Orang Digest, week of September 14, 2026
This Week's Episodes
- The LeanScale Podcast: Joey Gilkey, founder and CEO of TitanX (host Anthony Enrico, LeanScale co-founder), on why precision outbound beats volume dialing and where AI belongs in a sales motion.
- GTM Science: hosts Eddie Reynolds and Rachael Bueckert of Union Square Consulting, using a viral pizza meme to break down the six problems killing B2B outbound.
- 30 Minutes to Presidents Club: Armand Farrokh, former VP of Sales at Pave and ex-Carta SDR leader, on the step-by-step system for building an elite outbound SDR org.
- Verticals: Mike Duboe, general partner at Greylock and ex-Stitch Fix growth lead, on how AI agents reshape marketplaces - full write-up in Notion.
- AI to ROI: hosts Ray Rike (Benchmarkit) and Peter Buchanan (NewPlan) on the OpenAI vs Anthropic enterprise battle - full write-up in Notion.
The 3 episodes below are this week's deep dives; the other 2 are written up in full in the Notion archive.
Executive Summary
- Outbound isn't dead - unqualified volume is. Three shows agreed: the highest-activity rep is rarely the top performer, and better targeting fixes more than copy ever will (Gilkey on LeanScale, Reynolds on GTM Science, Farrokh on 30MPC).
- Parallel dialing wrecks connect quality. One team's 8,000 single-line dials beat another's 200,000 parallel dials on meetings booked, because parallel dialing degrades the answer experience and gets numbers flagged as spam (Gilkey on LeanScale).
- Stop selling to people with no demand. Sales can't create demand economically, only capture it: find people with an existing burning problem instead of converting the indifferent (Reynolds on GTM Science).
- Turning around a dead sales floor is an energy problem, not a skill problem. It shifts in 24 hours with private buy-in before public rollout, plus fast personnel cuts (Farrokh on 30MPC).
- The conversation, not the meeting, is the real unit of outbound value: every call is intel that should feed other channels, whether or not it books (Gilkey on LeanScale).
Framework of the Week
Joey Gilkey's four pillars of phone-led outbound (LeanScale), with the conversation as the diagnostic engine at the center. Get these four right and outbound scales predictably: the conversation in the middle tells you which pillar is broken.
- List: are you targeting the right accounts and contacts, with accurate data? "Done" means the list passed QA after the data provider handed it over - Gilkey's rule is never trust the raw output, since providers get the phone number wrong 15-20% of the time and titles wrong more often. The list is your ceiling: right title at the wrong account, or wrong title at the right account, and every conversation is wasted.
- Messaging: is the right message reaching the right contact? "Done" is A/B-tested scripts where the conversation shows which lands, not a clever hook - Gilkey tested the ten most popular openers and found no winner. Tone, pace, and deliberate pauses matter more than words.
- Rep: who delivers the message, and are they ramped to run the right inputs? "Done" is a rep whose connects convert, measured, not assumed.
- Follow-up: the systematic re-touch of the 85-90% of conversations that don't book. "Done" is a real cadence for circle-backs: booking a meeting 10-15% of the time is good, so most conversations fail, and that failure bucket is where future wins live.
Apply it this week: pull one rep's last month of conversations and trace each pillar - good accounts, one script outperforming another, connects converting, no's getting re-touched. Measure connect-to-meeting rate per pillar. Avoid treating booked meetings as the only output and throwing away the intel every conversation produces.
Benchmarks & KPIs Mentioned
- Cold call connect rate ~3-5% is the market benchmark (Gilkey, Farrokh); if you're below this, look at data quality and spam-flagged numbers before blaming reps.
- Cold call set rate ~20% for top SDRs (Farrokh); use as the bar separating A-reps from the pack on the phones.
- SDR activity target: ~25 accounts worked, ~3 contacts each (~75 sequenced), 200 dials, 4-5 meetings/week for mid-market (Farrokh); a starting frame to sanity-check your own SDR quotas.
- Booking a meeting 10-15% of conversations is good outbound (Gilkey); the inverse - 85-90% "failures" - is your follow-up pipeline, not waste.
- Cold email sequences at double-digit reply rates (Farrokh); a four-sentence email cleared this bar where three-paragraph templates didn't - a target for your sequence rewrites.
- CAC-payback floor for outbound: rule of thumb ~$40-50k ACV, though tightly-run motions push it lower (Reynolds); use to decide whether outbound belongs in a given segment at all.
- Reach rate: you'll only ever connect with ~200 of any 1,000-person list no matter how many rounds you dial (Gilkey); watch this to stop burning dials on the ~800 who will never answer.
Across the Shows
The dominant signal this week is a three-show consensus that outbound targeting, not messaging or effort, is where revenue is won or lost - and it cuts against the usual reflex of scaling headcount and dial volume to hit the number. Gilkey on LeanScale, Reynolds on GTM Science, and Farrokh on 30MPC arrived there independently. Farrokh's line is sharpest: the highest-activity rep is almost never the highest performer, and the wrong cold call on the wrong account is an F call no matter how good the rep. Gilkey supplies the economics: 8,000 single-line dials generated 25% more meetings than 200,000 parallel dials, a 30x efficiency gap, because volume degrades both the answer experience and carrier reputation. Reynolds frames it from demand: sales can't create demand, only capture it, so pointing reps at people with no burning problem wastes effort. The action for revenue leaders is concrete: audit where dials are landing before funding more of them, and treat target-list QA as a better investment than another sequence rewrite.
The real disagreement is on org design. Gilkey argues SDR headcount should be cut by at least half and rebuilt as a smaller, better-paid "special forces" team, plus a new MDR role that can't book meetings at all, only gather account intel for AEs. Farrokh, meanwhile, is still building and defending a classic 20-person SDR floor with weekly rhythms and activity thresholds. Both fit different motions: Gilkey's precision model works when connect rate and ACV justify $120k US reps; Farrokh's volume-plus-culture engine fits mid-market motions where at-bats still matter. The real question is which regime your connect rate and deal size put you in, not which guru is right.
The Deep Dives
The three episodes worth your time this week, with the mechanics behind the takeaways above.
Gilkey's core case study: one customer ran 200,000 parallel dials over 11 months at a 2% connect rate and 2% meeting rate; a rep doing 8,000 single-line power dials hit a ~9.8% connect rate and 12% meeting rate. Compounded, that's a 30x efficiency gap: the parallel-dial team needed 30x the dials for the same output. Parallel dialing breaks because firing five calls and patching through whoever answers leaves the rep unprepared, adds a mechanical bridge delay that makes people hang up, and trains carriers to flag the numbers as "spam likely," so nobody answers.
The deeper idea is "reach rate": on any 1,000-person list, you'll only ever connect with about 200 people no matter how many rounds you dial, and brute-force dialing takes roughly 7,200 dials to get there. Predicting which 200 will answer before you dial means reps only research people they'll actually reach, so prep effort stops being mostly wasted.
Takeaways & Implications
- If you're in RevOps: instrument connect rate and meeting rate separately by dial method, then pull your parallel-dial logs. If parallel dialing is inflating dial counts while meeting rates sink, you have a case to switch to power dialing and a spam-reputation problem to fix.
- If you're a CRO: reconsider whether your SDR floor should be volume-first. A smaller team with better connect data can out-produce a larger one - a comp and headcount decision, not a tooling one.
- If you lead marketing: treat sales conversations as an intel feed. Signals reps hear on calls beat third-party intent data and should shape copy, targeting, and named-account content.
Reynolds' frame: a pizza shop calling a past customer who's probably hungry runs better outbound than most B2B teams, because it nails intent, relevance, segmentation, and clean systems. Of six failure modes, the load-bearing one is intent. Sales can't economically create demand, only capture it. The weight-loss analogy: you want the person saying "I have to lose weight, it's my top priority," not the one who "might want to." Chasing the indifferent is the biggest fallacy in sales.
His priority order for who to work: expand existing customers first, then marketing hand-raisers, then engaged-but-quiet leads you reverse-engineer, then third-party intent, and only then pure cold. Reverse-engineering is the mechanic: look at who actually bought, find the shared signals (e.g. three pricing-page hits), and even a 1% conversion can justify outbound if the math works.
On relevance vs personalization: relevance is segmentation, not merge fields. One rep working one buyer type in one company type all day beats a rep context-switching across CFOs, CMOs, and COOs. And the discipline holds: run outbound profitability channel by channel and segment by segment, kill what doesn't pay, double down on what does.
Takeaways & Implications
- If you're in RevOps: audit outbound target lists against closed-won signals. Reverse-engineer traits and behaviors of buyers who actually converted, then rebuild target criteria around them - even a low but real conversion rate beats spray-and-pray.
- If you're a CRO: measure outbound on CAC payback and LTV:CAC per channel and per segment, not on activity or raw conversion. A motion unprofitable in SMB may pay in enterprise; find the profitable pocket and concentrate there.
- If you lead marketing: build segments narrow enough that reps work one buyer type at a time. Relevance at scale comes from segmentation, not AI-generated pseudo-personalization that prospects instantly clock.
Farrokh's turnaround thesis: a dead sales floor is an energy problem, not a skill problem. His 24-hour flip: before a team all-hands, he ran 16 hours of back-to-back one-on-ones, winning each rep privately and asking them to riff publicly on Thursday. When he then named each person's priorities in the meeting, the room's energy flipped, using peer pressure to convert the holdouts. Paired with "flip two, fire two, hire two": elevate two young guns who buy in, cut the two underperformers who won't take coaching (they'd otherwise scorch earth for nine months), hire two fresh for new energy.
Five weekly rhythms run 90% of SDR teams: Monday metrics meeting (commitments and consequences - Farrokh commits to something he controls, like 40 dials Friday, with a public consequence if he misses), Tuesday/Thursday teardowns (cold call, email, or account), Wednesday 1:1s, Friday team dials with every booked meeting posted publicly.
Targeting beats activity: skills raise a rep one letter grade, but the wrong account is an F no matter what. His minimum viable cold call: context-first opener, problem-first pitch (reverse pitch before solution), then agree with the objection and sell the test drive. The four-part cold email (what I noticed / the problem / the solution / a simple CTA) got sequences to double-digit replies where three-page templates failed.
Takeaways & Implications
- If you're a CRO turning around a team: win reps privately before the public rollout, and make the hard personnel calls in weeks, not months. You never recover floor energy after months of a dead room.
- If you're in RevOps: instrument the L1/L2 metric split - accounts worked and dials as leading indicators, and only drop to connect rate, set rate, and reply rate when meetings dip. This tells you whether a rep has a targeting problem or a skill problem.
Also Processed
- Verticals: Mike Duboe (Greylock) argues AI makes new markets addressable to the marketplace model, but agents threaten take rates on the "match," pushing value downstream to trust, guarantees, and proprietary catalogs.
- AI to ROI: Ray Rike and Peter Buchanan on OpenAI vs Anthropic - the enterprise winner will be decided by go-to-market execution and partner trust over 24-36 months, not one quarter's financials.
Full write-ups are in the Notion archive.
Worth Your Headphones
The LeanScale Podcast: "Outbound Isn't Dying. Yours Is." with Joey Gilkey. The 30x dial case study and the restaurant analogy for AI both land harder in Gilkey's own voice than on the page - he thinks in vivid metaphors and the energy of a founder who literally bet the farm carries the whole hour.