Do forward deployed engineers create real customer value, or do they substitute for a missing product?

Shyam Sankar · Jan 17, 2024

Building software for national security requires people who interact with customers to shape the product, because the intersection of domain and software is nuanced and the information asymmetry is massive

The problem is deep and complicated, and much of the software deploys into classified environments the engineers can't see into

Scope: describing early Palantir's specific situation with no clearances

8:43 20VC: Palantir CTO on The Broken Incentive Structure of How Governments Buy Defence, The Danger of Defence Spending at Historic Lows, How Elections and Wars Change Government Defence Buying & Why Budgets are Anti-Creative with Shyam Sankar

Nick Frosst · Sep 1, 2025

Forward deployed engineers are not a crutch for weak technology; enterprise software inherently has to be matched to how each business is set up

Consumer technology can work out of the box for everyone, but enterprise technology must be fitted to a customer's specific use case, so engineers accompanying the model are genuinely helpful

Scope: contrast drawn specifically between consumer and enterprise technology

40:07 20VC: Cohere Founder on How Cohere Compete with OpenAI and Anthropic $BNs | Why Counties Should Fund Their Own Models & the Need for Model Sovereignty | How Sam Altman Has Done a Disservice to AI with Nick Frosst

Harry Stebbings · Jan 9, 2026

The biggest lie in enterprise software is that you can get implementation and adoption without FTEs; selling to the largest companies requires headcount and implementation

Scope: scoped to selling into the largest companies in the world

25:46 20Sales: Why the Best Sales People are F**** in the Head | Why Remote Work is BS | Why Europe is a Nightmare for Recruiting Sales Teams | How to Build a Sales Machine | What Everyone Gets Wrong Today in Sales with Chad Peets

Harry Stebbings · Apr 20, 2026

You cannot sell into the enterprise without a forward-deployed services model

29:29 20VC: Everyone is Wrong; We Will Have More Developers in Five Years | Why Frontier Labs Will Be Way More Valuable Than They Are Today | Are SaaS Companies Cooked: Which Thrive & Which Die with Aaron Levie, Founder at Box

Harry Stebbings · Apr 20, 2026 · hedged

The claim that enterprise AI sales require a forward-deployed model rounds to being true, which is interesting given the rise of PLG-then-upsell-into-enterprise motions

Scope: 'rounds to being true'

29:41 20VC: Everyone is Wrong; We Will Have More Developers in Five Years | Why Frontier Labs Will Be Way More Valuable Than They Are Today | Are SaaS Companies Cooked: Which Thrive & Which Die with Aaron Levie, Founder at Box

Harry Stebbings · May 23, 2026 · hedged

Invisible CEO Matt Fitzpatrick claims that selling to enterprise requires a forward deployed engineer model, a claim the speaker questions

17:36 20Sales: The $100M CRO Bubble: Why Anthropic Are Causing a Comp Crisis | Why You Should Never Hire From Salesforce or Service Now | How to Hire, Train and Forecase in a World of AI with Chad Peets and Chris Degnan

Harry Stebbings · Jun 20, 2026

You cannot sell into enterprise without a fully fledged forward-deployed FTE model.

22:28 20VC: Why Remote Work is White Collar Fraud | Why Revenge and Patriotism are the Best Founder Traits | Two Questions Every Founder Needs to Ask | The Wild Story of Raising $1BN from Masa Son in an Hour Long Meeting with Ryan Peterson, Founder @ Flexport

Clay Bavor · Jul 4, 2026

Selling and implementing into the large enterprise still requires substantial human-led work — understanding business outcomes, integrating with bespoke tech stacks, and earning trust as a partner rather than a vendor

Large enterprises are snowflakes, among the most complex and regulated organizations in the world, so successful deployment depends on deep customer understanding and relationships

Scope: specific to companies serving large regulated enterprises

19:04 20VC: Open Models vs Frontier Models: Who Actually Wins? | The $100,000 Token Budget Every Engineer Will Need | Why Forward-Deployed Engineers Are the Future of Enterprise AI with Clay Bavor, Co-Founder of Sierra

Chad Peets · May 23, 2026 · hedged

There is something to the view that forward deployed engineers compensate for a weak product, though they can add real value in some situations

Big companies send hordes of FDEs into accounts to build the product they don't actually have and then call it all one product

Scope: It depends; FDEs add a lot of value in some worlds; You can get carried away with it

17:53 20Sales: The $100M CRO Bubble: Why Anthropic Are Causing a Comp Crisis | Why You Should Never Hire From Salesforce or Service Now | How to Hire, Train and Forecase in a World of AI with Chad Peets and Chris Degnan

Matan Grinberg · Jun 13, 2026

You can sell to enterprises today without a forward-deployed engineer model; needing FDEs to close a deal means your product is bad

The legitimate purpose of FDEs is acceleration — getting a customer to consume faster than they would on their own — not making the product function; using them as services makes you a services company like Accenture, not a product company

Scope: FDEs are fine when used purely to accelerate an account's ramp

69:18 20VC: Who Wins the Model War: OpenAI, Anthropic or Open-Source | Token Maxing, AI Hangovers & The Coming ROI Reckoning | Labour Displacement Fears are BS & Overblown | From Physicist to Sequoia Founder with Matan Grinberg, Founder @ Factory

Harry Stebbings · Jun 22, 2026

If your product requires FTEs to deploy in an enterprise, you have a bad product.

42:41 20VC: Nikesh Arora on the Frontier Model Problem: Breadth vs Depth | The Future of Token Costs | Memory Becoming the Moat | Where Value Accrues: Infra, Models, or Apps? | Why Enterprise AI is Not Ready & Systems of Record vs Systems of Intelligence

Chris Degnan · Oct 10, 2025 · hedged

It's better to productize the learnings from forward-deployed work so the product can be sold by presales engineers and implemented by post-sales services

If implementation needs a ton of people and that functionality isn't going into the product, you're a professional services company

Scope: acknowledges Palantir as a counterexample; stated as his preference for companies he advises

46:57 20Sales: Scaling Snowflake from $0-$3BN in ARR | Snowflake vs Databricks: My Biggest Lessons | Why Customer Success is BS and What Replaces It with Chris Chris Degnan

Clay Bavor · Jul 4, 2026

Sierra will reach far greater customer scale than Palantir's low-hundreds of customers, by accumulating deep domain expertise in specific industry verticals that makes deployments repeatable

The first deployments were by definition one-of-a-kind, but they have learned generalizable lessons — how to build a basket in retail, how to handle a healthcare claim status question, how to handle a checking account fee question — that can be applied at scale

Scope: Palantir customer count given as 'my understanding'

35:48 20VC: Open Models vs Frontier Models: Who Actually Wins? | The $100,000 Token Budget Every Engineer Will Need | Why Forward-Deployed Engineers Are the Future of Enterprise AI with Clay Bavor, Co-Founder of Sierra

Ron Gabrisko · Aug 4, 2025

Bundling a free technical resource with a small committed spend is what unlocks early enterprise adoption

The product was great but the binding constraint was a gap in customer talent and expertise; once a person set it up and delivered the first use cases, customers got it

Scope: Databricks' 'strategic customer program' as the example

16:20 20Sales: $0-$3.7BN: The Databricks CRO's Playbook to Build the Fastest GTM Engine in SaaS History | How Databricks Beat Snowflake | How To Build a Sales Org of 5,000 and Close $190M Deals with Ron Gabrisko

Clay Bavor · Jul 4, 2026

Because no one had ever deployed a customer-facing AI agent before, embedding engineers deeply inside design-partner companies was the only way to build the right product quickly

Understanding the business mechanics, people and business model at nearly the customer's own level was necessary; their founding engineer was effectively an employee of Weight Watchers, down to receiving performance review emails

Scope: describes the 2023-24 founding period

30:43 20VC: Open Models vs Frontier Models: Who Actually Wins? | The $100,000 Token Budget Every Engineer Will Need | Why Forward-Deployed Engineers Are the Future of Enterprise AI with Clay Bavor, Co-Founder of Sierra

Chris Degnan · May 23, 2026

The forward deployed engineer is a glorified professional services person

0:19 20Sales: The $100M CRO Bubble: Why Anthropic Are Causing a Comp Crisis | Why You Should Never Hire From Salesforce or Service Now | How to Hire, Train and Forecase in a World of AI with Chad Peets and Chris Degnan

Chris Degnan · May 23, 2026

A forward deployed engineer is a glorified professional services person and is definitionally not as strong as a core product engineer

Really good engineers want to work on the core product, not in the field; field-built product may never make it back into core, leaving the customer holding technical debt they must support

18:19 20Sales: The $100M CRO Bubble: Why Anthropic Are Causing a Comp Crisis | Why You Should Never Hire From Salesforce or Service Now | How to Hire, Train and Forecase in a World of AI with Chad Peets and Chris Degnan

Winston Weinberg · Jan 19, 2026

Enterprise AI companies should weight their go-to-market toward post-sales rather than pre-sales, following Microsoft's evolution

If you're bullish on AI you should be bullish on your own product's future value, so retaining a customer who pays $1M today matters more than landing new logos, because that customer could pay $100M later

Scope: acknowledges there is some land grab dynamic right now; we are at day one of product development

35:13 20VC: How Model Performance is Plateauing | Two Key Rules for Effective Deal-Making | Company Building Lessons from Keith Rabois, Brian Halligan and Pat Grady | Why Enterprise AI Adoption is Years Off with Harvey CEO Winston Weinberg

Brendan Foody · Jun 1, 2026

Defensibility comes from the forward deployed post-sales motion rather than from pre-sales go-to-market

Even the best salesforce loses pricing power once a savvy customer paying $1M a year realizes they can tell Claude to copy the product, whereas going deep with a customer and training agents on their tacit knowledge is very hard to recreate — which is why OpenAI and Anthropic are investing so heavily in forward deployed teams

Scope: partly agrees that go-to-market matters

39:34 20VC: Mercor CEO on Why Application Layer Companies Have No Defensibility, The Model is the Product | Token Spend Will Exceed Headcount Spend in 5 Years | The True Cost of Hiring AI Researchers in the Valley Today with Brendan Foody

Jesse Zhang · Sep 19, 2025

Forward deployed engineering is overhyped and doesn't fit every use case; when customer needs are consistent across accounts, a product-first approach beats it.

In their space the things different customers care about are pretty consistent, so building product scales faster and lets customers self-serve rather than messaging Decagon for everything.

Scope: they retain some forward deployed elements for large customers; founders' Palantir background drew them to the model

43:44 20VC: Why 90% of Founders Build Startups Wrong | Why AI Growth Rates are Sustainable & Remote Work is BS and the AI Talent War | Competing with Brett Taylor and Sierra: Who Wins the Customer Service War with Jesse Zhang, Decagon

Alexander Embiricos · Feb 21, 2026

You do not need forward-deployed engineers to adopt AI in the enterprise; the better path is shipping open-ended tools that individual employees use directly

FTE-driven workflow automation is limited by what leadership can envision top-down and what you can staff; the exciting future is everyone feeling superhuman via tools they are fluent with

11:33 20VC: Codex vs Claude Code vs Cursor: Who Wins, Who Loses | Will All Coding Be Automated - Do We Need PMs | The Real Bottleneck to AGI | The Three Phases of Agents and What You Need to Know with Alex Embiricos, Head of Codex at OpenAI

Alexandr Wang · Jun 12, 2024

Palantir was the first tech company to pursue deeply customized enterprise software, and was unfairly dismissed as a consulting company

Palantir's bet was that going into enterprises to build the perfect application on top of all their data would be far more valuable than what any other software provider could produce — and it did this before generative AI made the motion feasible

26:16 Palantir pioneered valuable customized enterprise software unfairly dismissed as consulting

Max Junestrand · Aug 15, 2025

Staffing customer success and post-sales at roughly twice the size of new sales, with forward-deployed lawyers, is the right approach in legal AI even though it is unstandard for SaaS

Customers start with no idea what AI can do for them and need hands-on help to reach concrete wins like doubling turnaround speed at the same quality

36:53 Post sales staffing must outweigh new sales with forward deployed experts in professional services ai

Max Junestrand · Jan 26, 2026

A forward-deployed model is necessary when selling to enterprises whose way of working you are changing, but unnecessary for simple process automation like contract redlining in a CLM system

Equipping an entire firm across practice areas, offices and seniority levels with AI proficiency is an enormous change management undertaking that requires heavy upfront manual investment, whereas send-a-document-get-redlines workflows can just be deployed

8:56 Fde needed only when the product changes how people work

Ryan Petersen · Jun 20, 2026

Enterprise sales doesn't require an embedded FTE model, but embedding staff inside a customer's workflow is an enormously powerful competitive advantage — a competitor with 130 embedded logisticians at a hyperscaler runs the RFP and always wins it.

When all decision-making runs through your embedded people, rivals can't get a foot in the door; but good FTEs who show up, perform and don't churn are hard to find.

22:38 Embedding is optional but becomes an account lock in moat

Chad Peets · Jan 9, 2026

Selling APIs requires forward deployed engineers because you are shipping a foundation for customers to build on, not an out-of-the-box product, and most customers lack the internal technical chops to build with it

APIs give customers a toolset rather than a finished product, so vendors must supply engineers to help them build

25:57 Api products require forward deployed engineers

Patrick Forquer · May 11, 2026

Implementation change management is the biggest capability the speaker learned at Braze and carried over to Legora, where the focus is on helping customers adopt the tools into their workflows rather than on the product alone

Even with a great product, the hard part is getting customers to adopt the tools into their existing workflows; at Braze consolidating multiple legacy tools took six to twelve months of stakeholder management and technical workstreams

4:54 Change management capability not product decides adoption

Patrick Forquer · May 11, 2026

Forward-deployed humans are genuinely required for agentic products in a way they weren't for SaaS, because an agent presents a blank page rather than a sequential click-path

SaaS onboarding could say 'you used to do it this way, now do it this way', but with an agent the user must specify what they want it to do — and most businesses don't decompose their work into goals, steps, tools and decision points, nor write it down

5:57 Agentic blank page makes forward deployed humans newly necessary

Patrick Forquer · May 11, 2026

The forward-deployed model is worth its cost for any deal at six figures ACV and above

Without that hands-on investment you risk selling software that never gets used; if you sell ten licenses you need all ten adopted and delivering value

7:50 Fde cost is justified above six figure acv

Jesse Zhang · Sep 19, 2025

Sierra's Salesforce-derived leadership pushes it toward a heavy-configuration, long-implementation approach, while Decagon's product-forward approach appeals to customers who want to own and iterate on the system themselves

Their leadership all came from Salesforce and carries that playbook

52:19 Leadership pedigree explains configuration heavy vs product forward strategy choice