Eilon Reshef (co-founder and CPO)02 January 2025

Inside Gong: How teams work with design partners, their pod structure, autonomy, trust, and more

7Frameworks

14Insights

Episode overview

Lenny Rachitsky interviews Gong co-founder and chief product officer Eilon Reshef about the company's autonomous, cross-functional pod model and its unusually intensive work with design partners. Reshef explains why customer collaboration makes nearly all shipped capabilities meaningfully used, how trust and fast decisions increase velocity, and why AI products still require specialist expertise and measurement. They also discuss the spiral method for learning, Gong's deliberately narrow initial customer profile, and mistakes Reshef carried forward from his previous company.

Key ideas

Gong organizes product work into autonomous cross-functional pods containing product, UX, engineering, and fractional writing and analytics support.
Each pod works closely with multiple design partners, often meeting weekly and using converging feedback to distinguish broad needs from one-customer requests.
A research coordinator recruits appropriate customers and schedules interviews so product managers can focus on learning and building.
Autonomy requires teams to solicit feedback and own decisions, while leaders accept less visibility in exchange for greater velocity, engagement, and product quality.
For close 51–49 choices within an established domain, Reshef favors deciding quickly because additional deliberation often does not materially improve the result.
LLMs do not replace AI expertise: teams still need data scientists, prompt expertise, evaluation systems, and clear measurements to improve beyond an initial version.
The spiral learning method repeatedly consults new experts until answers converge and additional conversations yield little new information.
Gong began with an intentionally narrow ideal customer profile to create repeatability and word-of-mouth before expanding.

Transcript available · source text is retained privately and is not published

Frameworks in this episode

Leadership4 steps

Autonomy as Selfish Leadership

Let people bring themselves to the work; you trade a little visibility for higher velocity, morale, and results.

Read →

Marketing4 steps

Small-Pond ICP Narrowing

Constrain your first ICP so tightly that customers talk to each other — then let word of mouth light the fire.

Read →

Strategy3 steps

The 51-49 Decision Rule

When options are near-even, just pick one fast — more data won't meaningfully improve a close call.

Read →

Innovation4 steps

The AI Expertise Layer

Don't outsource all AI to foundation models — keep expertise and measurement to break the quality glass ceiling.

Read →

Innovation4 steps

The Design-Partner Pod Model

Give each cross-functional pod 6-12 real customers to co-build with, so features ship pre-validated.

Read →

Productivity3 steps

The Research Coordinator Role

Borrow the recruiting-coordinator role to schedule all customer research, so PMs never lose time to logistics.

Read →

Self-Mastery3 steps

The Spiral Method for Learning Fast

Chain expert-to-expert referrals until you hear only repeats — that's how you know you've hit the bullseye.

Read →

People & resources mentioned

Attributed to the moment in the episode. Timestamps are approximate.

People · 7

Eilon ReshefMentionstoday my guest is Alon RF Alon is co-founder and chief product officer at gong
“today my guest is Alon RF Alon is co-founder and chief product officer at gong”
▶ Watch at ~1:00approx.
Lenny RachitskyMentionstoday we've got another very special compilation episode something I've been pulling on more and more with the podcast and the newsletter
“hey it's Lenny”
▶ Watch at ~3:30approx.
Matt AbrahamsMentionsMatt is a professor at Stanford University's Graduate School of Business where he teaches a very popular class on communication and public speaking
“host in Stanford Graduate School of Business lecturer Matt Abrahams sits down with experts”
▶ Watch at ~3:30approx.
Isaac AsimovCoinedhe's a science fiction writer
“there's a maybe I'll do a spoiler for Isaac Asimov he's a science fiction writer”
▶ Watch at ~27:30approx.
Amit BendovMentionsmy co-founder Amit who is the CEO
“I had this discussion with my co-founder Amit who is the CEO”
▶ Watch at ~29:00approx.
Gary OldmanMentionsmentioned
“this British TV series called slow horses uh with Gary Alman”
▶ Watch at ~49:30approx.
Rory SutherlandMentionsRory more than maybe anyone in the world he just sees the world through these these system lenses
“a previous guest had uh suggests Rory southernland has this pitch that you should have two dishwashers”
▶ Watch at ~52:00approx.

Resources · 37

GongCoinedsoftware · Gong
“Alon is co-founder and chief product officer at gong”
▶ Watch at ~1:00approx.
LoomUsessoftware · Loom
“companies are already powered by work OS including ones you probably know like vercell webflow and Loom”
▶ Watch at ~2:30approx.
WebflowUsessoftware · Webflow
“companies are already powered by work OS including ones you probably know like vercell webflow and Loom”
▶ Watch at ~2:30approx.
VercelUsescompany · Guillermo Rauch
“companies are already powered by work OS including ones you probably know like vercell webflow and Loom”
▶ Watch at ~2:30approx.
WarrantMentionssoftware · Warrant
“work OS also recently acquired warrant the fine grain authorization service”
▶ Watch at ~2:30approx.
ZanzibarMentionssoftware · Google
“Warren's product is based on a groundbreaking authorization system called Zanzibar which was originally designed for Google”
▶ Watch at ~3:00approx.
WorkOSRecommendscompany · WorkOS, Inc.
“you should consider work OS”
▶ Watch at ~3:00approx.
Google DocsMentionssoftware · Google
“originally designed for Google to power Google Docs and YouTube”
▶ Watch at ~3:00approx.
YouTubeMentionswebsite · Google
“originally designed for Google to power Google Docs and YouTube”
▶ Watch at ~3:00approx.
Auth0Mentionssoftware · Okta
“it's a dropin replacement for ozero”
▶ Watch at ~3:00approx.
Think Fast, Talk SmartRecommendspodcast · Matt Abrahams
“I want to recommend a podcast called think fast talk smart”
▶ Watch at ~3:30approx.
Stanford Graduate School of BusinessMentionscompany · Stanford University
“host in Stanford Graduate School of Business lecturer Matt Abrahams sits down with experts”
▶ Watch at ~3:30approx.
Gong ForecastMentionsproduct · Gong
“we launched a forecast product that was a pod working on that”
▶ Watch at ~6:00approx.
Google SheetsMentionssoftware · Google
“probably now it's already in Google Sheets always”
▶ Watch at ~18:00approx.
VantaRecommendssoftware · Christina Cacioppo
“get $1,000 off vanta when you go to vana.com plenny”
▶ Watch at ~23:00approx.
The Machine That Won the WarRecommendsother · Isaac Asimov
“he has this like a short story that the machine that won the war so you can look it up”
▶ Watch at ~27:30approx.
SpotifyMentionssoftware · Spotify
“Spotify has this kind of value they call talk is cheap”
▶ Watch at ~30:00approx.
NvidiaMentionscompany
“after that course I ended up buying Nvidia stock”
▶ Watch at ~32:30approx.
FigmaMentionssoftware · Figma
“I think figma calls their AI feature like first draft”
▶ Watch at ~35:00approx.
First DraftMentionsproduct · Figma
“I think figma calls their AI feature like first draft”
▶ Watch at ~35:00approx.
WebExUsessoftware · Cisco
“using WebEx which was the big one at the time”
▶ Watch at ~42:00approx.
Crossing the ChasmMentionsbook · Geoffrey Moore
“I did read Crossing casm and I told myself nah I can do way better than that”
▶ Watch at ~42:30approx.
American ExpressMentionscompany
“one of the Cosmetics companies in American Express in Cisco like different Industries”
▶ Watch at ~43:00approx.
CiscoMentionscompany
“one of the Cosmetics companies in American Express in Cisco like different Industries”
▶ Watch at ~43:00approx.
L'OréalMentionscompany
“we had one customer in I think it was L'Oreal”
▶ Watch at ~43:00approx.
The Ideal ExecutiveRecommendsbook · Ichak Adizes
“I think one that is sort of the starter one is I think it's called right now the ideal EX executive”
▶ Watch at ~46:30approx.
MismanagementMentionsbook · Ichak Adizes
“I think the original version is funnily enough I think it was called mismanagement”
▶ Watch at ~47:00approx.
Crucial ConversationsRecommendsbook
“I like crucial conversation”
▶ Watch at ~49:00approx.
Slow HorsesRecommendstv show · Will Smith
“this British TV series called slow horses uh with Gary Alman and it's it's a really kind of fun”
▶ Watch at ~49:30approx.
NetflixUsescompany · Netflix
“nowadays there's Netflix so you can find stuff”
▶ Watch at ~49:30approx.
AmazonUsescompany · Jeff Bezos
“I go to some Amazon or Ebay wherever I just buy a new one”
▶ Watch at ~51:00approx.
eBayUsescompany · eBay
“I go to some Amazon or Ebay wherever I just buy a new one”
▶ Watch at ~51:00approx.
Hanlon's razorRecommendsother
“it's called handon razor you can look it up Wikipedia whever”
▶ Watch at ~52:30approx.
WikipediaUseswebsite · Wikimedia Foundation
“it's called handon razor you can look it up Wikipedia whever”
▶ Watch at ~52:30approx.
LinkedInUseswebsite · LinkedIn
“I'm pretty available in LinkedIn is probably the best way”
▶ Watch at ~55:30approx.
Apple PodcastsRecommendssoftware · Apple Inc.
“you can subscribe to the show on Apple podcast Spotify or your favorite podcast app”
▶ Watch at ~56:00approx.
Lenny's PodcastCoinedpodcast · Lenny Rachitsky
“you can find all pasted episodes or learn more about the show at lenp podcast.com”
▶ Watch at ~56:30approx.

Spot an error or want something removed? Request a correction or removal.

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Hot Take· 1

Hot Take28:00

If It's a 51/49 Call, Just Decide — Even One-Way Doors

Eilon pushes making decisions fast, even large one-way-door ones, before you have all the information. His logic: when a choice is close (51/49 rather than 70/30), neither option is super wrong, so gathering more data or people rarely changes the outcome. He cites deciding not to buy a company that, in hindsight, wouldn't have changed Gong's position, and notes decision-making drains you like physical exertion.

Close calls (51/49) mean no option is badly wrong — just pick one
More data and more people rarely improve a genuinely close decision
A company they chose not to buy wouldn't have radically changed their position
Decision-making is depleting, like physical exertion — overthinking rarely raises quality

“you end up being like 51 40 49 no decision is going to be like super wrong”
Eilon Reshef · 28:30

#decision-making#speed#leadership

Explainer· 3

Explainer06:30

Every Gong Pod Builds With 6–12 Design Partners, Hand in Hand

Gong takes the design-partner concept to an extreme: every product pod works directly with a set of customers (often 6–12, sometimes as few as one for niche features) while building. Eilon argues customers know their pain better than the product team, so building alongside them removes the risk of shipping something nobody uses. This is unusual because many companies bar product teams from talking to customers at all.

Each pod works hand-in-hand with a set of design partners, not just one or two
A dozen partners is smarter than one so you build for the customer base, not a single account
Building this way removes the risk of shipping features that never get used
Contrasts with companies that forbid product teams from talking directly to customers

“I do believe customers know much better than that what they need”
Eilon Reshef · 16:30

#design-partners#product-teams#customer-development#b2b

Explainer09:30

The "Research Coordinator" Role Borrowed From Recruiting

Coordinating design partners across 25–30 pods is huge overhead, so Gong borrowed the recruiting-coordinator idea from talent acquisition. One person acts as a research coordinator: she takes the PM's target market and learning goals, sifts the customer base via a micro-CRM and micro email campaigns, and books the meetings so PMs just show up.

Idea borrowed from the recruiting-coordinator role in talent acquisition
PMs specify target market/ICP and what they want to learn
She uses a micro-CRM and micro email campaigns to source and schedule partners
Frees PMs from the scheduling burden of coordinating a dozen companies

“in our in our product team there's one person who's basically a research coordinator and she's responsible for reaching out”
Eilon Reshef · 10:00

#design-partners#operations#product-teams#scaling

Explainer42:00

The Power of a Narrow ICP: 5,000 Companies and a "Small Pond"

Gong deliberately started ultra-narrow — US, English, video conference over WebEx, selling software worth $1K–$100K — a bucket of only ~5,000 companies. Eilon's earlier company failed by spanning wildly different industries with no common lingo. A tight ICP creates a "small pond" where similar companies talk to each other, producing a rare B2B viral effect: a prospect became a customer because a salesperson they interviewed would only work at companies using Gong.

Deliberately narrow ICP: US, English, WebEx video calls, $1K–$100K software
Only ~5,000 companies fit the initial bucket
A previous company failed spanning industries with no shared lingo
A tight pond of similar companies creates a rare B2B viral effect

“there was no way we could scale it because everybody had their own lingo”
Eilon Reshef · 42:30

“well I'm only going to work for companies that use gong”
Eilon Reshef · 43:30

#icp#go-to-market#crossing-the-chasm#positioning

Story· 4

Story07:00

The PM Who Showed a Half-Built Product and Told the Customer to "Hit Save"

A PM demoed a not-yet-built forecast feature to a design partner. When the customer hit save he got an error, and the PM promised it would work in a week. Customers later said they appreciated that the team interpreted and digested their feedback rather than literally building what they asked for.

The team showed a half-built product before it functioned
They built to what the feedback meant, not the literal request
Customers valued seeing progress move according to their feedback

“I asked him to hit save he hit save and got an error message and I told him let's meet again in a week”
Eilon Reshef · 07:00

#design-partners#product-development#customer-feedback

Story17:30

Why Autonomy Is "Selfish" — the Picnic Potluck Lesson

Eilon frames giving teams autonomy as selfish: you get more out of people when you let them be themselves. He tells the story of a school picnic where a rigid ingredient list produced the lowest-common-denominator food, but telling everyone to "bring your own thing" produced a feast — people drove miles for specialties and brought their personality to the table. The same holds for software teams.

You get more from people when you let them work their own way, within limits
A fixed list drives everyone to the easiest, cheapest contribution
"Bring your own thing" produced a feast and repeated every year
Autonomy keeps people thinking and motivated, better short and long term

“I just think you get more from everybody if if you kind of let them be themselves and and do things in the way that…”
Eilon Reshef · 17:30

“here's a different method just tell everybody bring your own thing”
Eilon Reshef · 18:00

#autonomy#leadership#culture#management

Story20:00

He Never Installed Control Software on His Kids' Devices

Eilon applies the same autonomy principle at home: he never installed antivirus or screen-time software on his kids' devices, treating self-protection as their responsibility. After negotiating a two-hour computer limit, his daughter came back three days in a row asking him to install limiting software — flipping the dynamic so she owned the decision and he was helping.

He installed no protection or control software on his kids' devices by choice
Self-protection was framed as the child's own responsibility
His daughter later asked him to install the limiting software herself
When responsibility flips to the person, the parent becomes a helper

“could you please install the software on my machine so I can help me like control my limits and I love it when it's the…”
Eilon Reshef · 21:00

#autonomy#parenting#responsibility

Story45:00

Fail Corner: Going Horizontal Without a Focused ICP

Eilon's worst repeated mistake, from a previous year-2000 company, was going horizontal instead of specializing in a market. They landed three customers in three different segments, then scaled to ~20 salespeople on investor advice — all failing, because there was no true product-market fit and no focused, repeatable ICP. He vows to make new mistakes, not that one again.

Going horizontal instead of specializing was the core error
Three customers in three segments looked like traction but wasn't
Scaling to ~20 salespeople on investor advice before PMF failed
Root cause: no focused, repeatable ICP

“we didn't have a true focused ICP with like a very very repeatable product Market”
Eilon Reshef · 46:00

#failure#product-market-fit#icp#go-to-market

Tool· 2

Tool50:30

The $10 Two-Cutlery-Basket Dishwasher Hack

Eilon's favorite recent "product" is a second silverware caddy for the dishwasher. After losing and re-buying one, then finding the original, he realized keeping a second basket in the sink lets you continuously load cutlery while the machine runs or before you unload it. He notes a guest's related idea — Rory Sutherland's pitch that everyone should own two dishwashers, one clean and one dirty, so you never put dishes away.

Keep a second cutlery basket in the sink to load continuously
Costs roughly $10–15 and reorganizes the kitchen workflow
Rory Sutherland's variant: own two dishwashers, one clean and one dirty
No dishwasher maker thinks to offer a second basket

“if you have two of those baskets you put one of them in the sink and you can just like continuously load your Cutlery”
Eilon Reshef · 51:00

#product#life-hack#recommendation

Tool46:30

Book Rec: "The Ideal Executive" (originally "Mismanagement")

Eilon's top management-book recommendation is The Ideal Executive, originally titled Mismanagement. It defines people by four characteristics — producer, administrator, integrator, and change agent — and argues nobody embodies all four. He values it less for the specific model than for the habit of viewing yourself and your team through key characteristics, which makes leadership discussions far faster. He also recommends Crucial Conversations.

The Ideal Executive was originally titled Mismanagement
Four types: producer, administrator, integrator, change agent — no one is all four
Value is a shared vocabulary for high-velocity team discussions
Also recommends Crucial Conversations

“nobody wants to buy a book called mismanagement much rather buy a book that's called ideal executive”
Eilon Reshef · 47:00

#books#management#leadership#recommendation

Takeaway· 4

Takeaway15:00

Nearly 100% of Features Built This Way Get Used

Because features are co-built with design partners, close to 100% of what Gong ships ends up used by a significant number of people — more than 95% of capabilities, which Eilon believes is higher than most companies. Not everything gets charged for, and the occasional miss comes from value being narrow or quality gaps, not from nobody wanting it.

Very close to 100% of features end up used by a significant number of people
More than 95% of capabilities are used in a significant way
Misses are usually narrow applicability or willingness to pay, not lack of use

“I would say very close to 100% of the features we build end up being used by a significant number of people”
Eilon Reshef · 15:30

#design-partners#product-metrics#success-rate

Takeaway32:30

LLMs Don't Solve Everything — Keep Real AI Expertise In-House

Having built AI products longer than most, Eilon warns against swinging from "you need data scientists for every project" to "LLMs will solve everything." LLMs have huge utility but can't do specialized tasks like predicting deals. Companies still need core AI competencies and measurement (Elo-style ranking) to move past a V1, or they hit a glass ceiling.

Don't swing to the extreme of assuming LLMs solve everything
Specialized tasks like deal prediction need purpose-built models, not LLMs
Without measurement you can't tell if V2 beats V1
Lack of operational rigor around AI creates a glass ceiling

“hey llm is going to solve everything because LMS don't solve everything they have huge utility we use LMS over the place”
Eilon Reshef · 33:00

#ai#llms#product-strategy#machine-learning

Takeaway35:00

Treat AI Output as a "First Draft" — Figma's Framing

Eilon likes Figma's naming of its AI feature "first draft" because it sets the right expectation: not best, not great, but a good starting point. Knowing whether an LLM's output is 90% or 50% accurate changes how you build the product, design the workflow around it, and train users on what to expect.

Naming AI output a "first draft" conceptualizes it correctly
90%-accurate vs 50%-accurate output demands different product design
The framing shapes the workflow and what users are trained to assume

“figma calls their AI feature like first draft H which is a term I like”
Eilon Reshef · 35:30

#ai#product-design#llms#user-expectations

Takeaway52:30

Hanlon's Razor: Don't Attribute to Malice What Stupidity Explains

Eilon's go-to life motto is Hanlon's razor — never attribute to malice what is adequately explained by stupidity. He finds it genuinely useful because people so often read malice into behavior (a customer not replying, someone withholding feedback) when the real cause is that the person simply didn't know, didn't care, or wasn't trained.

The razor: never attribute to malice what stupidity adequately explains
People routinely misread neutral behavior as malicious
Usually the cause is ignorance or inattention, not ill intent
A funnier phrasing of "assume good intent"

“never attribute to melice that that which is adequately explained by stupidity”
Eilon Reshef · 53:00

#mental-models#life-motto#psychology

Key ideas

Frameworks in this episode

Autonomy as Selfish Leadership

Small-Pond ICP Narrowing

The 51-49 Decision Rule

The AI Expertise Layer

The Design-Partner Pod Model

The Research Coordinator Role

The Spiral Method for Learning Fast

People & resources mentioned

People · 7

Resources · 37

Insights & moments

◆Hot Take· 1

If It's a 51/49 Call, Just Decide — Even One-Way Doors

✶Explainer· 3

Every Gong Pod Builds With 6–12 Design Partners, Hand in Hand

The "Research Coordinator" Role Borrowed From Recruiting

The Power of a Narrow ICP: 5,000 Companies and a "Small Pond"

❝Story· 4

The PM Who Showed a Half-Built Product and Told the Customer to "Hit Save"

Why Autonomy Is "Selfish" — the Picnic Potluck Lesson

He Never Installed Control Software on His Kids' Devices

Fail Corner: Going Horizontal Without a Focused ICP

✦Tool· 2

The $10 Two-Cutlery-Basket Dishwasher Hack

Book Rec: "The Ideal Executive" (originally "Mismanagement")

▲Takeaway· 4

Nearly 100% of Features Built This Way Get Used

LLMs Don't Solve Everything — Keep Real AI Expertise In-House

Treat AI Output as a "First Draft" — Figma's Framing

Hanlon's Razor: Don't Attribute to Malice What Stupidity Explains

Hot Take· 1

Explainer· 3

Story· 4

Tool· 2

Takeaway· 4