AI Social Media Growth
0/15 complete

Module 01 · Building Your Content Engine

Build a Consistent, Testable Content System

A content system replaces “What should I post today?” with a weekly loop: collect audience problems, produce, publish, and review comparable evidence. AI can help organise drafts; it cannot decide what your audience values or guarantee distribution.

You will build a one-page operating system: one niche promise, three content pillars, three repeatable formats, a minimum cadence, and a bad-week-proof review ritual.

// concept

Build the Weekly Loop Around Audience Problems

Start with one audience, not a platform. Write:

// prompt — copy me2 lines
I help [specific audience] solve [recurring problem] by publishing [useful kind of content],
without [boundary I will not cross].

“I post about food” is broad. “I help Karachi office workers choose weekday lunches by showing the menu, portion and ordering steps, without health claims” guides topic choices.

Turn the promise into three content pillars: recurring problems, not vague subjects. Lunch-page pillars might be what is available, how to order, and safe storage/reheating. File sanitised customer questions and FAQs under them; private chats and order details are not prompt material.

The loop then has four stages:

StageWeekly actionEvidence saved
InputCapture five real questions under the three pillarsSanitised question list
ProductionBatch, then fact- and rights-checkApproved drafts
PublishingUse pre-chosen slots; record the post URLDate, platform, format, URL
ReviewCompare after the same windowMetric, questions, next test

// concept

Separate Pillars, Formats, and Cadence

A pillar is what problem you cover; a format is how it is delivered. Choose three formats suited to your skills and equipment:

  1. One-question explainer: answer one question in video or text.
  2. Process breakdown: show three to five steps and the key limitation.
  3. Before/after critique: improve a labelled sample; never expose client work without permission.

Set a minimum viable cadence: the smallest schedule that creates reviewable work without lowering quality. Two original posts per week is a planning hypothesis, not an algorithm rule. Consistency produces comparable practice and a recognisable promise; it does not force recommendation systems to distribute posts.

Ask: Could I complete this in a bad week—with deadlines, exams, load-shedding or low energy—while checking facts, captions and rights? If not, remove slots before publishing.

// concept

Write the One-Page Operating System

Keep this beside your calendar:

FieldDecision to record
Audience + problemOne group; one recurring problem
Niche promise“I help…” sentence with a boundary
Three pillarsThree problem areas that generate topics
Three formatsThree repeatable post structures
Weekly slotsDay, platform and format—not “post more”
Review ritualFixed window, job-matched metric, questions, next test

Sample system 1 — hypothetical Karachi food-business page: Audience/problem: nearby office workers need clear weekday lunch information. Promise: explain choices and ordering without health or delivery-time guarantees. Pillars: menu clarity; ordering; storage/reheating. Formats: menu walkthrough; FAQ carousel; kitchen-process clip. Slots: Tuesday walkthrough; Thursday FAQ; Saturday clip optional. Monday review: compare equal-age posts, inspect saves, completion where available and order questions; test one opening.

Sample system 2 — hypothetical freelance designer’s personal brand: Audience/problem: Pakistani small-business owners struggle to prepare a first visual-identity brief. Promise: explain design decisions without presenting samples as client results. Pillars: brief quality; visual hierarchy; handoff files. Formats: annotated sample; three-step document; screen recording. Slots: Wednesday LinkedIn document; Friday Instagram reel. Tuesday review: record relevant comments, saves and profile actions; hold format steady and test one pillar.

// worked_example

Worked Example

The hypothetical designer asks AI to organise rough notes. The sample prompt is:

// prompt — copy me7 lines
Act as a planning assistant, not a growth forecaster.
Audience: Pakistani home-business owners who struggle to brief a designer.
Problems heard repeatedly: unclear deliverables, conflicting colour requests, missing logo files.
Capacity in a bad week: two hours. Platforms: LinkedIn and Instagram.
Create one niche promise, three problem pillars, three repeatable formats, two weekly slots,
and a 20-minute review ritual. Do not predict reach or invent audience research.
Flag any plan that cannot fit inside two hours.

Sample AI draft one: Monday reel, Tuesday carousel, Wednesday stories, Thursday document, Friday video, plus daily comments; its pillars were “design, business, motivation.”

The draft exceeds two hours, and its pillars are not audience problems. Fix it with brief clarity, decision trade-offs and handoff readiness; use a sample carousel, short screen recording and checklist; publish Wednesday and Friday. On Tuesday, record both URLs after the same seven-day window, one job-matched signal such as checklist saves, and repeated questions. For an Instagram professional account, check Profile → View professional dashboard → See all insights. On LinkedIn, use Activity → Show all posts → View analytics. Confirm current labels and availability in official help.

// failure_cases

Failure Cases to Diagnose

6 cases to diagnose

  • Pillars are “AI, design, lifestyle.”

    Rewrite each as a recurring problem that produces five questions.

  • Formats change with every trend.

    Name a repeatable structure, not a sound, filter or meme.

  • The bad-week test fails.

    Remove a slot; do not cut fact, rights or caption checks.

  • Everything changes together.

    Hold topic, format and window stable; test one variable.

  • Views are the only note.

    Match evidence to the post’s job and read audience questions.

  • AI receives private messages.

    Extract the question; remove names, handles, numbers and order details.

// pakistan_angle

Pakistan Angle

If load-shedding or weak mobile data interrupts uploads, batch captions and compressed media offline, then publish during reliable connectivity. Do not declare one “best Pakistani posting time.” Test around your audience’s commute, work, Maghrib/Isha and, when relevant, Ramzan routine. A Karachi food page and nationwide LinkedIn audience need different tests.

Test English, Urdu script or Roman Urdu while holding topic and format stable; check Urdu and captions on a low-cost Android phone. Never paste WhatsApp customer numbers, addresses, JazzCash/easypaisa receipts or CNIC images into AI. Convert them to: “customers ask whether advance payment is required.”

// hands_on

Hands-On Exercise

7 steps

  1. List one audience and five sanitised recurring questions.

  2. Write the niche promise with one honesty boundary.

  3. Group the questions into three problem pillars.

  4. Choose three formats your current equipment supports.

  5. Assign two required slots; mark a third optional.

  6. Set a review window, job-matched metric, questions to inspect and one variable to change.

  7. Apply the bad-week test and cut until it fits.

// completion_rubric

Completion Rubric

6 checks — tick as you verify

0/6

// sources

Sources

// check_yourself

Check yourself

4 questions · answers and options are taken word-for-word from this course

0/4
  1. 1 / 4 · diagnose

    Your work shows this failure mode: “Pillars are “AI, design, lifestyle.”.” What does the lesson tell you to do about it?