Forge Academy: Content Ideas Engine

A useful content idea answers a question that matters to a particular buyer and can be supported by what your business knows. “Write about our service” leaves those decisions unresolved. A stronger starting point names the buyer's situation, the obstacle, the available evidence and the decision the content should help them make.

The content ideas engine in this lesson is a planning method built from Forge's structured knowledge. You use Node information and buyer–pain–proof relationships to develop candidate angles, then use documented creation and review workflows to prepare the content. This does not assume a separate interface module called Content Ideas Engine.

Begin with a buying situation, not a posting quota

Choose a single offer and ask what happens before someone considers it. Are they planning a change, comparing suppliers, investigating a recurring problem or trying to understand an implementation requirement? The situation determines which questions are useful now.

Our illustrative business provides office-relocation planning. Its operations buyer needs a workable sequence of responsibilities before moving day. Its finance buyer needs to understand the boundary between the planning scope and costs handled by other suppliers. The business has an approved sample dependency schedule and a scope checklist. These are hypothetical teaching materials, not a real customer case.

The idea “office moves are stressful” is too broad to help either buyer. “Which decisions must be made before the relocation schedule can be agreed?” is a more useful operations question. “Which relocation costs are outside this planning proposal?” is a different finance question. Both can arise from the same service without repeating the same message.

Map the relationships that can support an idea

Review the Node's description, value proposition and differentiator. Then inspect the buyers, their pains and the proof connected to those pains. The Forge User Guide explains how to add these relationships through the Node's buyer, pain and proof counts.

Look for missing links rather than a large quantity of entries. A pain without evidence may still support a carefully framed educational discussion, but it cannot support a claim that your service has solved that problem for customers. A proof item without a relevant buyer question may produce a technically correct post that nobody needs.

If you have 2 buyers, 3 pains and 2 proof items, simple multiplication produces 12 possible combinations. That arithmetic does not mean you have 12 valid ideas. Some evidence may address only one pain, and some pains may belong to only one buyer. Respect the actual relationships instead of connecting every item to every other item.

Use the Knowledge Base Setup lesson if the offer boundaries or evidence relationships need repair. An idea session should reveal knowledge gaps as well as produce drafts.

Turn a valid relationship into different kinds of help

Once you have a credible buyer–pain–proof chain, ask how the reader could use it. Several angles may be valid because they support different decisions. A diagnostic question helps someone recognise a problem. A process explanation helps them understand how work proceeds. An evidence demonstration lets them inspect what your offer actually provides.

For the fictional relocation service, the following candidate bank changes the reader's task while keeping each idea grounded.

Candidate idea	Buyer task	Evidence or boundary
Which decisions must come before a relocation timetable?	Identify prerequisites before agreeing dates	Show the dependencies in the approved sample schedule
Who owns the handover between the office team and the moving supplier?	Clarify responsibilities	Explain the responsibility entries included in the planning scope
What does the planning proposal include, and what remains a separate cost?	Review a budget request	Use the scope checklist and its explicit exclusions
How should a team read a dependency schedule?	Understand the proposed deliverable	Walk through one authorised example without implying a guaranteed move date
What should you bring to a relocation-planning discussion?	Prepare for a useful enquiry	List the information the service actually requires
When should the relocation plan be reviewed again?	Recognise a material change	Explain how a changed requirement can affect the agreed planning scope

The last idea requires the business to confirm its review process before publication. The table is a candidate bank, not permission to invent a procedure. Where the underlying answer is unknown, mark the idea as needing an owner response rather than asking AI to fill the gap.

Add a novelty constraint that changes the substance

A novelty constraint states what the next piece must contribute beyond recent content. It could introduce a different decision, a previously unexplained limitation, a new approved proof item or a useful worked example. Record the difference in a sentence.

For example: “The previous post explained the planning deliverables; this one helps finance distinguish included work from separately purchased services.” That difference is meaningful. Rewriting the same deliverables with a new opening sentence creates a variation, but does not by itself create a new buying question.

Review the recent publication record before accepting an idea. Look at the main claim, buyer, proof, requested next step and takeaway. Two posts can use different vocabulary while making the same argument. Conversely, two posts can use the same proof responsibly when one explains its limits and the other helps a different buyer inspect it.

Keep planned repetition intentional. An important offer may need to be explained more than once because different readers encounter different publications. Repeat the fact accurately, then give the new piece a clear purpose. Do not chase novelty by adding an unverified claim.

Write a brief that a draft can be judged against

For each selected idea, prepare a short working brief outside the interface if needed. Include the buyer question, the relevant Node, pain and proof, the single takeaway, the destination and the next action. Add any prohibited inference, such as “the sample schedule does not prove that every move finishes on time”.

Use this brief to review the result of Manual generation or your own commentary on the chosen surface. For long-form work on website_article, the documented Content Studio flow through Article Studio lets you choose a buyer, pain, proof and framework for the article. Verify the available workflow and destination requirements in your account.

Do not assume that a surface change automatically creates a new idea. A short social post might introduce the question while an article explains the dependencies in detail. The relationship is useful when the article supplies additional substance, and the post describes it accurately. Maintain the agreed voice with the Brand Voice Setup lesson.

Select ideas by readiness and usefulness

At the end of a session, divide candidates into 3 practical groups: ready to draft, needs a factual answer and not useful now. These are working categories for your team, not named Forge statuses.

• Ready to draft: a clear reader question, relevant evidence, a distinct takeaway and an available next step.
• Needs a factual answer: a useful question with an unresolved scope, proof or approval requirement.
• Not useful now: a duplicated argument, an irrelevant audience or a topic that does not support a current reader need.

Choose a small number of ready ideas and confirm that the person responsible can review them. A large idea bank is useful only if it helps selection. Turning every candidate into a scheduled post can magnify repetition and consume review capacity before the weak questions are resolved. Use the content calendar guide when assigning a realistic publishing rhythm.

Let the response improve the next question

After publication, inspect available supported metrics and the substance of relevant audience responses. A buyer asking about an exclusion may reveal a useful follow-up explanation. A repeated misunderstanding may point to an unclear phrase in the offer. An unavailable metric does not demonstrate that an idea failed.

Record the question that the next idea should answer, then verify the relevant facts before adding it to the bank. The Performance Loop lesson shows how to make one evidence-based update without claiming that a small observation proves audience preference or commercial causation.

For your next session, take one offer and produce 6 candidate questions. Explain the distinct reader task behind each, identify its evidence and select only those ready for drafting. Open Forge to check the underlying knowledge. If the team needs help turning its buyer problems into a credible first idea bank, request a demo focused on content ideas. See the Forge overview for how knowledge, creation and publication fit together.