The equity-kicker question is built on two premises that do not hold, and the real gate is a clause nobody has read
The question
Verbatim: "Does the equity-heavy FDE comp norm open a retainer-plus-equity-kicker pricing shape for early-stage data-team clients, and does that conflict with the phData W-2 / single-client-concentration guardrails?"
Follow-up from the 2026-05-27 FDE-pricing brief, carried forward unresolved through three subsequent briefs. Context: the founder is W-2 at phData (started 2026-05-26, DSA + Technical Account Lead), Florida-based; RDCO is the side entity.
Answer first
No, and do not spend more on this shape. Three findings, in descending order of how much they should change behavior:
The "phData W-2 / single-client-concentration guardrail" has no source. It is a citation loop. The phrase originates in the 2026-05-27 brief's open-follow-up section and is re-cited as an established constraint by [[2026-05-31-fde-scoping-pricing-vs-ai-consultant-framing]], [[2026-06-02-fde-retainer-band-pricing]], and [[2026-06-07-solo-fde-contract-structures]] without any of them ever sourcing it. A vault-wide grep finds no phData document, offer letter, or note establishing such a rule. The only real "single-client concentration" discussion in the vault is [[2026-04-15-commoncog-career-moats-confession]], where it describes Mammoth Growth at ~60% of RDCO revenue as a risk of the 1099 path. Taking the phData W-2 was the resolution of that risk, not the source of it. The compound phrase fused two unrelated things and then hardened into fact across four briefs.
The FDE comp norm provides no "cover" for a vendor equity kicker. The equity-heavy finding is real but is a fact about employment compensation. The report that establishes it explicitly does not cover contractor or consulting engagements (see below). A client's willingness to grant equity to an employee they hire says nothing about a vendor's pricing shape, and the "FDE buyers already expect equity in the picture" line from the parent brief is an inference the underlying data does not support.
The binding constraint is a different, unread clause, and it gates far more than the kicker. [[2026-04-16-client-reporting-automation-one-pager]] (line 102) has carried this open item since April: "Moonlighting clause check — phData's employment agreement may restrict outside consulting work. Need to read it before any client pitch goes out." That gate blocks all paid RDCO client work, cash or equity. Debating the kicker while the base retainer is ungated is optimizing the wrong layer.
Recommendation: drop the equity-kicker thread. Keep the standing cash-only shape from [[2026-06-07-solo-fde-contract-structures]]. Convert the four propagated "check the guardrail" follow-ups into one founder action: read five named clauses (listed below). That is a document read, not a research task.
What we already know (from the vault)
- The standing pricing position is cash-only and already settled. Paid fixed-fee pilot ($2K-$5K, 100% credited into the retainer) → 90-day-minimum retainer at ~$15K/mo, milestone-staged. Source: [[2026-06-07-solo-fde-contract-structures]], reaffirmed for the reporting line at $2-5k/mo in [[2026-06-13-managed-ai-data-services-pricing-models]].
- The vault has never taken a position on client equity. Every mention is speculative and immediately paired with the unresolved guardrail flag. [[2026-06-07-solo-fde-contract-structures]] down-weights outcome-based pricing for a solo specifically: a missed-outcome quarter is "a margin dent" for a studio with a bench and "closer to existential" for a solo. Equity is strictly worse than a success fee on that axis, since it is also illiquid.
- The vault has never read phData's signed employment agreement. Every phData-side statement traces to the founder-articulated house rule in
feedback_employer_client_content_boundaryplus generic public norms. The one place the vault explicitly flags an unread contract is the Mammoth Growth 1099 agreement in [[2026-07-13-mg-consulting-contract-ip-ownership-methodology]], which states: "the specific terms are unverified; do not treat the boundary as settled until a human reads the executed... agreement." That discipline was never applied to phData. - The equity-norm claim entered the vault as "permission, not a requirement." [[2026-05-31-fde-scoping-pricing-vs-ai-consultant-framing]] correctly hedged it, then still passed the guardrail conflict downstream as a live constraint.
What the web says
- Equity is 55-70% of FDE total comp at the top of the market in 2026, up from 35-45% in 2024. By tier: frontier labs 60-70%, applied-AI startups 45-60%, classic Palantir 40-55%, Fortune 500 enterprise AI 15-25%. Source: Perspective AI, "The 2026 Forward Deployed Engineering Compensation Report", published 2026-05-21. The directional premise of the parent brief is confirmed.
- That report's methodology is mostly self-reported, and it is a vendor blog. n=1,200 comprises 423 Glassdoor submissions, 187 Levels.fyi entries, 312 disclosed pay bands from job postings, 178 Team Blind screenshots, and 100 self-reports from the publisher's own network. Only the 312 posting bands are independently verifiable. Treat the tier percentages as MEDIUM confidence, directional not precise.
- The same report explicitly does not address contractor or consulting equity. Fetched and confirmed in this run: it covers W-2 employment structures only. This is the load-bearing gap. The parent brief's leap from "FDEs are paid in equity" to "clients will read a vendor equity kicker as normal" has no support in its own source.
- FDE equity is largely illiquid where it is largest. Frontier-lab grants are RSUs on private stock, 4-year vest with a 1-year cliff, with liquidity "via roughly annual secondary tender offers, not public markets" (Perspective AI, ibid.). Palantir FDSE compensation runs $171K-$295K with a $211K median on public, immediately-liquid PLTR stock (Levels.fyi, Palantir FDSE). The illiquidity travels with the equity shape and would travel to an RDCO kicker too.
- Florida's CHOICE Act does not touch moonlighting, outside employment, conflict of interest, or equity ownership. Effective 2025-07-01 (HB 1219). It covers non-competes and garden leave only, up to 4 years, presumptively enforceable for a "covered employee" earning more than twice the annual mean wage of the relevant Florida county, with courts required to grant preliminary injunctions unless the employee rebuts by clear and convincing evidence. Agreements not meeting its requirements remain governed by Florida's preexisting framework (Fla. Stat. 542.335). Source: McGuireWoods client alert, July 2025, fetched in this run. Note a date discrepancy: DLA Piper reports the effective date as July 3, 2025. Not material here; flagged rather than smoothed.
- Equity-for-services mechanics carry a real, immediate tax and admin cost. Advisory shares to a non-employee service provider are reported as: NSOs taxed as ordinary income at exercise plus self-employment tax; RSAs eligible for an IRC §83(b) election that must be filed within 30 days of grant, a window described as irrevocable and incurable; §83(b) does not apply to unexercised options. Sources: Cake Equity 83(b) guide, The Startup Law Blog, Velawood. UNVERIFIED against primary text: irs.gov returned 404 on two attempted fetches in this run. These are secondary sources. Do not rely on the 30-day rule or the SE-tax treatment for an actual filing without CPA confirmation.
Convergences and contradictions
- Converges: vault and web agree that outcome-contingent compensation is bad structural fit for a solo with no bench. The web adds that equity is worse than a cash success fee, because it stacks illiquidity and a tax event on top of the attribution risk the vault already identified.
- Contradicts: the vault treats the FDE equity norm as market permission for a vendor kicker. The primary source for that norm covers employment only and says nothing about vendor pricing. The inference does not survive contact with its own citation.
- Neither settles the actual question. No public source can tell us what the founder's signed agreement says. Everything on the phData side of this brief is inference from public norms, and it stays that way until a human reads the document.
Synthesis for RDCO
The honest verdict is that this pricing shape is not worth pursuing, and the more valuable output of the question is the citation failure it exposed. Four briefs across five weeks treated "the phData W-2 single-client-concentration guardrail" as a known constraint. It never was. It was a question mark in an open-follow-ups section that got re-cited as a premise, which is exactly the laundering failure mode the founder catalogued in feedback_workflow_agent_output_integrity (status: proposal read as live fact). The chain is worth retiring explicitly rather than quietly, because three downstream briefs currently carry a constraint that does not exist.
On the merits, even granting a permissive employment agreement, the trade is poor for a side thread. An early-stage advisory-scale grant is small, illiquid, and years from any liquidity event, while the costs land immediately: a §83(b) decision on a 30-day fuse, a 409A-dependent valuation, annual K-1 or reporting overhead, and a standing ownership interest that persists after the engagement ends. Set that against RDCO's actual bottleneck, which [[2026-06-28-productized-consulting-scalable-anchor-transition]] names as founder attention rather than capital. A kicker consumes attention and adds zero throughput. The cash retainer is the right instrument and it is already specified.
The genuinely useful reframe concerns which clause bites. The instinct is to worry about moonlighting, but for equity specifically the sharper exposure is a conflict-of-interest clause. Cash for a one-off engagement is a transaction that closes. Equity converts it into an ongoing ownership position in a third party, and if that client is, becomes, or competes with a phData customer or prospect, that position is a textbook disclosable interest that will surface on any annual COI attestation. Equity does not merely inherit the moonlighting risk; it creates a distinct and longer-lived one. That is a reason to prefer cash independent of what the agreement turns out to say.
Where the answer depends on the signed agreement, it depends entirely. These are the five clauses to check, by their usual headings:
- Outside Activities / Outside Employment — is prior written approval required, or only non-interference with duties? Approval-required is the common posture at consultancies and would gate the retainer itself.
- Conflict of Interest — does it prohibit, or merely require disclosure of, a financial or ownership interest in a customer, vendor, partner, or competitor? This is the clause that bites hardest on equity.
- Proprietary Information and Inventions Assignment (PIIA) — scope test: does it assign inventions "relating to the Company's business" regardless of when or where conceived, or is there a carve-out for work on own time without company resources? A broad version puts RDCO methodology at risk, cash or equity.
- Non-solicitation of customers/clients — separate from non-compete and often survives where a non-compete would not.
- Corporate opportunity — whether business opportunities arising through the role must be presented to phData first.
Note that Florida law does not help here. The CHOICE Act governs non-competes and garden leave only; ordinary outside-activity and COI obligations are contractual, and the agreement's own text controls.
Why this is in the vault
It retires a fabricated constraint that four FDE-pricing briefs ([[2026-05-27-forward-deployed-engineer-pricing-rdco-framing]], [[2026-05-31-fde-scoping-pricing-vs-ai-consultant-framing]], [[2026-06-02-fde-retainer-band-pricing]], [[2026-06-07-solo-fde-contract-structures]]) currently treat as real, and it redirects the open item from a pricing debate to the concrete unblocking action for RDCO's entire paid-services line: reading five named clauses in the phData agreement, an item open since 2026-04-16.
Open follow-ups
- What the phData agreement actually says. Not researchable. Founder action: read the five clauses named above.
- Whether the four downstream briefs should be edited in place to strike the guardrail phrase, or annotated with a pointer here. Vault-hygiene decision, not research.
- Whether advisory-share percentage norms (commonly cited as 0.25-1%) hold for a data-infrastructure vendor rather than a named advisor. Not pursued, since the shape is not recommended.
Related
- [[2026-05-27-forward-deployed-engineer-pricing-rdco-framing]] — the parent brief that opened this question
- [[2026-06-07-solo-fde-contract-structures]] — the standing cash-only contract shape
- [[2026-05-31-fde-scoping-pricing-vs-ai-consultant-framing]] — carries the unsourced guardrail phrase
- [[2026-06-02-fde-retainer-band-pricing]] — carries the unsourced guardrail phrase
- [[2026-06-13-managed-ai-data-services-pricing-models]] — outcome-based as capped upside, never the backbone
- [[2026-07-13-mg-consulting-contract-ip-ownership-methodology]] — the unread-contract discipline this brief applies to phData
- [[2026-04-15-commoncog-career-moats-confession]] — the actual origin of "single-client concentration risk" (Mammoth Growth)
- [[2026-04-16-client-reporting-automation-one-pager]] — the moonlighting-clause check, open since April
- [[2026-06-28-productized-consulting-scalable-anchor-transition]] — founder attention as the binding constraint
- [[2026-06-23-fde-risk-reversal-offer-design]] — risk-reversal without contingent compensation
Sources
Vault
06-reference/research/2026-05-27-forward-deployed-engineer-pricing-rdco-framing.md06-reference/research/2026-05-31-fde-scoping-pricing-vs-ai-consultant-framing.md06-reference/research/2026-06-02-fde-retainer-band-pricing.md06-reference/research/2026-06-07-solo-fde-contract-structures.md06-reference/research/2026-06-13-managed-ai-data-services-pricing-models.md06-reference/research/2026-07-13-mg-consulting-contract-ip-ownership-methodology.md06-reference/research/2026-06-23-fde-risk-reversal-offer-design.md06-reference/research/2026-06-28-productized-consulting-scalable-anchor-transition.md06-reference/2026-04-15-commoncog-career-moats-confession.md01-projects/services-offering/2026-04-16-client-reporting-automation-one-pager.md01-projects/phdata/index.md- Memory:
feedback_employer_client_content_boundary,feedback_workflow_agent_output_integrity
Web (fetched and read in this run unless marked)
- Perspective AI, "The 2026 Forward Deployed Engineering Compensation Report: What 1,200 FDEs Earn," 2026-05-21 — https://getperspective.ai/blog/2026-forward-deployed-engineering-compensation-report-1200-fdes (fetched; vendor blog, mostly self-reported data)
- McGuireWoods, "Florida Bolsters Enforceability of Certain Noncompete Agreements," July 2025 — https://www.mcguirewoods.com/client-resources/alerts/2025/7/florida-bolsters-enforceability-of-certain-noncompete-agreements/ (fetched)
- Levels.fyi, Palantir Forward Deployed Software Engineer salary — https://www.levels.fyi/companies/palantir/salaries/software-engineer/title/fdse (via search summary, not fetched)
- DLA Piper, Florida CHOICE Act takes effect — https://knowledge.dlapiper.com/dlapiperknowledge/globalemploymentlatestdevelopments/2025/florida-CHOICE-act-expanding-employer-non-compete-options-takes-effect (via search summary, not fetched; effective-date discrepancy noted)
- Cake Equity, "83(b) Election, Explained" — https://www.cakeequity.com/guides/83b-election (via search summary, not fetched)
- The Startup Law Blog, equity compensation guide — https://www.thestartuplawblog.com/the-complete-guide-to-equity-compensation-for-startups/ (via search summary, not fetched)
- Velawood, "Understanding 83(b) Elections" — https://velawood.com/understanding-83b-elections-a-crucial-step-for-startup-founders-and-employees/ (via search summary, not fetched)
- UNVERIFIED: IRS primary guidance on IRC §83(b) could not be retrieved; irs.gov returned HTTP 404 on two attempted URLs in this run. All §83(b) mechanics above rest on secondary sources and require CPA confirmation before any filing.
Follow-ups
- In publicly-available US employment agreements at comparable data/AI consulting firms (SEC-filed exhibits, published handbooks), what is the modal text of the "outside activities" and "conflict of interest" clauses, and which formulations actually prohibit holding an ownership interest in a client versus merely requiring disclosure? This is researchable from public filings, is not covered by any existing brief, and gates RDCO's entire paid-services line rather than just the equity question.
(Second candidate withheld: an E&O / advisor-liability question was considered and cut, since it is downstream of a pricing shape this brief recommends dropping.)