FDE Capture-vs-Create: Is There Latent Search Demand for "Forward Deployed" Data Services?
The question
"FDE capture-vs-create demand: is there existing search demand for 'forward deployed' data services?" — feeds the founder's live decision on whether the planned Sanity Check FDE term-claim piece is a capture-existing-demand SEO play or a create-the-category move (different content strategy, different success metric).
What we already know (from the vault)
- The 2026-05-29 baseline found a branded owner exists for the generic term (Palantir) but NO branded owner for "FDE for data teams"; it explicitly deferred quantifying search demand as an open follow-up — this brief is that follow-up. [[2026-05-29-fde-search-share-baseline]]
- The fractional-FDE whitespace check concluded the fractional + data-team + FDE-branded combination is open, but flagged the central risk plainly: whitespace may exist because "the category may not yet have buyer awareness." [[2026-05-28-fractional-fde-service-whitespace-check]]
- The Slotnick "FDE wars" note is the sharpest signal on WHO is using the term. Read directly: "the 'FDE wars' is operator/VC-side discourse... NOT buyers searching for FDE services," and data-team buyers "may not use the term yet" — supply-side salience, unproven demand-side salience. It also prescribes selling outcomes via the embedded model rather than leading with "FDE" branding to buyers who don't search it. [[2026-05-29-slotnick-cardinal-sin-platform-building-fde-wars]]
- The Utsubo read (recovered via QMD index) confirms even the closest "forward deployed" branded competitor targets bespoke AI app builds, not data teams — the data-qualified angle stays unclaimed. [[2026-05-29-utsubo-forward-deployed-studio-competitive-read]]
- The convergence thesis names RDCO's wedge explicitly: the "data-qualified FDE" — embedded delivery for data teams specifically — as the forward-deployed model converges across AI labs, big consultancies (Deloitte FDE), and boutique data shops. [[2026-05-13-fde-wave-convergence-rdco-thesis]]
- Note on vault state: several FDE source files in this thread are 0-byte (baseline, whitespace-check, Utsubo, pricing-framing) or placeholder stubs (convergence-thesis) on disk; their substance was recovered from the QMD index, which retains indexed content. Slotnick read in full from disk. Flagging rehydration as a hygiene follow-up.
What the web says
- I could NOT retrieve actual buyer search-volume numbers. Google Trends and Google Search both rejected automated fetches (HTTP 429), and no source surfaced a quantified Google Trends index or keyword-tool volume for "forward deployed engineer" as a BUYER query. Any number stated as buyer search volume here would be fabricated; there is none. (Direct measurement attempt, failed.)
- The strongest measured signal is JOB-POSTING growth, which is a supply-side proxy (employers hiring FDEs), NOT buyer search demand. Indeed data cited by The New Stack: FDE postings rose ~729% YoY, from 643 (Apr 2025) to >5,300 (Apr 2026). Bloomberry/Paraform cite ~1,165% YoY (Jan-Oct 2025) and "demand grew 10x in 18 months." (WebSearch — measured, but proxy for talent-market heat, explicitly NOT buyer search demand.)
- A revealing supply/demand mismatch, again on the HIRING side: postings grew ~800% (Jan-Sep 2025) while the candidate pool grew only ~50% (Bloomberry). This is a labor-market gap, not a buyer-discovery gap — it tells us employers want FDEs, not that data-team buyers are searching to hire an FDE SERVICE.
- Adoption is seller-driven: AI labs (OpenAI, Anthropic, Cohere), Salesforce (committed to 1,000 FDEs on Agentforce/Data Cloud), Deloitte (launched "Deloitte Forward Deployed Engineering"), and OpenAI's "The Deployment Company" are all adopting FDE as a GTM wedge. The pull is companies wanting to HIRE/BE FDEs, not data teams searching to BUY one. (WebSearch — proxy: GTM/operator discourse.)
- a16z is reported to have dubbed FDE "the hottest job in tech" — framed as a job/talent and GTM-strategy phenomenon, i.e. seller-side. I could not fetch the canonical a16z piece (404 on the guessed URL), so this is via search-result summaries only, not a primary read. (Label: secondhand, not primary-source verified.)
- Searches for the specific BUYER-intent combination ("fractional FDE for data teams," "applied AI services" buyer demand) returned essentially nothing on point — the buyer-side phrasing has near-zero web footprint (most hits were unrelated math/"fractional differential equations" noise). (WebSearch — proxy: absence-of-evidence, consistent with a term buyers don't yet type.)
- Net of proxies: job-posting/comp heat = high and MEASURED (729-1,165% YoY); operator/VC discourse heat = high (the "FDE wars" framing); buyer-side search/term-usage = unmeasured and apparently thin. Every retrievable proxy points the same direction — the heat is on the supply side.
Convergences and contradictions
- Strong convergence: vault and web agree the term is HOT on the supply side (talent + founders/VCs) and unproven-to-cold on the buyer/demand side. Slotnick's "operator/VC-side discourse, not buyers searching" maps exactly onto the web's measured 729-1,165% job-posting growth with near-zero buyer-intent footprint.
- No contradiction surfaced — the web did not reveal any hidden pool of buyer-side demand the vault missed; if anything the measured hiring numbers sharpen the vault's qualitative read.
- Honest gap, not a contradiction: nobody (vault or web) has the actual buyer search-volume number. The conclusion rests on the consistent DIRECTION of every available proxy, not on a measured buyer-query figure.
Synthesis for RDCO
The load-bearing call: on current evidence this is a create-the-category move, not a capture-existing-demand SEO play. The distinction the founder is deciding between resolves cleanly because every retrievable signal points to supply-side heat with no measurable buyer-side search demand. The term "forward deployed engineer" / "FDE" is salient among the people who want to hire FDEs or build FDE-led startups — it is not (yet) the phrase a data-team buyer types into a search bar when they have a pipeline problem and a budget. A capture play only works when buyers are already searching the term; the proxies say they are not, and the buyer-intent phrasings return near-empty. Treating the Sanity Check piece as an SEO capture play would optimize for a query that does not have buyer volume behind it.
What that implies for the piece's framing: write it as category-creation / demand-education, not as a keyword-targeted SEO landing page. The job of the piece is to teach a data-team buyer that "the embedded, forward-deployed delivery model" is the thing they actually want — to attach RDCO's name to the concept BEFORE buyers have a search habit for it, so RDCO is the brand they recall once the habit forms. This is a planting-a-flag move (consistent with the baseline's "the data-qualified term is open"). It rides a real and rising supply-side wave, which is the legitimate reason to move now rather than wait: the term is accruing salience fast, and being the first data-specific claimant is cheaper today than after an incumbent claims it.
The success metric must change accordingly. Do NOT measure this piece by organic search traffic or keyword ranking in the first 1-3 months — that metric assumes a capture play and will read as failure even if the piece succeeds at its actual job. Measure it instead by: (1) category-attachment signals — inbound/referral mentions, operator/VC reshares, citations that pair "RDCO" with "forward deployed for data"; (2) brand-recall and direct/branded-search lift over a longer horizon (6-12 months) as the term matures; (3) qualified-conversation starts attributable to the piece. SEO/organic capture becomes a LATER, lagging metric once (if) buyer search demand materializes — worth instrumenting from day one as a leading indicator of the category tipping from create to capture, but not the launch scorecard.
Evidence-strength caveat, stated plainly: this recommendation is directionally well-supported but rests on proxies, not a measured search-volume figure. The single highest-value thing that would harden or overturn it is one real Google Trends / keyword-tool reading on "forward deployed engineer" and the data-qualified variants. If that reading came back showing meaningful and rising buyer-side volume, the call would shift toward a hybrid (create-led, capture-instrumented). On what is retrievable today, create-the-category is the right default.
Open follow-ups
- Pull an actual Google Trends curve for "forward deployed engineer" + "FDE" + "forward deployed for data" via a browser-automation path (the plain-fetch route is 429-blocked) to convert this from proxy-based to measured.
- Get one keyword-tool reading (volume + difficulty) on the data-qualified variants to set a realistic capture-vs-create threshold.
- Rehydrate the six 0-byte FDE stub files in 06-reference/ — content exists in the QMD index but not on disk.
- After the Sanity Check piece ships (if it ships), instrument branded/direct-search lift as the leading indicator of the category tipping from create toward capture.
- Decide whether to claim the generic term, the data-qualified term, or both — the baseline left this open and it changes the piece's headline.
Sources
- [[2026-05-29-fde-search-share-baseline]]
- [[2026-05-28-fractional-fde-service-whitespace-check]]
- [[2026-05-29-slotnick-cardinal-sin-platform-building-fde-wars]]
- [[2026-05-29-utsubo-forward-deployed-studio-competitive-read]]
- [[2026-05-13-fde-wave-convergence-rdco-thesis]]
- WebSearch result sets (2026-05-30): job-posting growth figures (729% YoY Indeed; 1,165% YoY; 10x/18mo; 800% vs 50% candidate-pool) sourced from The New Stack (thenewstack.io/forward-deployed-engineer-fde-openai-google/), Bloomberry (bloomberry.com/blog/i-analyzed-1000-forward-deployed-engineer-jobs-what-i-learned/), Paraform (paraform.com/blog/forward-deployed-engineer-demand-quadrupled), Pragmatic Engineer (newsletter.pragmaticengineer.com/p/forward-deployed-engineers). All supply-side (hiring) — no buyer search-volume figures published.
- Buyer-intent / a16z-framing context: Fast Company, IT Brew, Pave — secondhand summaries, a16z primary piece not directly fetched.
- Failed/blocked fetches (logged for honesty): google.com/search and trends.google.com (HTTP 429, automated-fetch blocked); a16z.com/forward-deployed-engineer (404, guessed URL not verified) — no numbers taken from these.