Yongping becomes visible in the places buyers already are — in person and in the feed.
- 12 industry events, Aug–Feb
- Panel or speaking seat at two Tier 1 events
- Weekly LinkedIn activity on his own profile
- Commenting on his network's posts, not just his own
Turn what he already knows into reusable source material. No crew, no studio.
- One photo per event
- One 90-second phone video answering a real question
- Quarterly interview sessions to bank article material
- Questions logged from every call and event
One capture becomes four assets across three voices.
- 6 expert articles added to existing services pages
- Video on YouTube, embedded on the same page
- Posts from Yongping, the Noble page, and BioBuzz
- FAQ and video schema so AI assistants can cite it
Every asset routes to the same place: a study-design call with Dr. Chen.
- Noble newsletter plus his personal send, per article
- Four-part nurture series ending in one ask
- Post-event follow-up sequence from his own inbox
- Warm reactivation sequence across his BD network
Yongping saying Noble is good is marketing. Chris Frew tagging him in a BioBuzz recap is a referral in front of the whole region. Both of the LinkedIn posts we already have show 77 and 258 reactions on posts he happens to be standing in — the reach is there and free, it just isn't naming him.
Search and AI assistants both reward pages that answer a specific question in a specific voice, from a named person with credentials. Noble has scientists who can do that from direct experience, and a competitor's marketing agency cannot fake it. Adding those answers to pages that already rank is the cheapest way to get credit for expertise Noble already has.Video is hosted on YouTube rather than self-hosted because YouTube is the most-cited domain in Google's AI Overviews. That is a hosting decision, not a channel strategy — see the note in section 04.
A sequence from Yongping's own inbox, referencing an article he wrote and an event they both attended, is a fundamentally different email from a Noble marketing blast. That is the entire argument for building this around a person rather than the brand.
Noble Life Sciences
- 17 years in DC–Baltimore preclinical CRO work
- Director of Operations, Washington Biotechnology (5.5 yrs) — ran the shop, not just the science
- Scientific Advisory Board, Lahjavida
- Postdoctoral fellow, Johns Hopkins School of Medicine
- Ph.D., University of Science and Technology of China; Clinical Medicine, Anhui Medical University
Twelve events, three tiers, one content deliverable each
Every event below produces exactly one photo, one 90-second video, and one LinkedIn post. Tier decides how much preparation goes in beforehand — Tier 1 events get a pre-written question for Yongping to ask from the floor and a target list of three people to meet.
One ask that changes the economics. Before September, Noble should ask BioBuzz for a speaking or panel slot at two of the Tier 1 events rather than buying tickets to all twelve. Yongping on a panel produces a photo, a video, a quote, and a mention in the recap post — four assets from one evening. Yongping in the audience produces one photo.
Six posts, three voices
The BioBuzz posts are drafts written for Chris Frew — they only exist if he agrees to name Yongping, so the ask is part of the plan, not an assumption. Photos below are real, from events Yongping has already attended.
The posting order matters more than the copy. Yongping posts first. Noble's company page reshares 24 hours later, not the same day — a same-day company reshare tells LinkedIn's ranking system the post is corporate and suppresses its reach in personal feeds. Chris Frew's recap runs whenever he runs it; we never ask for timing, only for the mention.
A repeatable article format, built the way AI assistants cite
Every article in the programme is built to the same template, and the mockup below is the first of them worked through in full so the pattern is visible. Each one is added to an existing noblelifesci.com services page following the approach we built for Pace: exact-question headings, a two-sentence answer above the fold of each section, a visible video embed, named expert attribution, and FAQPage plus VideoObject schema. Six are scheduled through February, and the format keeps producing after that for as long as Dr. Chen keeps getting asked new questions.
Do not build one page per question. Google names that as scaled content abuse. All ten answers live on one page, inside one FAQ block.
Add to a page that already ranks rather than creating a new URL. Pages already in Google's top ten supply the largest single share of AI Overview citations.
Text above video. AI systems cite text, not footage. The video earns the YouTube presence; the paragraph above it earns the citation.
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [{
"@type": "Question",
"name": "How long does an IND-enabling package actually take?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Six to nine months for most small
molecules, measured from the first dose-range-
finding study to the final signed report..."
}
}]
}
Plus a VideoObject block for the embed, carrying name, description, uploadDate, duration, thumbnailUrl and contentUrl. Both go in the page head; no visitor sees them, and they are what makes the answers eligible to be lifted individually.
- Top 5 mistakes that get a preclinical study repeated — Oct, pegged to Pharma Clinical Innovation
- GLP or non-GLP: which of your studies actually need it — Nov, cost-driven, high commercial intent
- Choosing an animal model: 10 answers from the preclinical bench — Dec, Yongping's deepest technical ground
- What a first-time sponsor should ask on a CRO site visit — Jan, comparison-intent, links to facilities page
- Cell and gene therapy preclinical: what is different and why it costs more — Feb, pegged to CONNECTpreneur
Video here is a page asset. Each clip exists to make the article section better and is hosted on YouTube because YouTube is the most-cited domain in Google's AI Overviews — a hosting decision, not a channel strategy. This plan does not include YouTube channel management: no titling convention, upload cadence, transcript standard, or playlist structure.
That is a deliberate omission, and worth naming because the evidence for video is strong. Across 75,000 brands, Ahrefs found YouTube mentions correlated with AI-assistant visibility more strongly than any other factor measured — 0.737, against 0.664 for branded web mentions and 0.266 for domain authority. Note the wording: that measures how often a brand comes up across YouTube generally, including in other people's videos, which is a different thing from running a channel. If Noble wants to pursue it properly, it is a separate scope and should be resourced as one.Ahrefs, Top Brand Visibility Factors in ChatGPT, AI Mode and AI Overviews, December 2025. Correlation, not causation.
Every number in the ten answers is a draft for Dr. Chen's review. The figures above are written from general preclinical practice to show the format and the voice. Nothing publishes until he has confirmed each one against what Noble actually delivers, and any claim we cannot stand behind gets cut rather than softened.
Two senders, two completely different emails
The same article goes out twice. Noble's version is a designed newsletter to the full list. Yongping's version is plain text from his own address to a segment of a few hundred people who know him — and it is the one that produces replies.
Article 1. Establishes Yongping as the person answering, not the company. Pure value, single link, no CTA beyond reading.
The event video, embedded. Same answer in his voice and his face. Recipients who ignored the article often watch the clip.
First commercial content: the capability set, framed as “here is what the answers above look like as a study.” AAALAC, GLP, both species in one building.
The only ask in the series. From Yongping, offering to review a study design free. Anyone who books moves to the sales process; anyone who doesn't stays on the newsletter.
Two sequences from Yongping's inbox
Sequences are one-to-one: they send from Yongping's connected inbox, enroll one contact at a time, and unenroll automatically the moment someone replies or books. That is the right tool here — a marketing email blast from a personal address is neither, and would burn the relationship.
Both sequences below are drafts of the structure. Every line is customisable to Dr. Chen — his phrasing, his level of formality, how directly he asks, which articles each step links to, and the timing between steps. We would sit with him for an hour, rewrite the copy in his own words, and adjust the cadence to whatever he can realistically keep up with. A sequence that does not sound like the person sending it is worse than no sequence at all.
Setup requirements. Yongping's inbox must be connected to HubSpot, his meetings link created, and both sequences built under his user so sends come from him rather than from a shared address. Sequences count against a per-user daily send limit, which is why the reactivation sequence is capped at 25 enrollments a week — that cap also happens to be the number of replies one person can actually answer well.
What we count, and what we need from Noble
Everything below it is a leading indicator. If impressions, followers, and open rates all rise and this does not move by month six, the campaign is not working and we change it.
Reported monthly, judged quarterly. Preclinical buying cycles are long and a founder who reads Article 1 in September may not need a CRO until January. Reviewing pipeline monthly would kill a working campaign in month two. We report the leading indicators every month, and make continue-or-change decisions at the end of month three and month six.
- Sign-off on the twelve events and the travel budget for the two out-of-state trips
- Which services page Article 1 gets added to — you said you would pick it
- Ninety minutes with Dr. Chen to record the first three videos and fact-check the ten answers
- His inbox connected to HubSpot, and his BD contacts imported and tagged
- An approach to Chris Frew about a panel seat at two Tier 1 events
- The SEMrush prompt research export, to confirm the order of the six articles