Influence Β· Milestone Companion: Translation Is the Job
What this milestone asks
Deliver a public talk of thirty minutes or more β online or offline β fully recorded, and organize the Q&A into a public article afterward. The stage behind it: enterprise evangelism is translation, not persuasion; prepare three demos (code for engineers, ROI for management, experience for end users); keep a case library recording background β approach β outcome β failure modes; and protect credibility above all β never promise what you cannot deliver, never relay what you have not verified. Thirty minutes is the threshold where structure, stamina, and audience management all become real; the Q&A article is where the talk's value doubles.
Why this milestone
Stages 1β3 built the inputs β verified knowledge, craft, a publishing practice. Stage 4 converts them into influence, whose currency in enterprise settings is translation. Engineers want code and failure modes. Management wants cost and efficiency language: what does this replace, what does it cost per transaction, what breaks at scale. End users want to touch the thing and feel the difference. The same capability must survive all three translations without distortion β that is the entire job, and it is why the three-demo discipline exists: a code walkthrough that thrills engineers reads as risk to a CFO, and an ROI slide that persuades a CFO reads as vapor to an engineer.
The recording is not vanity; it is your case library's raw material. And the Q&A article outlasts the talk: a published "everything the audience asked, answered with sources" piece gets forwarded into companies you will never visit, and proves what enterprise buyers most want to see β that you answer hard questions without hand-waving.
How to use the resources
Study stagecraft at the highest level. David Malan's CS50 lectures remain the benchmark for pacing and stage control β watch how he handles a large room, deploys silence, and recovers from misbehaving demos. Chris Anderson's TED Talks: The Official TED Guide to Public Speaking covers the thirty-minute format directly: one through-line, stories as load-bearing structure, ending early rather than late.
Watch how frontier presenters handle hostile depth. Stanford CS25 Transformers United β rotating top-tier researchers β shows how experts address mixed expert audiences: dense but signposted, humble about limits, precise about sources. That register, softened, is what an enterprise technical audience respects. Mu Li's paper structure (background β method β experiments β impact) maps directly onto the enterprise talk: context β approach β measured outcome β what it means for you.
Build the case library before you need it. Every project from Stages 1β3 becomes a case entry: background, approach, outcome, and β most valuable β failure modes. Failure modes earn the most trust in a skeptical room; anyone can recite successes. Ground field-level claims in the Stanford HAI AI Index report. When enterprises ask how adoption spreads, cite the Berkeley RDI LLM Agents MOOC as a case study in organizing a massive agent-education operation: guest lineup, assignments, certificates, community.
Prepare the three demos as separate artifacts. Code demo: a walkthrough of a real agent run with visible traces. ROI demo: one spreadsheet, one before/after workflow, conservative numbers. Experience demo: something an end user operates themselves in ninety seconds. The Stage 2 live-demo iron rules still bind: backup recordings for everything, nothing shown that has not succeeded three times.
Engineer the Q&A. Plant nothing, prepare everything: draft answers for the ten questions your Stage 3 FAQ says are most likely. After the talk, transcribe the actual Q&A, answer each question in writing with sources from your fact ledger, and publish within a week while the audience still remembers asking.
Find the stage. InfoQ and QCon are the established editorial and conference channels; local meetups are where you earn submission to them. A thirty-minute meetup slot with a recording fully satisfies this milestone β do not wait for a keynote invitation that is not coming yet.
Common failure modes
- One demo for all three audiences. The most common enterprise-evangelism failure. The fix is not a better demo; it is three different ones.
- Promising the roadmap. "It will do that soon" is a promise you cannot deliver and a claim you have not verified β a double violation. Say "not yet, and here is what currently breaks."
- Relaying unverified claims because a respected source said them. Your credibility asset is precisely the Stage 1 verification habit; do not spend it on hearsay.
- Treating Q&A as defense. The hardest question is your next article's headline. Evangelists who deflect lose the room; ones who say "I don't know, here's how I'd find out" win it.
- Letting the recording die in a folder. Unpublished recordings compound nothing.
Acceptance checklist
- A public talk of 30+ minutes was delivered, online or offline, with date, venue/platform, and audience size recorded.
- A complete recording exists and is accessible (publicly, or internally with permission noted).
- Three distinct demos (code / ROI / experience) were prepared; the audience-appropriate one was shown under the live-demo iron rules.
- The talk had one through-line the audience could state back β verified by asking at least two attendees.
- The case library holds at least three entries in background β approach β outcome β failure modes form.
- A Q&A article covering every question asked is published, with sourced answers, within one week of the talk.
- No claim in the talk or article lacked a traceable source or an explicit "unverified" label.
Going further
The Q&A article ends the milestone but starts an asset class: every future talk reuses its questions; every enterprise conversation reuses the case library. Natural next steps: a recurring meetup slot, a course distilled from your Stage 3 FAQ, and a standing page aggregating recordings, articles, and methodology β your public credibility dossier. Influence stops being something you pursue and becomes something your archive does on your behalf. Keep the ledger honest, keep the cadence, and the archive compounds.