Content strategy

Channels and cadence#

Two Instagram accounts, launched August 2026: directional

• @eddiecheng.ig: Eddie's personal account, video-first, the lead generation engine. Target 4 to 5 videos per week.
• @penangmedia: company account, graphics-first (carousels, statics, some motion). Target 3 posts per week.

Content is batch-produced in advance; Eddie's filming availability is the deliberate bottleneck and is protected through a fixed weekly batching session. Every personal video passes Eddie's sign-off before publish.

The channel decision is deliberate: the buyer's working day happens inside Meta, so paid funnels and Instagram meet them there. LinkedIn was evaluated and rejected as a channel. Any future channel addition is a separate strategic decision, not a content-level one. directional

Channel roster beyond Instagram (decided by Eddie, 2026-08-13 interview)#

• Instagram: the nurture channel, full stop. Content plus ads run through it, with the goal of being top of mind when the business consideration moment arrives. verified
• YouTube: aspirational secondary. He wants consistent videos and has published sporadically, but no committed cadence exists yet. The March 2026 cluster-strategy thinking (own "Meta ads for DTC brands" for 90 days at one to two videos a week before expanding into DTC growth, performance creative, and AI) stays on file as the playbook for whenever YouTube gets real investment. verified
• Twitter: parked. The 50k+ followers were built on viral stories and are mostly untuned to the offer, and he hasn't posted consistently in about two years. The purpose-built Twitter content platform exists but nothing is running; revisit is queued behind higher priorities. verified

Current priority stack, in his words: the working session funnel and its ads first, Penang Media lead gen funnel second, Instagram content for both handles third. verified

Pillars#

The decision behind the number#

Problem it solves: founders make decisions from isolated metrics without checking what changed elsewhere in the market, ad account, or business.

The signature story comes from VIBAe's localized advertising. The team tested native-language ads in markets including Germany, France, Spain, and Sweden. Some localized campaigns reported below-average ROAS and appeared to be poor performers. At the same time, established English campaigns serving those countries saw an uplift in ROI. This did not prove that the localized ads caused the improvement, but it showed that switching them off based on isolated campaign ROAS would have been premature. The next iteration moved beyond translation toward culture-first concepts while continuing to measure the effect across each market and the wider ad account. verified by Eddie

Point: "A result is only useful if it changes what you do next."

Framework: Local Result Versus Total Effect. A weak local result with a negative total effect means stop or redesign. A weak local result with a positive total effect means keep learning and iterate. A strong local result with a negative total effect calls for a check for displacement or hidden costs. A strong local result with a positive total effect can scale with guardrails.

Action: before switching off an apparent loser, compare its isolated result with what changed across the same market and wider ad account. Choose one decision: stop, investigate, or scale.

The business behind the ad account#

Problem it solves: founders try to scale through paid media before proving that demand, margins, and lifetime value can support it.

The signature story comes from a discovery call with a Finnish energy-device company. Its electricity-monitoring device generated €40 in revenue excluding tax and cost a little over €20 to produce at its existing volume. Even a hypothetical €20 acquisition cost would leave almost no contribution. Possible recurring revenue of €1 to €2 per device each month could have changed the calculation, but it had not been proven. Follow-up keyword research found high Finnish search demand for spot electricity prices, where users wanted free information from utility apps, and a much smaller category of searches with buying intent for the hardware. Eddie declined the retainer and recommended utility distribution or proof of recurring revenue first. Paid media could become an amplifier once the business had traction and economic headroom. verified from call and follow-up email

Publication note: keep the company anonymous. The economic figures are decision context and require separate clearance through the claims register before use in marketing content.

Point: "You cannot make a good ad decision without understanding the business behind it."

Recurring points: "The ad account does not decide whether you can scale. The economics do." "A profitable-looking ad account can still be losing the business money." "ROAS is not the business."

Framework: The Amplifier Test. Check four things before investing in paid media: traction, the signal being amplified, economic headroom, and the measurable trigger for more spend. If an answer depends on hope rather than evidence, validate it before trying to scale it.

Action: answer the four Amplifier Test questions using real evidence and numbers before approving a paid campaign. If one answer is missing, set a small validation test and spending limit instead of building a scale plan.

Creative that compounds#

Problem it solves: creative teams produce hundreds of variants without a reliable record of which concepts, hooks, and formats created traction.

The signature story comes from a creative workflow meeting with Desiree. A task-based system could show that creative had been produced, but one task could contain several ads with different combinations of concepts, hooks, and formats. It could not show which territory the team had explored. That led to the family-tree model. Concepts form the trunks, hooks and formats form the branches, and results show which branches to grow, cut, or explore. Without that map, creative production becomes "hit and run" and the next testing cycle loses the benefit of the last one. verified from 2026-03-31 team call

Point: "If your creative process has no memory, every month starts from zero."

Framework: The Creative Tree, within the Branded Response Flywheel. The tree maps Concept → Hook → Format, then records each combination as validated, unsuccessful, or untested. It becomes active during the Iterative Creative Formula, learns during Testing and Intelligence Extraction, and directs budget during Profit-Driven Scaling.

Action: map one active concept, including every hook and format already tested. Mark each combination as validated, unsuccessful, or untested. Write the next brief by pairing a validated element with one promising untested branch.

The framework visual prompts remain content-production assets. They do not belong in the brand brain.

Strategic thesis#

The content engine is not a parallel workstream to the paid funnel; it is the funnel's actual front end. A £197 paid session is a meaningful ask from someone the buyer discovered this week, so familiarity precedes conversion, and the first sixty days of content should be proof-dense rather than hook-dense. New accounts mean a visible social proof gap early on; that is expected and temporary, provided the content shows the goods. directional

Measurement#

Views are the primary reach signal. Profile visits per 1,000 views is the sanity check that distinguishes genuine buyer reach from broad-audience spectators. directional

Hook routing#

New hooks enter the working session funnel first as Meta headline and primary-text variants. The strongest identity-led concepts can then move into the Instagram script queue. Every variant must pass the fact bank, claims register, voice rules, and Eddie's sign-off before publication. directional

Related: offer architecture, hook bank, receipts standard, voice guide.