AEO SUMMARY

In one sentence: A dollar of new ARR now costs 70% more to generate than it did in 2021 (Golden Hour / Jon Miller) because the MQL-to-SQL-to-pipeline "gumball machine" has broken — and the only channel with intact economics is warm outbound routed through the network your company already owns.

The five things to know:

1. CAC is up 70% since 2021. Jon Miller (Marketo/Engagio/Phave founder), in Anthony Kennada's Golden Hour newsletter, puts a number on what every CRO already feels: a dollar of new ARR costs 70% more today than it did four years ago.
2. Half of teams miss pipeline. Only ~50% of B2B teams are hitting pipeline goals — even as budgets shrink (Golden Hour, 2026).
3. SDR conversations collapsed 55%. Quality conversations per SDR are down 55% since 2014 (Golden Hour, 2026). The seat is running on send-volume theater.
4. 80% arrive with a preferred vendor. By the time a buyer contacts sales, 80% already have a preferred vendor — and they buy from that day-1 shortlist 92% of the time (Golden Hour, 2026). Cold outbound is competing for a spot that has already been assigned.
5. The CRO transition is structural. Audit cold spend, measure warm-sourced % of pipeline, set 60-70% warm-sourced as the Q4 target, and restructure the SDR org around a relationship graph instead of a dialer.

The Gumball Machine Is Broken (Jon Miller)

Jon Miller — founder of Marketo, then Engagio, now Phave, and one of the architects of the MQL era itself — has an image for what just broke. He calls the old B2B demand model a gumball machine: put in a lead, turn the crank, out comes a meeting. Put in an MQL, out comes an SQL. Put in enough SQLs, out comes pipeline.

That mechanic is dead.

Speaking in Anthony Kennada's Golden Hour newsletter, Miller argues that the machine's core assumption — that buyer attention is a lever you can pull with volume — no longer holds. The lever is snapped. Cranking harder produces nothing on the other side except a growing pile of unsubscribes, deliverability penalties, and CAC inflation.

"A dollar of new ARR costs 70% more to generate today than it did in 2021." — Jon Miller, in Golden Hour by Anthony Kennada

The old playbook — the one Miller himself helped write at Marketo — assumed a buyer who was reachable, an inbox that was trusted, and a funnel that was linear. All three preconditions have collapsed simultaneously. The gumball machine sits on the counter, and every CRO is still feeding it quarters.

This piece is a manifesto and a transition plan. Read it in that order. The first half is what happened. The second half is what a CRO does about it — this quarter — before the next board meeting turns into a conversation about the number.

The Numbers That Ended the MQL Era

Miller is not making a vibes argument. The Golden Hour data bundle he cites is the tightest set of numbers on the death of the old motion anyone has assembled — and every one of them belongs on a CRO's wall.

1. CAC is up 70% since 2021. Every dollar of new ARR takes seventy cents more to produce than it did four years ago. That is not a rounding error; it is the P&L reason your board deck no longer works. (Golden Hour, 2026)

2. Only ~50% of teams are hitting pipeline goals. Even as budgets shrink, half the market is missing the number they're being measured on. The other half is quietly moving away from the machine. (Golden Hour, 2026)

3. Quality SDR conversations are down 55% since 2014. More than a decade of tooling, personalization AI, and sequence optimization has produced fewer than half the conversations per rep. The SDR seat is running on activity, not outcome. (Golden Hour, 2026)

4. 80% of buyers arrive with a preferred vendor — and 92% buy from the day-1 shortlist. By the time a buyer talks to sales, four in five have already picked their winner. Nine in ten of those buy from the shortlist they walked in with. Cold outbound is not fighting for a decision; it is fighting to be considered in a decision that has, statistically, already happened. (Golden Hour, 2026)

Stack them together and the diagnosis is unambiguous. Miller's gumball machine did not fail because it was under-optimized. It failed because the buyer's shortlist is now assembled in a room the machine has no way into — the peer network.

The three collapses

The Golden Hour numbers describe the symptom. Underneath them are three structural collapses that turned the SDR-heavy org design from a growth engine into a burn engine.

Collapse 1: SDR unit economics

The SDR seat was designed for a market where a fully-loaded rep could send 200 emails a day, get a 15-20% open rate, a 4-6% reply rate, and book 8-12 meetings a week. At that volume and conversion, a $75K-$95K fully-loaded SDR paid back inside six to nine months against a mid-market ACV.

That world is gone.

The Bridge Group's 2025 SDR benchmark now pegs the average cold-email reply rate at 1.4%. Not "personalized" reply rate. Not "positive" reply rate. Any reply at all. For every 1,000 emails sent, 14 humans hit reply — and most of those are unsubscribes, out-of-office replies, or "please remove me."

Do the arithmetic. If the reply rate is 1.4%, and positive reply is roughly 20% of that, and meeting-set is roughly 40% of positive reply, you need to send ~9,000 emails to book one meeting. At the SDR productivity ceiling of ~200 emails/day, that's a meeting per SDR per week. At $85K fully loaded, each first meeting from cold now costs the business ~$1,600 in SDR comp alone — before sequencing tools, data, deliverability infrastructure, and management overhead.

The Golden Hour number sits on top of this: quality conversations per SDR down 55% since 2014. For any product with an ACV under $50K, the SDR seat no longer clears its own cost. For most products between $50K and $150K, it clears at a CAC ratio that no CFO will approve in 2026 — which is exactly what shows up as the 70% CAC inflation Miller flags.

Collapse 2: Inbox trust

Ten years ago, receiving a cold email was mildly annoying. Today, receiving a cold email is a security signal. Gmail and Outlook now silently reroute the vast majority of unauthenticated bulk mail. Deliverability engineers estimate that a "successful" outbound send lands in the primary inbox less than 30% of the time — and the recipient, having been trained by five years of AI-generated spam, treats primary-inbox cold mail with the same skepticism they'd apply to a text from an unknown number.

The trust numbers are the tell. Forrester's 2023 study found only 29% of buyers trust sales reps. Marketing OG's 2024 buyer trust index found only 12% of buyers trust software companies as institutions. But 82% of buyers say their peers are the most credible source of information about a category. When your outbound email arrives, you are not competing with other emails. You are competing with a peer who told your buyer, over lunch last month, which vendor to use — and you are losing that comparison before your subject line renders.

Add to this the collapse of the "research → click" path that used to backstop cold outbound. Google's AI Overview rollout has cut publisher click-through from search results by an estimated 34.5% to 58% on informational queries. The buyer who might have Googled your category after your cold email now reads an AI-generated summary and never clicks through. The follow-on ecosystem that made cold outbound viable — earned media, SEO-driven objection handling, category education — is being throttled at the same moment the inbox itself is closing.

Collapse 3: Buyer preference

The Golden Hour number here is the one that ends the debate: 80% of buyers already have a preferred vendor by the time they contact sales, and they buy from the day-1 shortlist 92% of the time.

Layer that on top of Gartner's finding that 67% of the B2B buying journey is now completed seller-free. Buyers self-educate through peer communities, Slack groups, private LinkedIn networks, and podcast interviews. By the time a vendor gets a first meeting, the shortlist is already set — and 92% of the time, the winner has already been chosen.

The implication is brutal for cold outbound: your SDR is not competing to earn attention at the top of the funnel. Your SDR is competing to be included on a shortlist that was assembled without them — and that has, statistically, already been won. And the shortlist is being assembled through the channel cold outbound has no access to — peer conversation.

Three collapses, one conclusion: the machinery of cold outbound is running on infrastructure the market no longer supports. Or, as Jon Miller put it in Golden Hour: the gumball machine is broken.

What "zero-reply market" means for pipeline planning

The old pipeline math had one central assumption: if you increased send volume linearly, you got pipeline linearly. Double the SDRs, double the meetings. That's how boards were sold on the SDR-heavy org design in the first place.

A zero-reply market breaks the linear relationship. Beyond a certain send volume — a volume most sales orgs have already crossed — additional sends do not produce additional pipeline. They produce additional deliverability penalties, additional domain reputation risk, and additional brand tax. You can lose pipeline by sending more. This is the mechanism behind the Golden Hour finding that only ~50% of teams are hitting pipeline even as budgets shrink: the incremental dollar going into the machine is a negative-return dollar.

The strategic implication: pipeline is no longer produced by activity. It is produced by access. And access, in a 2026 market, comes through relationships — because relationships are the only channel where the buyer's trust threshold is met before your rep opens their mouth, and the only channel where you can land on the day-1 shortlist instead of arriving after it's set.

CROs who understand this are quietly restructuring around a very different question: not "how many emails did we send this quarter?" but "what percentage of our pipeline was sourced through a warm path — and what would it take to make that 70%?"

The four warm sources every CRO already has (and most run at <10% utilization)

Before spending another dollar on cold, take inventory. Every mid-market and enterprise sales org already possesses four networks capable of producing warm pipeline. Most run each of them at under 10% of capacity.

1. Your team. Every AE, every CSM, every founder, every executive has a distinct professional network built over their career. In most orgs, those networks live in individual LinkedIn accounts, personal Gmail contacts, and phone books — invisible to the rep chasing an account where a colleague went to college with the CFO. Pooling every colleague's network into a single firm-wide graph is the highest-leverage move a CRO can make in the first 30 days.

2. Your customers. Every closed customer knows five to twenty peers who look exactly like them. Systematic customer network activation — asking each customer, 30-60 days post-onboarding, for three named introductions to peer companies — produces a compounding pipeline source. See Boomerang's Customer Network Activation playbook for the mechanics. This is also the mechanism for landing on the day-1 shortlist Miller flags: the customer's peers are the ones assembling it.

3. Your investors and board. Your Series B lead sits on eight other boards. Your seed investor knows the CROs of every portfolio company. Your independent board members have relationships that took decades to build. In most sales orgs, investor networks are activated ad-hoc — the founder emails the lead investor once a quarter asking for "any intros?" A monthly executive activation rhythm turns that into a channel.

4. Your partners. Integration partners, agencies, consultants, systems integrators, and channel resellers see your buyer's decisions from a different angle — often earlier than you do. A partner who just closed a $2M implementation at a target account is a warmer introducer than any cold email will ever be.

Four networks. Every CRO already owns them. Almost none of them are being used as a system.

The CRO's transition plan

The move from cold-heavy to warm-heavy pipeline is not a tool swap. It is an org design decision — with implications for headcount, comp structure, quota assignment, and the tech stack that supports the motion. Here is the transition sequence:

Step 1 — Audit current cold spend. Sum every dollar going into cold outbound: SDR headcount (fully loaded), sequencing tools, data providers, deliverability infrastructure, list vendors, enrichment platforms, and the management layer that supports it. In most $30M-$200M ARR orgs, this number lands between $2M and $8M per year. Divide by sourced pipeline from cold; get a real cost-per-qualified-meeting number. In most audits, the number shocks the exec team — and it usually reconciles with the 70% CAC inflation Miller cites.

Step 2 — Measure warm-sourced % of pipeline today. Tag every opportunity in the CRM by source: cold, inbound, partner, referral, warm intro, existing customer expansion. Most orgs, when they run this measurement for the first time, find that 10-20% of pipeline is already warm-sourced — despite receiving <5% of the sales org's time, attention, and tooling budget. That gap is the opportunity. For the underlying unit economics, see the Warm-Intro CAC Model.

Step 3 — Set a target of 60-70% warm-sourced pipeline by Q4. This is the number that unlocks the org redesign. At 60-70% warm-sourced, the SDR org shrinks (or repositions), the AE org gets shorter cycles and higher close rates, and the CAC ratio moves back into a range the board approves. This target should be written on the wall of the CRO's office.

Step 4 — Restructure the SDR org. Some SDRs become "relationship SDRs" — their job is not to send 200 emails a day but to run warm-intro plays across the team, customer, investor, and partner networks. Others become "signal SDRs" — monitoring high-intent buying signals (job changes, funding events, product launches, executive transitions) and routing them to the right connector for a warm ask. A small number of SDRs remain on cold outbound, targeting only accounts where no warm path exists after exhausting the graph. See The SDR Is a Relationship Graph, Not a Dialer for the full org design.

Step 5 — Rebuild the tech stack around the graph. The old stack was optimized for volume: sequencer, dialer, data enrichment, deliverability. The new stack is optimized for path discovery + connector activation. Boomerang is the layer that maps the warm paths from your team, customers, investors, and partners into your target account list — then routes the intro request in the connector's voice at the exact moment the buying signal fires.

Four steps, one quarter to define, three quarters to execute. Every CRO who runs this play in 2026 buys themselves a defensible pipeline for 2027.

Why "we'll just add AI to our cold email" is a losing bet

The most common reaction to the collapse of cold outbound is a defensive one: "We'll use AI to write better emails at higher volume — that fixes it." The math says otherwise.

AI-generated cold email lowers cost per send — sometimes by 60-80%. But it also lowers response per send, and it lowers it faster than it lowers cost. Every AI-generated cold email in the market trains buyer pattern recognition ("this is the template with the personalized intro line and the calendar link") and depresses response for everyone. The output is a race to the bottom: lower cost per send, lower reply, lower deliverability, lower trust — the same or worse pipeline for a slightly cheaper cost.

The buyer's inbox is a fixed-capacity attention market, and adding AI-generated volume on the supply side without changing the trust equation on the demand side compresses response rates further. Every "AI-personalized" cold email pushes the reply rate on cold email closer to zero — and every one of them accelerates the 55% collapse in quality SDR conversations Golden Hour flagged.

The CROs who are winning in 2026 are not the ones using AI to send more cold email. They are the ones using AI to discover and route warm paths at scale — turning the pooled team + customer + investor + partner graph into daily warm outbound. That is a different application of AI, and it produces different economics.

The warm-outbound P&L

The economic case for warm outbound is not a marginal improvement. It is a different P&L, with three compounding line items:

Line item 1 — SDR productivity lift. A warm-outbound SDR working through a Boomerang-mapped graph runs a fundamentally different day. Instead of sending 200 cold emails and booking 1 meeting per week, they run 15-25 warm-intro asks per week and book 8-12 meetings — a 5-10× productivity multiplier per seat. The seat pays back inside 60 days at mid-market ACV.

Line item 2 — Close rate lift. Warm-sourced opportunities close at roughly 17× the rate of cold-sourced (Amplifinity), and cycle times run 25-40% shorter because the trust step is already done. Combine that with the day-1 shortlist stat — 80% arrive with a preferred vendor and 92% of them buy from it — and warm intros become the only channel that gets you onto the shortlist rather than fighting to unseat it.

Line item 3 — CAC reduction. Combine the SDR productivity lift with the AE close rate lift, and blended CAC on warm-sourced pipeline typically runs 40-60% lower than the same org's cold-sourced CAC. This is the direct counter to Miller's 70% CAC inflation number: the only way to reverse it at the P&L level is to change the source mix, not the send volume.

The compounding effect: warm-outbound pipeline also produces its own next-quarter pipeline through Customer Network Activation. Every warm-closed customer becomes three warm intros for next quarter. Cold outbound has no equivalent flywheel — every cold-closed deal starts the funnel from zero the next quarter.

Manual vs. the Boomerang engine

The plays described above can be run manually — with a spreadsheet, LinkedIn Sales Navigator, and a lot of nagging your team for intros. That works up to a point. Here is what changes when the same motion runs through a purpose-built engine:

Boomerang is the layer that turns warm outbound from a hobby into a channel — the same way Outreach and Salesloft turned cold email into a channel a decade ago.

90-day CRO transition plan

Days 1-15: Measure. Run the audit. Tag every opportunity in the pipeline by source. Compute cost-per-qualified-meeting for cold vs. warm. Present the numbers to the exec team. Set the Q4 target: 60-70% warm-sourced pipeline.

Days 16-30: Pool the graph. Consolidate every rep, exec, board member, and customer contact into a firm-wide graph. Tag connectors by strength. Identify your 50-100 strongest connectors across team, customer, investor, and partner networks. Boomerang deploys the graph in this window; manual teams take 45 days.

Days 31-60: Activate the plays. Launch three plays in parallel: (1) Customer Network Activation — 30-day cadence asking every customer for three peer introductions; (2) Executive Activation — monthly rhythm surfacing 10-15 target accounts to the CEO, CRO, and board for warm-intro asks; (3) Signal-Based Warm Outbound — job change alerts on every target account CFO/CRO/Head of Ops, routed to the best connector in the graph. See the pipeline generation complete playbook for the play-by-play mechanics.

Days 61-90: Restructure. Move 40-60% of SDR seats from cold to warm/signal roles. Rebuild the SDR comp plan around warm meetings booked, not emails sent. Retire the tools that no longer fit the motion. Report the new pipeline source mix in the Q4 board deck.

The delta at the end of 90 days: pipeline mix moves from 15% warm-sourced to 40-50% warm-sourced (with a path to 70% inside another quarter), CAC on new pipeline drops 30-40%, and the SDR org shifts from a cost center under scrutiny to a channel with defensible economics.

For the deeper economic case, see The Trust Collapse P&L, The State of Warm Intros 2026, and What Is Warmbound?.

Frequently asked questions

What is the gumball machine problem? The "gumball machine" is Jon Miller's frame for the MQL-to-SQL-to-pipeline model he helped invent at Marketo: put in a lead, turn the crank, out comes a meeting. Speaking in Anthony Kennada's Golden Hour newsletter, Miller argues the mechanic is broken — buyer attention is no longer a lever you can pull with volume. Cranking harder now produces unsubscribes, deliverability penalties, and CAC inflation instead of pipeline. The four numbers Golden Hour bundles alongside the frame — 70% CAC inflation, ~50% of teams missing pipeline goals, 55% drop in quality SDR conversations, and 80% of buyers arriving with a preferred vendor — describe why.

Why has CAC inflated 70%? Jon Miller's Golden Hour number — a dollar of new ARR costs 70% more than it did in 2021 — is the aggregate of three collapses. SDR unit economics broke as cold-email reply rates fell to 1.4% (Bridge Group, 2025). Inbox trust broke as Gmail/Outlook rerouted the majority of unauthenticated bulk mail. Buyer preference broke as 67% of the buying journey moved seller-free (Gartner) and 80% of buyers arrived with a preferred vendor (Golden Hour). The 70% inflation is the market repricing the SDR-heavy motion after its underlying assumptions failed. Reversing it requires changing the source mix — not sending more of the same email.

Is cold outbound really dead, or just harder? It's not dead — it's uneconomic for most segments. At a 1.4% reply rate (Bridge Group 2025) and quality conversations per SDR down 55% since 2014 (Golden Hour), the SDR seat no longer clears its cost for products under $50K ACV, and clears only marginally for products under $150K ACV. Cold outbound survives in narrow use cases: very high ACV enterprise (where a single meeting justifies enormous send volume), regulated markets with no other viable channel, and greenfield categories where no warm path exists yet. Everywhere else, the math has broken.

How is warm outbound different from just "asking for referrals"? Referrals are passive: a customer happens to mention you. Warm outbound is active and systematic: you map every warm path from your team, customers, investors, and partners into a firm-wide graph, monitor buying signals across your target account list, and route intro requests to the best connector at the exact moment the signal fires. Referrals happen. Warm outbound is engineered. Boomerang is the layer that makes it a repeatable, measurable channel.

What percentage of pipeline should be warm-sourced? For most B2B software companies with an ACV above $30K, the target is 60-70% warm-sourced by Q4 of Year 1 of the transition, with a path to 75%+ in steady state. The remainder is inbound (marketing-sourced) plus a narrow cold-outbound cohort targeting accounts with no warm path. This mix produces the CAC ratios boards are asking for in 2026 — and it's the only mix that gets you onto the day-1 shortlist that decides 92% of purchases (Golden Hour).

Do we fire the SDR team? Almost never. Most SDR teams get repositioned, not reduced. A well-run transition moves 40-60% of SDR seats from cold sequencing to relationship activation (running warm-intro plays across the graph) and signal monitoring (watching for buying signals and routing them to connectors). Cold sequencing shrinks to a small specialist cohort. Total SDR headcount often stays flat while pipeline output doubles.

Won't warm outbound "run out" as we exhaust our network? No — because customer network activation creates a compounding flywheel. Every warm-closed customer produces three warm intros for next quarter. The graph grows faster than you can exhaust it. The failure mode is the opposite: most orgs dramatically underutilize the graph they already have, running warm sources at under 10% of capacity.

What tools do I need to make this work? The minimum viable stack: a CRM with clean opportunity source tagging, a relationship intelligence layer that pools your team's networks, a signal engine for job changes and funding events, and a warm-intro orchestration layer. Boomerang is the orchestration layer — it sits on top of your CRM and data providers, maps warm paths into your target accounts, drafts intro requests in the connector's voice, and closes the loop when meetings book. The old sequencer/dialer/enrichment stack shrinks or exits.

Related reading

• The State of Warm Intros 2026
• The Day-1 Shortlist Is Golden (sister piece)
• The Warm-Intro CAC Model
• 3 Moats AI Can't Copy — Relationships, Experiences, Craftsmanship (sister piece)
• Customer Network Activation: The 2026 Playbook
• The Pipeline Generation Complete Playbook
• The SDR Is a Relationship Graph, Not a Dialer
• The Trust Collapse P&L
• What Is Warmbound?

Schema markup

Build the warm-outbound engine for your sales org

Boomerang is the warm-outbound orchestration layer for CROs restructuring away from cold. It pools every rep, executive, board member, customer, and partner network into a firm-wide graph, monitors buying signals across your target account list, and routes intro requests to the best connector at the exact moment the signal fires — closing the loop when the meeting books.

The pipeline motion your competitors are still trying to fix with more cold email. Book a 15-minute walkthrough →