Chapter 08 · Execution
Two of the five formats ship from assets already in the vault.
Ordered by dependency, not ambition. The first two weeks are unblocking and baselining; nothing new is posted until the landing surfaces are safe. The single largest investment lands in phase three, once the cheap formats have produced real numbers to aim it with.
The four phases
Ninety days, in dependency order
Days 1–14
Unblock and baseline
Nothing new ships until the landing surfaces are safe and the starting numbers are written down. Almost all of this is already staged.
- Publish the staged conexeu.com fixesCanonicals, the twelve rewritten meta descriptions, the heading-tree repair. Already verified on conexeu-bioregeneration.webflow.io; production is unchanged.
- Resolve /platform/in-work---aesthetics-cxuA page titled “IN WORK” is publicly reachable on the exact vertical this plan is built on. No post can link to aesthetics until it is fixed.
- Settle the legal-name string and the brand-canon conflictEDGAR spells it Conexeu Sciences Inc., no comma. And the vault brand guide and the live stylesheet disagree on type and palette — templates cannot be built twice.
- Baseline the eight target queries in AI and GoogleChapter 07’s scoreboard is worthless without a before. Record it now, before anything ships.
- Legal rails: one reviewed comment reply, one caution-line placement standardImprovised replies from a Nasdaq issuer about an investigational device are the largest uncontrolled risk in the programme.
- Open the Facebook page and reserve the TikTok handleFacebook launches; TikTok is reserved only. Profile completion is an entity-graph signal even before a first post — name field, bio keywords, EDGAR legal name.
- Write the TikTok decision memoOne page: what may be said unbranded, who moderates, how a comment describing a reaction is escalated and logged, who signs off, what is retained. Play 05. A documented no closes the question for the quarter.
Days 15–30
Ship what already exists
Two formats need no new production. Get them out, and let them generate Conexeu’s own benchmark data instead of borrowed benchmarks.
- Cut six macro material films from the vault stillsFormat 05. 53 photographs already shot. 6–12s, seamless loops, texture audio, no claims. Mastered 9:16. The comparator frames stay back for legal.
- Ship all six to ReelsReels, not TikTok — overlapping audience, an account that already exists, and a comment surface small enough to moderate at 695 followers. These six are the measurement instrument for short-form.
- Build and post the GLP-1 seven-slide carouselChapter 02’s storyboard, as specified. Instagram at 1080×1350 and LinkedIn as a native PDF document post — the highest-engagement format on that platform.
- Start the founder cadenceOne post a week each from the CSO and CMO, from their own profiles, first person, reviewed. This is the play with the highest measured ratio in the whole plan.
- Fix the follow graph and the biosFollowing 13 accounts is a broadcast posture. Follow the injectors, societies, KOLs and GLP-1 clinicians; rewrite both bios for in-app search.
Days 31–60
Build the mechanism asset
One render, six placements. This is the quarter’s single largest investment and the one with the highest measured ceiling in the category.
- Commission the sub-dermal mechanism renderFormat 01. One continuous descent, three named layers, ending on host cells populating the lattice. Legal review at storyboard, not at render lock.
- Cut it three ways and ship across the three live channels in one week12s, 22s, and 45s with CSO narration. Same asset, three surfaces, one week — so the comparison between channels is clean. The 12s cut is mastered 9:16 and held in reserve for TikTok.
- Lift stills into the carousel and the aesthetics page heroSlide four of the proof carousel and the vertical page both need the cross-section. One render pays for six placements.
- Publish the reframe as page text with question-shaped H2s“Occupy, provoke, structure” has to exist as a citable paragraph in the first 30% of a page, not only as a carousel slide.
- Write the Organization node and raise the Equisolve duplicate as a tickettickerSymbol, logo, identifier, and a sameAs array carrying both social profiles. The competing IR Corporation node is a vendor change outside our access.
Days 61–90
Compound and read the results
Hold cadence, run the reframe series, and produce the first honest read on what worked on Conexeu’s own account rather than on published benchmarks.
- Run the category reframe as a six-post seriesFormat 04. Test Reel against three-slide carousel on the same argument; the carousel usually wins saves and the Reel usually wins reach. All three columns stay unbranded — this is the format with the most fair-balance exposure.
- Six weeks of one GLP-1 question per week from the CSOFormat 03. One question per clip, shot at the bench, burned-in captions, posted from her own profile.
- Re-check the eight target queries against the day-1 baselineThe only way to know whether the AEO work moved anything. Monthly from here.
- Report conversations, not followersQualified inbound — clinician, investor, injector, partner — attributed to a named post. Follower growth is the leading indicator, not the goal.
- Decide on the comparator macro framesHighest-value and highest-risk images in the vault. Either legal clears them with data behind them, or they stay back for another quarter. Do not let this drift undecided.
The scoreboard
What gets measured, and what the number has to beat
Six numbers. Each has a baseline read on 20 August 2026 and a target that is beatable, because a target nobody believes gets quietly dropped in month two.
| Measure | Baseline (2026-08-20) | Day-90 target | Why this number |
|---|---|---|---|
| Best single post, engagements | 82 (Nasdaq film, LinkedIn) | Beat 82 twice | The mechanism film and the GLP-1 carousel should each clear the company’s current ceiling. If neither does, the format thesis is wrong. |
| Founder share of programme engagement | 0% | >50% | Personal posts out-earn brand pages 1.6× on rate. If the founders are not carrying half the programme, the cheapest play in the plan is not running. |
| Instagram cadence gap, longest | 34 days | ≤5 days | Dormancy makes each return post get re-tested against a cold audience. This is the highest-leverage discipline fix and it costs nothing. |
| Reels completion rate | Not tracked | >50% on 4 of 6 macro films | 50% completion is where short-form distribution measurably accelerates. The six macro films are the instrument that tells us whether this audience will watch material behaviour at all. |
| TikTok question | Open, undocumented | Signed memo either way | Play 05. The deliverable is the decision, not an account. A documented no is a result — it stops the question being reopened on instinct every quarter. |
| Target queries with a Conexeu citation | Baseline in phase 1 | 3 of 8 | Eight queries in Chapter 07. Three is a real result for one quarter against journals and established practices. |
| Qualified inbound attributed to a post | Not tracked | Tracked, with a number | The actual objective. Clinician, investor, injector, partner. Everything above is a leading indicator for this row. |
Baselines read off the live posts and profiles on 2026-08-20. Benchmarks behind the targets are published 2026 datasets, cited in Chapters 02, 03 and 06.
The two things that will actually kill this plan
Cadence, and legal latency. Not creative quality — the creative direction is settled and two formats need no production at all.
The 34-day gap in the Instagram feed is the failure mode already on the record, and every asset in here is promotional material for an investigational device, which means a review queue sits between the idea and the post. Build the review into the calendar as a step with a duration, not as an approval that happens when someone gets to it.