M ARR stage when the role requires executing established processes at scale. Companies that don’t anticipate this evolution end up with teams that can’t grow with the business, forcing painful performance management and organizational restructuring.
Adam’s hiring discipline at both Miro and Apollo centers on evaluating candidates for the next 12-18 months of company evolution, not just the current state. This requires clear thinking about how the business will change and what capabilities will matter in that future state. At Apollo’s current stage, the company is transitioning from early enterprise sales to scaled enterprise motion. The capabilities that matter over the next 18 months include process adherence, coaching receptivity, and ability to operate within defined playbooks while still thinking strategically. These differ from the capabilities needed 18 months ago, when the priority was entrepreneurial problem-solving and comfort with ambiguity.
The interview process explicitly tests for future-state capabilities even if they’re not immediately necessary. Candidates get asked about experiences operating in structured environments, not just startup chaos. How have they adapted to new processes? How do they balance following playbooks with strategic thinking? Have they successfully transitioned from building to scaling? The best candidates have experience with both modes and can articulate how their approach changes based on organizational stage. Candidates who’ve only worked in early-stage environments or only in established corporations often struggle with the transitions that fast-growing companies require.
This forward-looking hiring approach creates some tension with immediate needs. A candidate who’s perfect for the next 18 months might be slightly overqualified for today’s requirements. They might command higher compensation than the current market rate for the role as it exists today. They might have capabilities that won’t be utilized for 6-9 months. Hiring them anyway is the right decision because the alternative, hiring for today and replacing them in 12 months, is far more expensive and disruptive.
The financial math supports this approach despite higher upfront costs. Hiring a rep at $150K base who can grow with the company for 2-3 years is cheaper than hiring a rep at $120K base who maxes out after 12 months and needs to be replaced. The replacement costs include recruiting fees, ramp time, lost productivity, and cultural disruption. More importantly, the better hire contributes to building institutional knowledge and process maturity that compounds over time. The rep who grows with the company for three years becomes a leader who can onboard and coach the next wave of hires.
This philosophy extends beyond individual hires to team composition. Apollo deliberately maintains a mix of experience levels and backgrounds, ensuring the team has people who’ve operated at the next stage of scale. When the sales team was 20 people, they hired several reps who’d worked in 100+ person sales organizations. These reps brought perspective on what breaks at scale, which processes matter, what cultural elements preserve or destroy performance. They couldn’t implement all their knowledge immediately, the organization wasn’t ready, but their presence meant the company could anticipate problems rather than just react to them.
The risk of this approach is over-hiring, bringing in people so far ahead of current needs that they become frustrated or disengaged. The mitigation is honest conversations during the interview process about organizational stage and growth trajectory. Candidates need clear understanding that they’re being hired partly for capabilities that won’t be fully utilized immediately. The best candidates find this exciting, they’re joining specifically because they want to help build the next stage. Candidates who need immediate utilization of all their skills are wrong-fit regardless of capability level.
The results of this hiring discipline show up in retention and performance metrics. Apollo’s sales team shows 90%+ 12-month retention, which is extraordinary for a scaling organization. Performance distributions are tighter, fewer reps at the bottom end of performance because the hiring bar filtered them out. Most importantly, the team scales smoothly without the organizational restructuring that typically happens when companies realize their team can’t execute at the next level of sophistication. The team evolves with the business rather than being replaced by it.
