Chris Arsenault

Category: The Musings

Long-form writing on technology, leadership, AI strategy, and execution inside complex organizations.

  • Marketing Ops as Revenue Infrastructure

    Marketing Ops — the function that runs the data, systems, and processes behind marketing — is treated by most organizations as a back-office cost center. The companies that treat it as revenue infrastructure produce dramatically better outcomes from the same marketing investment.

    Understanding why Marketing Ops is actually revenue infrastructure, and what changes when it’s resourced that way, is one of the more impactful insights in modern revenue organizations.

    The Common Treatment

    Marketing Ops, in most organizations, is staffed lightly. Maybe one or two people for a marketing org of 30. Their job is to “make the systems work” — fix broken integrations, clean up data hygiene issues, generate reports for executives.

    Treated this way, Marketing Ops is reactive. They respond to broken things. They don’t build the underlying infrastructure that would prevent the breakage. They don’t invest in the experimentation systems that would let marketing iterate faster. They don’t build the data architecture that would make attribution more honest.

    The outcome: marketing functions on a shaky foundation. Campaigns launch with imperfect targeting. Reports show inconsistent numbers. Experimentation is slow because the infrastructure to run experiments doesn’t exist. The CMO gets blamed for execution issues that are actually infrastructure issues.

    The Infrastructure Treatment

    The companies that treat Marketing Ops as infrastructure invest in it differently. Specifically:

    1. They staff it as a real function.
    A 30-person marketing org has 5-7 ops people. The ops team is led by a senior leader who sits at the marketing leadership table, not as a back-office function. The investment ratio looks heavy compared to industry average — and produces output that justifies it.

    2. They build experimentation infrastructure.
    The ability to run a structured marketing experiment in two weeks, instead of two months, changes what marketing can learn and how fast. Companies with strong Marketing Ops can run 10x more experiments per quarter than companies with weak Marketing Ops, and the compounding learning advantage is enormous.

    3. They build data architecture as a strategic asset.
    Customer data, campaign data, attribution data — all of it lives in systems that are designed to be queried, not just stored. The marketing team can answer questions in hours that competitors answer in weeks (if at all). The data infrastructure becomes a competitive advantage.

    4. They build automation that compounds.
    The repetitive operational work of marketing — list segmentation, campaign production, lifecycle messaging — gets automated systematically. The team’s time goes to higher-leverage work. The infrastructure produces output that doesn’t require ongoing manual labor.

    The ROI Argument

    The argument for investing in Marketing Ops at infrastructure level is straightforward: every marketing dollar produces more output when the infrastructure is strong. A campaign that targets the right audience produces 3-5x the response of a campaign that targets a poorly-segmented list. An experiment that runs in two weeks produces learning that compounds across subsequent campaigns. Honest attribution lets the team double down on what’s working.

    The math: each additional Marketing Ops hire often produces more revenue impact than the equivalent additional marketing campaign hire. But the impact is indirect — it shows up as better performance across all the campaigns, not as a specific campaign’s contribution. Which is why most CMOs underinvest. The direct attribution to a specific Marketing Ops hire is impossible.

    What to Watch For

    If you’re a CEO evaluating marketing investment, three signals about Marketing Ops capability:

    1. How fast can your team launch a structured experiment?
    If the answer is more than four weeks for anything beyond a basic A/B test, the infrastructure is weak.

    2. How consistent are the numbers across reports?
    If the same metric shows different values in different dashboards, the data architecture isn’t trusted.

    3. How much campaign work requires manual labor?
    If list-building, segment-pulling, or campaign production takes hours of manual work each time, the automation infrastructure is missing.

    Marketing Ops isn’t a cost center. It’s the foundation everything else stands on. Treat it that way and the marketing function performs accordingly.

    The companies that have made this shift are running circles around competitors with the same total marketing budget. The infrastructure is the multiplier.

  • OEM Partnerships That Compound

    OEM partnerships — where one company’s product gets embedded into or distributed through another company’s offering — are one of the highest-leverage partnership structures available, and one of the hardest to get right. The companies that build them well create durable revenue streams. The ones that build them poorly burn engineering cycles for negligible return.

    The pattern that separates the two is structural.

    What an OEM Partnership Actually Is

    In an OEM partnership, your product is being sold by or through another company’s commercial motion. This is different from co-selling (where both companies sell to the same customer separately) and different from referrals (where one company hands prospects to another).

    OEM means the partner’s customer-facing motion includes your product. Sometimes invisibly (your product is embedded in theirs). Sometimes visibly (your product is sold alongside theirs as a recommended companion). Either way, the partner’s go-to-market is doing work for you.

    Why OEM Is Hard

    Three structural challenges that kill most OEM partnerships:

    1. Misaligned incentives.
    The partner’s reps don’t get paid for your product the way they get paid for theirs. So they don’t sell yours unless forced to. The OEM relationship exists in contract but not in motion.

    2. Engineering investment without revenue.
    OEM partnerships often require integration work — APIs, white-label customizations, joint product development. The engineering investment happens upfront. The revenue happens (maybe) in the future. Companies that do the integration without ensuring the commercial motion is structured to produce revenue end up with technical debt and no return.

    3. Customer support ambiguity.
    When your product is embedded in the partner’s offering, who supports the customer when something breaks? If the partner’s support team handles it, they need to know your product well — which requires training they may not invest in. If your team handles it, you’re spending support resources on customers you don’t have a direct relationship with.

    What Makes OEM Partnerships Work

    The OEM partnerships I’ve watched generate real revenue share three structural features:

    1. The partner’s reps are comped on your product.
    Not just contractually — operationally. Their comp plan includes credit for your product. Their quotas include some assumption of your product’s revenue. Without this, the partnership exists in legal but not in execution.

    2. Joint customer success ownership.
    A specific person on each side owns the customer experience. When something breaks, the resolution path is clear. When the customer wants to expand, the expansion conversation has owners on both sides. Ambiguity in customer ownership is one of the most reliable killers of OEM partnerships over 18 months.

    3. Engineered reciprocity.
    The OEM motion isn’t one-directional. Either you’re providing strategic value to the partner that they couldn’t get elsewhere, or there’s a reciprocal flow of value back from your customer base to theirs. Without reciprocity, the partner eventually feels they’re doing more than they’re getting and the motion fades.

    When OEM Makes Sense

    Three conditions where OEM is worth pursuing:

    1. The partner has reach you can’t replicate.
    They sell into a market where your direct motion would take years to build. Their reach is the value.

    2. Your product makes their offering meaningfully more valuable.
    Customers buy your product because it’s part of their bundle, not as a separate decision. The bundling creates value the customer wouldn’t get otherwise.

    3. Both sides have aligned strategic priorities.
    The OEM motion is core to both companies’ growth, not a side bet. When the priorities align, both sides invest in making it work. When they don’t, the partnership atrophies whenever either side has competing priorities.

    When to Walk Away

    If the partner won’t commit to comping their reps on your product, the OEM is going to fail. Don’t pursue it.

    If customer support ownership can’t be cleanly defined, the OEM is going to produce friction that costs more than the revenue. Don’t pursue it.

    If reciprocity isn’t designed in, the partnership has a 12-18 month half-life. Pursue with eyes open and don’t expect it to compound.

    OEM partnerships that compound are rare because the structural requirements are hard. The ones that work are some of the most durable revenue motions available. The ones that don’t are some of the most expensive partnership investments to write off.

    Build for the structural conditions, or don’t build at all.

  • The CEO’s Week in Deals

    How a CEO spends their time on deals — when they engage, when they disengage, how they sequence — has more impact on revenue than most CEOs realize. Most spend their time on deals reactively, jumping in when called, with no overarching architecture.

    The CEOs who spend their deal time deliberately produce more revenue per CEO-hour than the ones who don’t.

    The Common Pattern (Reactive)

    Most CEO deal involvement looks like this:
    – Sales team calls when they need executive sponsorship
    – CEO joins the call, brings whatever value they can in 45 minutes, exits
    – Repeat across many deals, with no consistent prioritization
    – CEO ends the week feeling like they spent a lot of time on deals without knowing if it was the right time

    This pattern has a few problems. The CEO becomes a generic resource that any rep can call on. Their time gets used for deals that don’t actually need executive intervention. The deals that genuinely warrant CEO time may not be the ones that get it.

    The Architected Pattern

    The CEOs I’ve watched run deal time well architect it. Specifically:

    1. They identify the 5-10 deals per quarter that warrant CEO time.
    Not every deal needs the CEO. The deals that do are usually: top-10 strategic accounts, deals that are stuck and need executive escalation, deals where competitive pressure requires CEO-to-CEO engagement, and deals at thresholds that justify executive investment. The CEO knows which deals these are at the start of the quarter, not in the middle when reps escalate.

    2. They engage proactively, not reactively.
    The CEO reaches out to the customer’s executive sponsor before being asked to. Schedules a conversation. Builds the relationship outside any specific deal pressure. When the deal needs executive engagement later, the relationship already exists.

    3. They protect time for deal work.
    The CEO blocks specific time on the calendar for deal conversations — both internal reviews and external customer conversations. The blocks are protected. Other meetings work around them. The deal work happens because it’s been prioritized at the calendar level, not because it competes for time with everything else.

    What a Good Week Looks Like

    A CEO running architected deal time might look like this in a typical week:

    • 90 minutes Monday morning: pipeline review with the head of sales, focused on the top 10 deals that need executive attention
    • 30-45 minutes per day: one customer-executive call or in-person meeting
    • 60 minutes Friday afternoon: review of the week’s customer engagement and adjustment for next week

    That’s roughly 5-6 hours per week on deal work. Less than most CEOs spend, used more deliberately.

    What to Hand Off

    Just as important as what the CEO does on deals is what they don’t. Several activities CEOs often do that they probably shouldn’t:

    Routine deal coaching. This is the VP of Sales’ job. If the CEO is coaching reps on standard deal mechanics, the VP isn’t doing their job, and the CEO is filling a gap they shouldn’t be filling.

    Approval theater on standard discounts. If the discount is within normal range, the CRO or VP of Sales should have authority. Forcing standard discounts to the CEO clogs the CEO’s calendar with low-value decisions.

    Customer conversations any senior person could handle. Not every customer conversation requires the CEO. Many require a senior leader who’s not the CEO. Defaulting to the CEO when a senior leader would suffice trains the customer to expect CEO engagement on every meaningful conversation.

    The Diagnostic

    Audit your last four weeks. What percentage of your deal time was on the top 10 deals versus on deals you didn’t choose? What percentage was proactive (you initiated) versus reactive (someone called you)? What percentage created lasting relationship value versus served a single transactional moment?

    If the answers skew reactive, generic, and transactional, the architecture is missing.

    CEO deal time is one of the most leveraged resources in any company. Spending it deliberately is one of the highest-leverage decisions a CEO makes. Most don’t make it deliberately. Which is why most CEO deal time produces less than it could.

  • The Right Answer to ‘What Do You Do?’

    “What do you do?” is the most common question in any professional setting and the one most professionals answer badly. The answer you give shapes how the asker thinks about you for the duration of the relationship — and most professionals give an answer that produces no useful follow-up.

    The right answer is harder to construct than it seems and more valuable than its simplicity suggests.

    The Wrong Answers

    Most people answer in one of three predictable ways:

    The job title answer.
    “I’m the VP of Sales at Company X.” Accurate. Useless. The asker now knows your title but has no reason to remember it, no way to follow up, and no entry point for genuine conversation.

    The company description answer.
    “I work at Company X. We help mid-market businesses optimize their operations.” This is marketing-speak. The asker tunes out at “help mid-market businesses.” Generic positioning produces generic responses.

    The exhaustive answer.
    “Well, we do a lot of things — we’re a SaaS platform that combines CRM, marketing automation, and customer success in one workflow, and we serve customers in financial services, healthcare, and…” The asker has lost interest by the third clause.

    All three answers fail because they’re optimized for technical accuracy rather than for opening a useful conversation.

    The Right Answer

    A good answer to “what do you do?” has three properties:

    1. It’s specific enough to be memorable.
    Not “I help companies with sales.” Something like “I help founders figure out when to hire their first salesperson.” The specificity gives the asker something to grab onto.

    2. It invites a follow-up question.
    A good answer leaves a door open. The asker should naturally want to ask “how do you do that?” or “tell me more about that.” The answer plants a hook that pulls the conversation forward.

    3. It positions you, not your company.
    What you personally bring matters more than your company description. The asker may not remember your company name. They will remember “the person who helps founders make first-sales-hire decisions.”

    The Construction

    The structure I’ve found works well:

    “I [specific thing I do] for [specific kind of person] who [specific situation they’re in].”

    Examples:
    – “I help BD operators figure out which of their relationships are actually productive and which are just busy work.”
    – “I work with founders in their first $1M of revenue who need to figure out whether they’re ready for their first sales hire.”
    – “I write about relationship-driven revenue for operators who think trust matters more than tactics.”

    Each version is specific. Each invites a follow-up. Each positions the person rather than the company.

    The Test

    After you give your answer, watch what happens. If the asker says “interesting” and changes the subject, your answer didn’t work. If they say “tell me more about that” or “actually, I have a question…” — your answer landed.

    This is iterable. Try variations. Notice what produces engagement. Refine until you have an answer that consistently opens conversations.

    The 30 seconds it takes to deliver a good answer to “what do you do?” is one of the highest-leverage moments in professional networking. Most people use those 30 seconds badly. The ones who use them well build network capital faster than the ones who don’t.

    Construct your answer deliberately. Test it. Refine it. Then use it everywhere — at conferences, in meetings, in introductions. The compound returns over a career are larger than they look from a single conversation.

  • Market Development is an Operating System, Not a Campaign

    Most companies run market development like marketing with a longer sales cycle.

    They pick a segment. Build a list. Write a message. Launch a campaign. Attend an event. Count meetings. Then they do it again next quarter. Rinse and Repeat.

    The work can look busy and still fail to build anything durable.

    When the campaign ends, the target list goes stale, the lessons stay in individual notebooks, and what was once a promising opportunity is handed to a team that was not involved in shaping it. Nobody can explain why one market moved and another did not. The next campaign starts from nearly zero.

    This is episodic outreach, not market development.

    Market development is an operating system. Its job is to continuously convert an uncertain market into a sequence of better decisions, and the distinction matters.

    A campaign has a start date, an audience, a message, and an end date. An operating system decides what the organization notices, how it interprets the signal, who can act, what evidence is required, where work gets routed, and how the result changes the next decision.

    Campaigns can run inside that system and they cannot substitute for it.

    Why the Campaign Model Breaks

    The campaign model usually breaks in five places.

    First, the list becomes the market.

    The team starts with names in a database and assumes those names represent the opportunity. But the real unit of opportunity may be a property portfolio, a development corridor, a municipality, a partner ecosystem, a buying committee, or a population of projects reaching the same decision window.

    Second, activity replaces evidence.

    Emails sent, calls made, event scans, and meetings booked are easy to count. They are weak proof that a market exists, that a problem is urgent, or that anyone has authority to act.

    Third, qualification happens inside the seller’s head.

    One person sees a strategic opportunity. Another sees an unqualified lead. A third sees an operational problem that belongs somewhere else. Without a shared evidence standard, the loudest interpretation wins.

    Fourth, handoff means forwarding an email.

    The receiving team gets context late, disagrees with the assumptions, or lacks capacity. The originating team calls this resistance. The receiving team calls it poor qualification. Both are partly right.

    Fifth, learning is not retained.

    The market answers a question, but the answer never changes the segmentation, message, qualification rule, partner model, or investment thesis. The organization generates experience without building knowledge.

    An operating system fixes these failures by connecting seven components.

    1. Define the Unit of Market

    Start by deciding what you are actually developing.

    It may be an account. It may be a place. It may be a portfolio of properties, a cohort of construction projects, a municipal priority, an ecosystem route, or a recurring decision shared by several industries.

    This sounds semantic, but it isn’t.

    If the unit is wrong, ownership will be wrong; data will be fragmented, outreach will be aimed at one contact while the decision sits across six organizations. The CRM record will look complete while the market picture remains broken.

    Write the unit down and give it a stable identifier. Define what belongs inside it and what does not.

    2. Build a Signal Layer

    Market development should begin with observable change, not generic intent.

    In my current work, a building permit, capital plan, land acquisition, leadership change, public agenda, funding notice, procurement, expansion, renewal, or operating failure can create a decision window. The signal does not prove an opportunity but it does beg for investigation.

    Use official and permissioned sources where possible. The U.S. Census Bureau’s Building Permits Survey, for example, provides national, state, and local statistics on authorized residential construction. The FCC’s National Broadband Map provides location-level availability information reported through the Broadband Data Collection process. No one source tells a complete commercial story; triangulating your data will contribute to responsible market sensing when their scope and limits are understood.

    Every signal needs a source, date, confidence level, owner, and expiration rule: a stale signal should not stay green because nobody validated it.

    3. Map Decision Windows

    A buyer journey describes how someone moves toward a purchase. A decision-window map describes when a choice becomes expensive, slow, or impossible to revisit.

    Those are different things.

    In the built environment, a utility-design release may matter more than a marketing response. In enterprise technology, the security review may matter more than the demo. In a public opportunity, an approved procurement path may matter more than executive enthusiasm.

    Good market developers ask four questions:

    1. What decision is coming?
    2. Who has authority to make it?
    3. What is their why?
    4. When does the window close and what’s impacted by it?

    The how becomes the relevance of your outreach: is the message attached to a real decision, or just a persona description?

    4. Separate Momentum From Evidence

    Commercial movement and evidence maturity are related, but they are not the same. Track the two dimensions separately.

    A senior executive can be enthusiastic while feasibility is unknown and a technical team can validate a use case while no buyer has authority or budget. A signed document can exist while implementation ownership is unresolved and a decision remains unowned.

    For every opportunity, distinguish among:

    • What we know.
    • What we assume.
    • What we still need to prove.
    • Who can accept the proof.
    • What decision the evidence supports.

    This removes a great deal of pipeline theater and it also makes a respectful no easier. A weak opportunity does not need another follow-up email, it needs missing evidence, a later decision window, a different route, or a stop decision.

    5. Route Authority Explicitly

    Market development is orchestration work, but it should not become an empire that absorbs every decision.

    Product owns product truth. Finance owns the financial standard. Legal owns legal interpretation. Security and privacy owners decide within their authority. Delivery teams accept implementation work. Public-sector specialists govern public processes.

    The market-development team owns the quality of the package moving between them. That package should state the problem, affected population, decision date, evidence, assumptions, risks, requested decision, and proposed owner. A handoff is complete only when the receiving owner accepts it, rejects it with a reason, or routes it to the correct destination.

    6. Run a Control-Loop Cadence

    Cadence is not meeting volume, it is the rhythm by which the system senses, decides, acts, and learns.

    A practical rhythm looks like this:

    1. Weekly: Review new signals, aging decisions, evidence gaps, handoffs, and exceptions.
    2. Monthly: Review market theses, portfolio movement, capacity, campaign results, and control issues.
    3. Quarterly: Decide which markets to expand, redesign, hold, or exit.
    4. After every material pursuit: Record what changed the decision and what the system should do differently next time.

    The meetings are not the system. The decisions, records, and changed behavior are the system.

    7. Measure Outcomes, Not Motion

    Activity still matters, obviously it tells you whether the team is doing the work. However, that should not be confused with the value of the work.

    An outcome-led scorecard should examine seven things:

    1. Market evidence quality.
    2. Qualified decision progression.
    3. Delivery or customer outcomes.
    4. Economics.
    5. Repeatability.
    6. Partner and customer trust.
    7. Controls and learning.

    Calls, posts, events, and meetings sit underneath those measures as diagnostics. These explain performance, but they should not manufacture it.

    This is where many teams get uncomfortable. Activity is immediate and attributable, market outcomes are delayed and shared. The operating system has to preserve accountability without pretending one person caused a multi-party result.

    Build the Minimum Viable System

    You do not need a reorganization or a new technology platform to start. In the first 30 days:

    • Pick one market play.
    • Define its unit of market.
    • Name five credible signals.
    • Map the three most consequential decision windows.
    • Create one evidence record.
    • Define two accepted handoffs.
    • Establish a weekly decision review.
    • Choose three outcome measures and three stop conditions.

    Run it for 90 days and resist the urge to scale it because the team is excited. Scale it when the records are usable, the handoffs are accepted, the decisions are faster or clearer, and the learning is changing behavior.

    The point is not process for its own sake, it is organizational memory.

    A good operating system allows a company to recognize a pattern once, test it responsibly, route it correctly, and reuse what it learned. Over time, the advantage compounds. The team does not merely know more people. It knows how the market actually moves.

    Campaigns create moments. Operating systems create memory, judgment, and repeatability.

    Build the system, then make the campaigns earn their place inside it.

  • Late-Stage Discount Discipline

    Late-stage discounting is where most deals lose margin they didn’t have to lose. The pressure to close converges with the buyer’s leverage at exactly the moment when the seller is most vulnerable to giving up value.

    The sellers who hold margin in the late stages do specific things differently from the ones who don’t. The discipline isn’t natural; it has to be built deliberately.

    Why Late-Stage Discounting Happens

    Several pressures converge in the last 20% of a deal:

    The seller has emotional investment.
    After months of cycle, the seller wants the deal to close. The pain of losing it after this much work feels worse than the pain of giving up margin to win it. This emotional asymmetry favors discounting.

    The forecast pressure is acute.
    The deal is forecast for this quarter. Losing it means missing the quarter, which means uncomfortable conversations with management, the board, and possibly the team. The discount feels small relative to the cost of the miss.

    The buyer has new leverage.
    At late stages, the buyer knows the seller wants the close. They negotiate more aggressively because they sense the seller’s vulnerability. Procurement amplifies this — their job is to extract the maximum discount, and the late stage is when they have the most leverage to do so.

    Fatigue.
    After months of negotiation, the seller’s energy for resistance is low. Saying yes feels easier than saying no, especially when the discount sounds reasonable in the moment.

    What Disciplined Sellers Do

    The sellers who hold margin late share a specific set of practices:

    1. They establish pricing logic early.
    Before late-stage pressure hits, the pricing has been explained with specific rationale. Not “this is what it costs” but “the implementation is $X because of Y specific work, and the annual is $Z because the value delivered is W.” Pricing built on logic is harder to discount, because reducing the price requires arguing against the logic.

    2. They tie discounts to specific reciprocal commitments.
    If a discount is given, it’s traded for something. Multi-year commitment. Faster payment terms. Reference rights. Expanded scope. Discounts given without reciprocal commitment train the buyer that asking produces results — and the asks compound on every subsequent deal.

    3. They escalate discount requests instead of accepting them.
    The seller doesn’t have authority to grant the discount on their own. The request has to be escalated, evaluated, and decided by a level above the seller. This both buys time and signals that the discount isn’t a normal accommodation. Buyers ask for fewer discounts when they know the ask will trigger an escalation rather than a quick yes.

    4. They walk away from bad-economics deals.
    The hardest discipline is the willingness to lose the deal rather than take it at terms that won’t be profitable. Sellers who will walk away from a deal at a 40% discount win deals at 25% discounts that the never-walk seller loses entirely. The walk-away credibility is the leverage.

    The Cultural Move

    Holding late-stage margin requires organizational support. If management punishes missed deals more than discounted deals, the discipline collapses. If management punishes both equally, but rewards discount discipline visibly, the behavior holds.

    The CROs who have the strongest margin discipline have built a culture where “we walked away” is celebrated when the alternative was bad-economics. The ones who haven’t have teams that quietly discount their way through every quarter.

    Late-stage discount discipline is built early. Pricing logic established in week three is what holds in week thirty. Without the early foundation, the late-stage pressure wins.

    Build the foundation. Hold the line. The margin you save compounds across every deal you close.

  • Why Your Best Customer Isn’t Your Best Reference

    The customer who’s most loyal, most enthusiastic, and most aligned with your product is often a poor reference. This is counterintuitive, and it costs companies pipeline they don’t realize they’re losing.

    Understanding why your best customer might be your worst reference — and how to choose references better — is one of the more useful unlocks in enterprise sales.

    What Makes Someone a Bad Reference

    A bad reference isn’t necessarily an unhappy customer. It’s a customer whose situation, language, or perspective doesn’t match what the prospect is trying to evaluate.

    Several patterns:

    1. Too small.
    Your best customer is a 50-person startup that uses your product enthusiastically. The prospect is a 5,000-person enterprise. The startup’s experience is irrelevant to the enterprise’s evaluation. Their happy use case doesn’t translate.

    2. Too far ahead.
    Your best customer adopted your product three years ago and has built sophisticated workflows on top of it. The prospect is in early adoption. The reference talks about advanced use cases the prospect can’t yet imagine. The conversation creates anxiety, not confidence.

    3. Too aligned.
    Your best customer agrees with you about everything. They never push back. They love every feature. The prospect, who is naturally skeptical, hears the reference as a fan club rather than as an honest assessment. The prospect leaves the call doubting whether the reference is real.

    4. Wrong industry.
    Your best customer is in a vertical you happen to have product-market fit in. The prospect is in a different vertical with different constraints. The reference’s success doesn’t transfer.

    What Makes Someone a Good Reference

    A good reference shares context with the prospect — same scale, same industry, same stage of adoption, similar challenges — and gives an honest assessment that includes the hard parts.

    Three specific qualities:

    1. They had a specific problem similar to the prospect’s.
    The reference can describe a problem the prospect recognizes. Not generic (“we needed to grow”) — specific (“we had three weeks to roll out a system across 12 sites and our previous vendor’s implementation had failed”). The specificity is what creates resonance.

    2. They acknowledge what was hard.
    The reference talks about implementation friction, decisions they’d make differently, places the product fell short. This builds credibility. A reference who describes the relationship as flawless reads as suspect. A reference who describes it as mostly great with specific warts reads as honest.

    3. They have similar context.
    Same general industry. Same general scale. Similar buying committee structure. The prospect can map the reference’s situation to their own. Without that mapping, the reference is academic.

    How to Build a Reference Program That Works

    Most reference programs are organized around customer enthusiasm. The best customers get listed first. This is the wrong organizing principle.

    Better: organize references around prospect-segment match. For any given target prospect, the right reference is the customer whose situation most resembles the prospect’s, regardless of how enthusiastic that customer is overall.

    This means cultivating references across your customer base — not just the top 10. Some of your most useful references will be customers who are moderately enthusiastic about your product but whose context perfectly matches what specific prospects need to see.

    The Practical Move

    Audit your reference list. For each name on it, ask:
    – Is their context similar to our typical prospect’s context?
    – Are they at a similar scale?
    – Did they have a similar problem?
    – Are they capable of giving an honest assessment that includes hard parts?

    The customers who score highest on those questions are your best references — even if they’re not your most enthusiastic customers overall.

    The customer who loves you most is sometimes the one who can sell you least to a stranger. Pick references for fit, not for fandom.

  • Your Website Is a Sales Rep

    Most B2B websites are designed by marketing teams optimizing for design and brand. They’re judged by visual appeal, message clarity, and SEO performance. They’re rarely judged by the metric that actually matters: how well they would perform if they were a sales rep on commission.

    The reframing helps. A good sales rep, working a website’s worth of inbound interest, would do specific things that most websites don’t.

    What a Good Sales Rep Does

    A good rep handling inbound interest:

    Qualifies fast. Within the first conversation, they know whether the prospect is in the right segment, has budget, has authority, and has timeline. They don’t waste time on unqualified inbound.

    Speaks specifically. They reference the prospect’s industry, role, and apparent situation. Generic positioning isn’t part of their playbook.

    Surfaces the right next step. Not always a demo. Sometimes a discovery call. Sometimes a piece of content. Sometimes an introduction to someone the prospect should also meet. The next step is matched to where the prospect actually is.

    Removes friction. Calendar links. Direct contact info. Easy pathways to the next step. The rep’s job is to make it easy for the qualified prospect to keep moving.

    What Most Websites Do

    Most websites do the opposite of all four:

    They don’t qualify. Every visitor sees the same homepage. The CFO of a Fortune 100 sees the same opening as a graduate student doing research. No segmentation. No filtering.

    They speak generically. “We help companies like yours grow.” Empty calorie language designed to offend nobody and resonate with nobody.

    They offer the same next step to everyone. “Request a demo.” Whether the visitor is ready or wildly not ready, the call to action is the same.

    They add friction. Forms with 12 fields. Email gates on basic content. Live chat that never has a human behind it. The mechanics of converting interest into engagement are clunky.

    What a Sales-Rep Website Does

    If you redesigned your site as a sales rep would, three things would change:

    1. The homepage would have multiple paths.
    Different segments would see different framings. Someone arriving via a security-focused search would see security framing. Someone arriving via an industry-specific keyword would see industry framing. The website would route to the framing that matches the visitor’s apparent context.

    2. The next-step call to action would vary by visitor signal.
    First-time visitor with thin signal? Offer a piece of content that matches their apparent interest. Returning visitor with deep signal? Offer a meeting. Account-named visitor (matched against your target list)? Offer a direct conversation with a senior person.

    3. Friction would be ruthlessly removed at every step.
    Forms would be short. Content would be ungated unless gating was strategic. Calendar links would be easy. The next step would always be one click away.

    The Counterargument

    Marketing teams will object: “We need to capture leads. Form fields are how we qualify. Gated content is how we measure intent.”

    These objections are mostly wrong now. The lead capture optimizer that worked in 2018 doesn’t work the same way in 2026 — buyer behavior shifted. Buyers research extensively before they’re willing to fill out a form. By the time they do, they’ve already decided. The form became a conversion barrier rather than a qualification tool.

    The websites that win in the current environment are the ones that prioritize ease of buyer engagement over breadth of lead capture. Less data per visitor, but the data is from higher-intent visitors who actually convert.

    If your website were a sales rep, would you keep them on the team? If the answer is no, the website needs work — not aesthetically, but functionally.

    The sales-rep frame produces sharper decisions about what to put on the site, how to route visitors, and what to ask of them. Use it.

  • The Referral Partnership That Works

    Most referral partnerships fail because they’re built on hope and goodwill rather than structure and incentive. The few that work share specific structural features that distinguish them from the many that don’t.

    After watching dozens of referral partnerships across multiple industries, here’s what separates the working ones from the dead ones.

    What Makes Most Referral Partnerships Fail

    The typical referral partnership starts with an executive handshake. “We refer business to each other when relevant.” Both sides agree, mean it, and — for about three months — actually do it occasionally.

    Then it fades. Why?

    Nobody owns it operationally. Sales teams have quotas tied to direct revenue. Referring a deal out to a partner is, in most comp plans, neutral to negative for the rep. They don’t get credit for the referral. They might lose attribution they could have claimed by working the deal directly.

    No tracking, no accountability. When neither side tracks who referred what, both sides assume the other isn’t holding up their end. Often both sides are right. Without explicit tracking, the perceived imbalance kills the motivation to refer.

    Asymmetric value flow. One side benefits more than the other. The one benefiting less stops referring. The one benefiting more notices and stops too, because reciprocity has died.

    What Makes Referral Partnerships Work

    The functional ones share three structural choices:

    1. Compensated reps.
    The reps doing the referring get something for it. Spiffs. Quota credit. Recognition. The compensation doesn’t have to be large — it has to be visible. Reps who get $250 for a closed referral refer. Reps who get nothing don’t, regardless of executive enthusiasm.

    2. Tracked attribution.
    Both sides track referrals in a shared system. Not perfectly — but well enough that imbalances are visible. When one side has referred 30 deals and the other has referred 3, the conversation can happen explicitly rather than the relationship dying silently.

    3. Defined trigger conditions.
    “Refer when relevant” is too vague. “Refer customers in segment X with need Y to partner Z” is specific. Specific triggers convert into actions. Vague intentions don’t.

    The Mechanics of Setting It Up

    Three concrete moves to actually build a working referral partnership:

    1. Map the trigger conditions explicitly.
    Sit down with the partner and define, in writing: when should we refer to you? When should you refer to us? Specific customer profiles, specific situations, specific signals. The conversation forces clarity that ad-hoc referrals never produce.

    2. Build the spiff into rep compensation.
    Both sides commit to comping their reps for closed referrals from the partner. Even a modest amount. The comp signal changes behavior in ways executive enthusiasm doesn’t.

    3. Set a quarterly review with explicit metrics.
    Number of referrals each direction. Conversion rate. Revenue attributed. Without the review, the asymmetries that kill partnerships develop invisibly. With the review, they get addressed early.

    When to Walk Away

    Not every referral partnership is worth setting up. The ones not worth it have two characteristics:
    – The customer overlap is theoretically present but operationally weak
    – One side has dramatically more deal flow than the other in the relevant segment

    In both cases, the partnership tends to be lopsided from the start, and no amount of structure compensates. Better to recognize this early and not invest in a partnership that’s structurally unlikely to balance.

    The referral partnerships that work require operational discipline most companies aren’t willing to invest in. The ones that don’t fail because nobody designed them to succeed beyond the executive handshake.

    If you’re going to do referral partnerships, do them with structure. Otherwise, don’t pretend the casual version will produce the value the structured version would have.

  • The Hire You Shouldn’t Have Made

    Every founder has made at least one hire they should have recognized as wrong in the interview process and didn’t. The interview felt fine. References checked out. The person seemed smart, experienced, and culturally fit. Three to six months in, the relationship is broken — the founder is spending disproportionate time managing the hire, the hire isn’t delivering the output expected, and nobody on either side is sure what went wrong.

    The hire you shouldn’t have made usually had signals in the interview that got rationalized away. Learning to see those signals before you make the offer is one of the highest-leverage hiring skills.

    The Signals Most Often Rationalized Away

    “They’re smart but didn’t ask great questions.”
    A candidate who doesn’t ask sharp questions in an interview is signaling one of two things: either they don’t deeply care about the role, or they don’t have the curiosity that the role requires. Either is a problem, and both show up post-hire as passive execution on poorly-defined tasks.

    The rationalization is usually “they were nervous, that’s why they didn’t ask questions.” Sometimes that’s true. Often it isn’t. The candidates who don’t ask questions in the interview are usually the candidates who won’t ask them on the job.

    “They interviewed well but something felt off.”
    If you finished the interview unsure why you felt the hesitation, trust the hesitation. The unexplainable feeling of “something is off” is usually the interviewer’s pattern recognition detecting something below the conscious level. It’s not a reason to reject outright — but it’s a reason to dig harder in the next round, not to paper over it with “well, they checked all the boxes.”

    The rationalization is usually “I’m being too picky” or “we need to fill this role.” Both are signals that the hiring process is being pulled by urgency rather than by standards.

    “References were fine but generic.”
    Fine-but-generic references are often a bad sign. Real strong references describe specific situations, specific outcomes, specific growth the candidate produced. Generic references — “great to work with, hardworking, reliable” — often indicate references who are being polite but can’t point to specific excellence.

    The rationalization is “they got good references.” They got references. The references were lukewarm on specifics. That’s data, and it’s usually ignored.

    “They have the right experience but haven’t done this specific thing before.”
    Experience in adjacent roles is valuable, but the specific thing you’re hiring for often has specific dynamics the candidate won’t understand until they’re in it. If you’re hiring someone to build a motion they haven’t built before, price that correctly — they’ll learn, which takes time, and they may learn poorly.

    The rationalization is “they’ll figure it out.” Sometimes they do. Often they struggle, and the struggle is visible at month four.

    Why These Signals Get Ignored

    Three reasons:

    1. Urgency.
    The role has been open too long. The team needs relief. The expansion plan requires this hire by a specific quarter. The pressure to fill overrides the discipline to wait.

    2. Sunk cost.
    You’ve invested hours in this candidate. Multiple interview rounds. Multiple conversations. Saying no means starting over. The cost of starting over is visible and immediate. The cost of hiring wrong is distributed and delayed. The math, rationally, still favors starting over — but the psychology favors finishing.

    3. Conflict avoidance.
    Raising the concern with the hiring team means arguing against a candidate everyone has discussed positively. That’s socially costly. Staying quiet and hoping is cheaper in the moment.

    The Practice

    Before extending any senior offer, I now do a personal exercise: write down what specifically concerns me about this hire, if anything. Force the concerns to be articulated. Then ask: would I hire this person if I didn’t have this specific role open right now?

    If the answer to the second question is “probably not” — don’t extend the offer. The specific role open isn’t a good enough reason to hire someone you wouldn’t otherwise hire.

    The hire you shouldn’t have made is almost always the hire where you saw the signals, rationalized them, and moved forward because moving forward was easier than not.

    Don’t rationalize. Trust the signals. The cost of not filling is almost always less than the cost of filling wrong.