Carlos Delatorre, CRO @Harness: Why Every Sales Rep Should Do Pipeline Generation | E1251

Summary

• Self-sourced pipeline generation is Carlos Delatorre’s one absolute non-negotiable in hiring reps — because “excellence in pipeline generation equals excellence in sales,” and fewer than 10% of reps can actually do it. A rep who has never sat in the “Dark Lonely Room” — “the 19th phone call where no one has picked up and I’ve left the same voicemail 19 times” — is unlikely to make the 20th call, even if they claim otherwise in interviews. He keeps nobody on his teams who can’t PG.
• Outbound in 20125 is not dead — “that’s crazy talk” — roughly two-thirds of meetings are still set by cold call, and Delatorre is “really glad” competitors believe otherwise. His proof specimen: a Dun & Bradstreet CISO who thought one call got the meeting, when rep Dave Bole’s book showed 11 emails, 8 calls, and a befriended exec admin’s Post-It note — 18 touches invisible to the buyer.
• The operating system behind “Pipeline Generation Tuesdays”: 20 target accounts per quarter, ~2 focus accounts declared each Thursday, weekend research into 25 contacts and 2-3 hypotheses per account, marketing’s “PG Passport” Friday noon, manager prep review Monday morning, role plays Monday afternoon, all-day calling Tuesday with catered meals and applause — then a Friday celebration Carlos runs himself. PG plus prep should consume ~30-35% of a rep’s time.
• The unit economics: 2-2.5 meetings a week is the sweet spot of good PG; 3:1 rep economics might be possible pre-product-market-fit but “once you’re scaling you want to be at 5:1 or better”; outbound works “certainly down to 50k” deal sizes and may start to break down below. To leaders pleading drought, Harry’s framing is: “We are not lacking for pipeline — the thing we’re lacking for is ramped salespeople that can go pursue it.”
• Founders shouldn’t hire a sales leader for the first two to three years — “the CEO really has to develop the plays” — and should avoid lone rangers: MongoDB attrition analysis found solo reps in a city churned ~4x higher, so the minimum hub is two salespeople plus a sales engineer. Identify misfires at boot camp, not month 18 — milestone-based ramp makes failure visible in weeks.
• Closing is dead as a lever; discounting rarely manufactures a deal: “I don’t believe salespeople sell. Buyers buy” — all a rep can do is create an environment conducive to the purchase decision, and in a considered purchase the champion “either exists or doesn’t… whether it’s 250k or 200k is not going to make the difference. It’s binary.”
• Rare operator candor on his own failures: his CEO stint struggled because the real problem was product-market fit, not go-to-market — “I didn’t have a toolkit” — and a painful hire was an EMC relationship-seller whose steakhouse Rolodex produced nothing. Best strategy he’s seen: likely Navan’s COVID bet-the-company pivot — 90% revenue drop, “burn the ships” into expense management, and go upmarket while nobody was traveling.

Deep dive

1. Sales is more science than art — and hiring starts with DNA you can’t change

• Delatorre’s opening frame: when people say someone was “born a salesperson,” they usually mean charisma — which “doesn’t hurt, but it’s not the core of it.” His deeper claim: “art is what we call something when we don’t understand it yet” — Picasso decomposes into colors and shapes, and sales decomposes the same way. “It’s more science than art, candidly.”
• The hiring trap is over-rotating on past success as a predictor. His formula instead: define the future job to be done, list the attributes and skills it requires, then interrogate the candidate’s accomplishments and life journey for innate attributes. “First the DNA, then the skills” — DNA can’t change; skill gaps just lengthen ramp, so you need explicit negotiables and non-negotiables.
• Mechanics worth stealing: opinions form through interviews and back channels; the multi-day “challenge” presentation comes only at the end, as validation of an already-formed conviction — asking for that work early is unfair to candidates.
• Why hiring committees fight: “we tend to index around our own strengths” — the high-EQ interviewer misses that a candidate is loose on process; the process-oriented one misses that the candidate is awkward.

2. Challenger scar tissue and self-sourced pipeline

• Deal-type experience beats domain experience — “better to teach the domain and hire as many of the sales skills as you can” — unless the domain is molecular-modeling-specific. His checklist: challenger-company success (near non-negotiable — “it’s so much harder to operate at a company where you don’t have a big logo that sells for you”), self-sourced pipeline generation (absolute non-negotiable), and comparable deal complexity, size, and stakeholder count (negotiable if the candidate is smart and driven — a $50-150k mid-market rep can step up to $500k enterprise deals).
• Why PG history can’t be waived: the Dark Lonely Room. “If they haven’t been in that lonely place, the likelihood that they are willing to make that 20th call is low — even if they thought they were willing upfront.” The rep must have seen the other side: “knowing that deal exists because I PG’d it — there’s a pride that comes with that.”

3. Founders: build the motion before hiring a leader — and avoid lone rangers

• For a brand-new company, don’t hire a sales leader yet: “the CEO really has to develop the plays, figure out the message.” Hire junior ICs and SDRs, collocate, “duke it out and refine the message” — build the organization only once the motion is repeatable. The juniors will need heavy lifting (“you may have to remind them that they should wear a shirt to a meeting”), but “that’s a beautiful part of the journey only a founder can really do well.”
• Same logic for enablement: an early-stage company “wouldn’t know what to build.” Construct V1 of training after ~20 closed customers — which means ~60 losses too, “so I’ve got 8 sales cycles of experience” — and keep it to two or three days.
• He agrees with a hire-two-reps rule attributed to likely Jason Lemkin, but for a different reason than competition: MongoDB’s attrition analysis found lone-ranger reps churned roughly 4x higher. The ten-second lean-over question becomes a scheduled meeting — “forget it.” Minimum hub: two salespeople plus a sales engineer.
• On remote: “all sales teams in-person are much more productive — 100%, no question about it.” Faster learning, contagious morale (“if I’m having a bad day but the person next to me is having a great day, that gives me hope”), and managers who can see and fix what a rep doesn’t know they’re doing wrong.

4. Why every AE must generate their own pipeline

• Three mechanisms beyond “more pipeline is like cowbell — the more the better.” PG un-sticks live deals: stuck too low, or trapped in one division? “You PG into the person you need to be at” — driving higher win rates on existing business. And outsourced PG breeds entitlement: he watched one company where AEs defined a workable lead as customer-has-agreed-to-buy-after-demo — “that’s like 80% of the sales cycle is done” — forcing enormous marketing and SDR spend just to feed the reps.
• When reps source their own, gift pipeline from SDRs, marketing, or partners gets “treated like gold” and converts better. Harry’s addition — which Carlos endorses — is accountability: no more “demand gen gave me shit leads.” The framing: AEs are responsible for generating the pipeline that supports their own quota; everything else is “a cherry on top.”
• Time allocation, answering Chris Degnan’s six-to-eight-meetings constraint: one day a week (~20%) of pure outbound plus 10-15% prep and follow-up — roughly 30-35% of a rep’s time. The justification is categorical: “excellence in pipeline generation equals excellence in sales” — a big-pipeline rep with sloppy execution still hits; a beautiful executor without pipe misses.

5. The PG operating system: Thursday declaration to Tuesday action

• Quarterly: each enterprise rep picks 20 top target accounts, validated with the manager on day one or two. Weekly: Thursday night, declare ~2 focus accounts in Salesforce. Friday through the weekend: research — execs on social media, the 10-K, job postings, what engineers gripe about on Blind — distilled into 2-3 hypotheses per account, 25 relevant contacts per account (names, emails, mobiles) and conversation starters for the top 10.
• The support system: marketing ships a “PG Passport” every Friday at noon — the week’s releases and case studies pre-boiled into two or three reusable sentences — and managers forward it with account-specific advice (“my neighbor works in HR at this account — she’d take a call to validate your contacts”). Harry’s pushback: is that scalable? Carlos: an hour or two, and the manager only needs a spark on two to four of ten accounts.
• Monday morning is a 1:1 where the manager sharpens the prep (“Jamie Diamond is not going to respond to an email about cloud cost management”); Monday afternoon is role-play training. Tuesday is all-day calling with catered meals — “we laugh when somebody gets hung up on, we applaud when somebody scores a meeting” — then beers, then a Friday wrap-up Carlos runs himself, celebrating meetings and the opportunities born of prior weeks’ PG. “Most salespeople on the planet are not doing what we do.”
• The design evolved from Parametric Technology’s unstructured “power hours,” which were limited because “you can easily go for an hour and never talk to anyone” — and without even one conversation, “it’s really hard to bring your A game for the next power hour.”

6. Outbound in 20125 is not dead — “that’s crazy talk”

• Harry channels the consensus that outbound is dead; Carlos’s response is total: “I strongly disagree… and I’m really glad other companies don’t [do it] — I hope a lot of those companies are our competitors.” Cold calling still sets about two-thirds of meetings; email’s job is to soften, so the name “lands with some sort of a positive aura” when the phone rings. On AI’s infinite personalized email: “you will — I’ll grant you that you will” — but voice remains the most effective channel.
• The load-bearing anecdote: rep Dave Bole and the Dun & Bradstreet CISO, who opened the meeting with “it must be nice to be able to just call a CISO and have him bring you in.” Dave’s book showed 11 emails, 8 phone calls, and a befriended executive admin who Post-It-noted his screen — “he had no idea about the 18 other touches.” The lesson: “you have to break through the noise.”
• How rare is this? Asked what share of reps are actually good at PG: “less than 10%. It’s elite” — and highly correlated with elite selling overall. Does he carry anyone who can’t? “No.”

7. The math: two meetings a week, 5:1 at scale, works down to $50k

• Don’t over-obsess on meetings booked — the easiest thing to track. 2 to 2.5 meetings a week is the sweet spot of genuinely good PG (a zero week can still be a good week); what deserves obsession is qualified pipeline coming out the other end.
• Harry stress-tests the business case: 8 meetings a month, 25% converting at $100k each is $1.2M against a $200-250k fully-loaded rep — roughly 5:1 only if everyone is ramped. Carlos’s rebuttal: that math is cold-territory-only — add expansions, partner leads that arrive “70% sold,” and SDR flow, and 4:1 solo becomes 6:1 blended. Rules of thumb: 3:1 might be possible pre-product-market-fit; “once you’re scaling you want to be at 5:1 or better.”
• Outbound economics hold “certainly down to 50k” and may start to break down below (junior reps, lower OTE); enterprise average deals run $200-400k. And Harry’s framing of the market complaint is: “We are not lacking for pipeline. The thing we’re lacking is ramped salespeople that can go pursue it.”

8. Ramp by milestones — fire at boot camp, not month 18

• Decompose ramp into milestones from week one: training entrance exam, role plays, meetings booked out of boot camp, reaching an economic buyer by month three. Companies that start worrying at months 9-12 and part ways at 18 are torching money and the rep’s time. He has cut people at boot camp: “we apologize… because we hired them. But why wait nine months and waste their time and our opportunity cost?”
• His enablement build: model the hypothetical fully-ramped rep, break every skill into competencies learnable in three hours or less, sort by ideal modality (solo, Zoom, classroom), and sequence toward the earliest possible new logo. The result is a ~9-month journey — three week-long trainings (boot camp, intermediate, advanced) plus a 3-hour cohort deep-dive. The predictive tell: “once somebody can land a new logo with the core product, the likelihood that they’re going to be successful is very high.” Sales cycle: ~120 days.
• Reps dial fast — “on the phone in the first couple of days,” pre-boot-camp, off a messaging framework with sample questions and proof points; the first meeting is typically set before boot camp ends.
• On deal reviews: MEDDPICC (crediting Mark Goldberg’s episode) is “the soundest framework I’m familiar with” — run it when needed, early-messy or late-forecast. You know it’s working “when you see salespeople talking to each other without a manager involved and they’re using the terminology… when qualification just becomes part of the culture.”

9. Verticalize the playbook, not the team — closing is dead, discounting rarely manufactures deals

• Centralizing industry knowledge so reps are “more relevant more quickly” — yes, even for horizontal products. But verticalizing the sales force “cuts against meritocracy”: if financial services is “the cat’s meow… is the team in New York amazing, or do they have all the banks?” Mix territories for fair opportunity; the only exception is ERP-grade divergence where one rep genuinely can’t sell both.
• Discounting can’t manufacture a deal in a considered purchase: the champion willing to fight the economic buyer “either exists or doesn’t — whether it costs 250k or 200k is not going to make the difference. It’s binary.” Post-decision budget accommodation is fine (250 to 180 to move fiscal years). And when a deal slips anyway: “pain is a great teacher… the answer has to be something we failed to do.”
• The deadest tactic of all: closing. “This idea that you can coerce a customer or do some psychological gymnastics to get someone to sign — first of all, that deal is probably going to churn… I don’t believe salespeople sell. Buyers buy. All salespeople can do is create an environment that is conducive to that purchase decision.”
• On SaaS churn and the seat glut, he’s calmer than the consensus: over-bought seats have “mostly worked through the system already,” but the future demands reps driving customer outcomes — “the big sales rep with the fancy suit that comes in, does the big deal and walks off — those days are long gone.” Answering Dave Kellogg’s farmer-vs-hunter line: account managers must PG too (into other divisions), and an AM driving real value makes the account “very difficult for a hunter to crack.” Mature-territory mix: ~1/3 PG, 20-30% customer care, the rest prosecuting deals — with dials varying by territory.

10. Scar tissue: the CEO detour, the EMC hire, and Navan’s bet-the-company pivot

• To GTM leaders eyeing the CEO seat: a company hiring a new CEO has a problem set — diligence it and honestly map your abilities against it. His own run: he believed the problems were go-to-market; they were product-market fit. “I didn’t have a toolkit… I wasn’t particularly effective at guiding the company through refining product-market fit, and it took me a while to figure that out. It was a painful experience” — he’d brought over Keith Butler to run sales along the way.
• His most painful hiring mistake: a relationship-seller off big numbers at EMC. Three months of “fancy lunches at steakhouses in New Jersey” with senior people “completely irrelevant to the product,” and nothing came of it — the man simply “lacked the skills to go and call a stranger and deliver a value prop.” Expensive, hired outside standard parameters, on Carlos’s own advocacy.
• The best sales strategy he’s witnessed: the travel company he joined in January 2020 (likely Navan), which lost ~90% of revenue when the world locked down in March. Founder Ariel came back from vacation saying “by God, we’ve got to build an expense management solution — and burn the ships,” pivoting ~90% of engineering overnight while the exec team thought it “a bit dramatic” — everyone still assumed COVID would last weeks. Simultaneously the company went upmarket: “were you happy with your travel program before COVID? What better time to overhaul it than now, when no one’s traveling?” The two pivots “saved the company… it would have died.”
• The origin-story close, from his medical-equipment days: an office manager said she’d already bought the competing Becton Dickinson blood analyzer. Carlos had a BD unit in his closet — so he offered it free if his likely Coulter Counter demo lost, drove all night through a snowstorm to Jackson, California, demoed the doctor live on Friday, and flipped the order. On AI’s next five years: it automates research and prep, so productivity trends up and attrition down.