What a content engine is, and what it is not
Most social media offers sell you people: a community manager, a designer a few hours a week, a calendar that fills up. The Social Content Engine sells you a system. You own the accounts, the content and the model.
| A content engine is | A content engine is not |
|---|---|
| An editorial model: three to five recurring themes the brand owns, each with a reason to exist | A list of posting days waiting for ideas |
| Formats that can be produced every month without a new brainstorm | A one-off campaign concept that runs out in six weeks |
| A voice sheet that AI and people write against, in Spanish and English | A tone "we'll get a feel for over time" |
| Shooting days planned around formats, so one morning feeds several weeks | A photographer booked when someone remembers |
| A schedule you approve in one sitting, published on time | Daily "can you approve this?" messages |
| A monthly report on what moved a number, and the decision that follows | A screenshot of follower growth |
Month one: the setup
The first month is not a lighter version of the following ones. It is a different job, and it is quoted as part of the same proposal.
- Brief and audit. One session with the people who own the brand. What it sells, to whom, what it refuses to say. We read the last months on every platform you have, and what the competitors and references you name are doing.
- Editorial model. The themes the brand will talk about, what each one is for (awareness, consideration, proof, community), and how much of the month it gets. This is a two-page document, not a deck, and it is the thing the whole engine runs on.
- Voice sheet. How the brand sounds: sentence length, words it uses and words it never uses, how it handles humour, prices, complaints and news. AI writes against it, people check against it, and it lives in your files, not ours.
- Formats. For each platform, the recurring formats: what a piece looks like, how long it is, what a shoot needs to feed it. Formats are chosen for repeatability, because a format that needs a new idea every week is not a format.
- First shoot and first month's content. One shoot day, planned against the formats. Then the first month of pieces, produced, reviewed against the voice sheet, and scheduled.
By the end of month one you have the model, the voice sheet, the format library, a bank of footage and a month of content queued.
The monthly rhythm
From month two, every month looks the same. Predictable is the feature.
| When | What happens | What you see |
|---|---|---|
| Week 1 | Editorial planning against the model and the previous report; briefs for the month's pieces | The month's plan, for one approval |
| Week 2 | Shoot day (when the month needs one); AI-assisted production of drafts, variants and captions in ES and EN; voice check on every piece | Mid-month check: the pieces, ready to approve or comment |
| Week 3 | Revisions, final edits, scheduling across the two or three platforms; community answers routed to the right person | The calendar, filled and timed |
| Week 4 | Publishing continues; reporting on the previous month's pieces: what moved, what did not, why | The monthly report, one page, with a decision at the end |
The number of pieces per platform and the number of shoot days are fixed in your quote and do not drift month to month. If the brand has a launch, an event or news, the engine bends around it, and a page for that moment is the Launch Page.
What the brand provides
- Access to the accounts, which stay in your name.
- A brand book, or the closest thing you have. If you have nothing, the voice sheet becomes the brand book for social.
- People and products in front of a camera on shoot days.
- One person who can approve a month in one sitting.
What is measured
Follower counts are reported because people ask for them, but they are not what the report is about. Each month's report answers four questions, per platform:
- Which pieces moved a number, and which number? Reach, saves, shares, profile visits, link clicks, messages, sign-ups. Each format is judged on the metric it was built for.
- Which themes are carrying the account? So the editorial model gets more of what works and drops what does not.
- What did the audience do next? Visits to the site, form fills, WhatsApp messages, wherever the content pointed.
- What changes next month? One decision, written down, and the plan for week one follows it.
When it is not the right product
- You need one campaign, not a system. A launch, a season, an event. That is the Launch Page or a Campaign Microsite, with a start and an end.
- You want a person's voice, not a brand's. A founder or executive on LinkedIn is the Founder LinkedIn presence, a different method with interviews and a personal voice sheet.
- You cannot get people in front of a camera. The engine can run on product, design and text formats, but it runs better with faces. If that is off the table, we plan for it, and we tell you what it costs in reach.
- Approval takes weeks. The rhythm above assumes one approval per month. If legal review of every caption takes ten days, we build a smaller, slower engine or advise you to keep it in-house.
Where it comes from
The method is not new; the tooling is. LOBODS, the social media agency Dmytro Lobod co-founded in 2016 and still runs, has spent ten years building content systems for 350+ brands in 12 countries. Editorial models, format libraries, shoot days planned against formats, monthly reports with a decision at the end: that is how LOBODS has worked for a decade. The Social Content Engine takes that practice and runs the iterations with AI, so the same discipline costs a brand less time and less money.
Where to go next
- A moment that needs its own page this week? The Launch Page, 72 hours.
- Want to know whether ChatGPT and Gemini mention your brand at all? The AI Visibility Audit, 5 days.
Tell me which platforms, what you sell and who can approve a month in one sitting.