How to use Fable to plan your marketing for the next 3 months

Fable is a strong AI model, and it is a genuinely good marketing strategist with one catch: it is only as honest as the numbers you give it. Feed it your real follower counts, your real traffic, your real sales, and it will build you a plan grounded in your actual situation. Feed it vague optimism and it will hand back generic advice you could have got from any blog post.

This guide gives you the full process: what to gather, the exact prompt to paste into Fable, and how to work with the answer. It takes about an hour to prepare and produces a week-by-week plan for the next 3 months. (The method works with any strong AI model, but the prompt and the review habit below are tuned for how Fable responds.)

Step 1: Gather your inputs (30 to 60 minutes, be brutally honest)

The plan is only as good as this step. Collect the following. Where you do not know a number, write "unknown" rather than guessing. "Unknown" is useful data; a guess is poison.

Your goal

  • The ONE business outcome the next 3 months should move. Not "grow my following". Something like "15 new paying customers", "30 booked calls", "500 email subscribers who match my buyer". Pick a number.

Your accounts, one honest line each

  • Every social account in scope: platform, handle, follower count, posts published to date, typical likes and comments per post, posting frequency. Include the embarrassing ones. An account with 30 followers after 90 posts is exactly the kind of fact the AI needs.

Your sales reality

  • How many paying customers or sales do you have now? How many came in the last month? What does a customer pay you? What is the path from stranger to sale (see your content, visit your site, sign up, buy)?

Your traffic and tracking

  • Monthly website visitors if you know them, and crucially: can you tell where a visitor or customer came from? For most small businesses the honest answer is "no, almost everyone shows up untracked". Say so. Fixing this will likely be part of your plan.

Your content engine

  • What you currently make and how. What is easy for you to produce repeatedly (talking to camera? writing? screenshots of your work?), what you hate making, how many hours a week you can genuinely give this. A plan built on 20 hours a week you do not have is worthless.

Your audience guess

  • Who actually buys from you (not who you wish would). Any pattern you have noticed in real customers: sector, size, situation, the problem that made them buy.

Existing commitments

  • Anything already planned or scheduled (a launch, an event, content already in the calendar). The AI should build around these, not ignore them.

Constraints

  • Budget (often zero, that is fine), tools you have, platforms you refuse to be on, anything off the table.

Step 2: Set the ground rules

Before you paste the prompt, understand the three rules baked into it:

  1. Sales-first, not reach-first. Likes and followers are means, not ends. Every recommendation must trace to your goal number.
  2. Honesty about measurement. If you cannot tell where customers come from, the plan must include fixing that, otherwise you will finish the 3 months no wiser than you started.
  3. A plan you will actually execute. The AI must fit the plan to your real capacity, not an imaginary content team.

Step 3: The prompt (copy, fill the brackets, paste)

You are a senior social media and growth strategist for a small business.
Your job is to produce a 3-month social media strategy whose success is
measured in SALES, not reach. Reach is a means; every recommendation you
make must state how it moves the sales goal.

## The business
[One paragraph: what you sell, to whom, at what price, and how someone
goes from stranger to customer.]

## The goal for the next 3 months
[Your one number, e.g. "15 new paying customers attributable to social".]

## Current state (real numbers, verified today)
Accounts:
[Platform / handle / followers / posts to date / typical engagement /
posting frequency. One line per account. Include the bad ones.]

Sales: [current customers, last month's new sales, price point]

Traffic and tracking: [monthly visitors if known, and an honest statement
of whether you can attribute visitors/customers to a source. If you are
mostly blind, say so plainly.]

## My content engine and capacity
[What you make now, what is easy/hard for you to produce, hours per week
you can honestly commit, tools you have.]

## Who actually buys
[Your best evidence-based read of your real customers.]

## Existing commitments
[Anything already planned or scheduled that the strategy must build
around, not replace.]

## Constraints
[Budget, off-limits platforms, anything else fixed.]

## Rules for you, the strategist
1. Sales-first. Every recommendation ends with the mechanism by which it
   drives sign-ups or sales. If a recommendation only grows reach, either
   connect it to sales or cut it.
2. Be honest about my current state. Do not flatter the numbers. If an
   account or tactic is not working, say so and say what changes.
3. If my tracking is blind, fixing attribution is a first-class part of
   the strategy, not a footnote. Give me the simplest fix that works
   (tagged links, one question at sign-up, a dedicated link per platform).
4. Fit the plan to my stated capacity. Prefer one repeatable content
   system over many one-off ideas.
5. Include an email-capture mechanic. Followers are rented; an email list
   is owned. Recommend a lead magnet and the exact capture flow
   (e.g. "comment WORD and I will send it to you", link-in-bio, DM
   automation), matched to my platforms and tools.
6. Before writing the strategy, list up to 5 questions whose answers
   would materially change the plan. Then answer them yourself with
   labelled assumptions and proceed. Do not stall waiting for me.
7. Where you recommend a format (video length, style, structure), commit
   to a specific recommendation with reasoning, not a menu of options.

## Produce the strategy in exactly this structure
1. CURRENT-STATE READ. The honest picture from my numbers: what is
   working, what is not, and the single biggest thing holding back sales.
2. POSITIONING AND TARGET AUDIENCE. Who the content is for, what I should
   be known for, and why that person would buy.
3. CHANNEL PLAN. Which platforms to focus on, which to maintain, which to
   drop, with reasoning tied to where my buyers actually are.
4. CONTENT SYSTEM AND FORMATS. A repeatable weekly engine (pillar content
   plus the assets spun off it), the specific formats to use, and the
   format recommendation for any content type I flagged as undecided.
5. LEAD MAGNET AND EMAIL CAPTURE. What the magnet is, why my audience
   wants it, the exact capture mechanic per platform, and what the email
   sequence does after capture (drive to a sale, not just a newsletter).
   Tailor the capture mechanic to my specific industry/niche; one
   mechanic does not fit all, so say how to adapt it to mine.
6. THE 3-MONTH CALENDAR. Phased week-by-week (or fortnightly) plan:
   what ships, on which account, in which format, and what each phase is
   trying to prove. Build around my existing commitments.
7. METRICS AND THE ATTRIBUTION FIX. The 3 to 5 numbers I track weekly,
   the specific fix so I can see where sign-ups come from, and the
   checkpoint criteria at weeks 2, 6 and 12 for doubling down or killing
   each tactic.
8. QUICK WINS: WEEK 1. The 5 highest-leverage actions I can complete in
   the first week, each doable in under 2 hours.
9. PAID, A SIMPLE STARTER PLAN (follow-up section). A simple, effective
   way to use a small paid budget with little paid experience: which ONE
   platform to start on and why, the single campaign type to run, a
   sensible starting daily budget, exactly what to promote (usually the
   lead magnet or the best-performing organic post), how to tell if it
   is working within the first 2 weeks, and the beginner mistakes to
   avoid. Quick wins first, keep it small and measurable, tie every
   pound to sign-ups or sales.

Step 4: How to run the session

  • Interrogate, do not just accept. When the strategy arrives, push on anything vague: "you said post Reels 3x a week, give me the first 6 topics", "you recommended a lead magnet, write its outline". Fable gets sharper under cross-examination.
  • Challenge the flattery. Ask directly: "what is the weakest part of this plan and what would a sceptic say?" Make it argue against itself once.
  • Ask for the week-1 list as tasks. Have it rewrite section 8 as a checklist with time estimates. Do those five things before touching anything else.

Step 5: Live with the plan

  • Put the calendar somewhere you see daily.
  • Track your 3 to 5 metrics weekly, badly is fine, consistently is mandatory.
  • At weeks 2, 6 and 12, paste your actual numbers back into the same chat and ask: "here is what actually happened, adjust the plan". The review loop is where the compounding happens; the first plan is just the starting guess.

That is the whole method. An hour of honest preparation, one well-structured prompt, and a review habit. Fable does the strategy work; you keep the judgement and do the showing up.