What if my prompt conflicts with product data, research, or formatting settings?

When generating content, WriteText.ai combines multiple sources of information such as your product data, target SEO keywords, image analysis, AI-driven research and even your own instructions or prompts, to produce optimized, relevant text. It then applies your template’s formatting and layout rules.

Sometimes these things can conflict. To make sure you stay in control, WriteText.ai follows a clear hierarchy of instruction priority so it knows which to follow.

Important caveat — programmatic rules come first

Before any input priority is considered, WriteText.ai applies programmatic/styling rules you’ve set in your template. These are hard-coded formatting and layout settings that the AI cannot override, including:

• Heading level
• Bold, italics, or other text styling
• List vs. paragraph format
• Section ordering and placement

These rules always win because they are enforced after the AI generates the text.

Example: 
If your template section is set to bulleted list, but your prompt says “write in paragraph form”, the output will still be a bulleted list.

Hierarchy of input priority (highest to lowest)

Once formatting rules are applied, the system follows this order:

1. User-provided prompts or instructions
   1. Includes Custom prompts (Template), Special instructions (Template and Bulk), and Additional prompt (Single) fields you enter when generating content.
   2. These take top priority because they reflect your explicit creative direction.
   3. These are optional — if your product data and research are accurate and you don’t need to add other details, you may not need them.
   4. If you do use them, be careful not to add contradictory instructions.
2. Product data, image analysis, and target keywords
   1. Data pulled directly from your product setup in your ecommerce backend, including attributes, meta fields or custom fields, featured image, and other product details you manually input.
   2. Also includes your target keywords, whether manually entered or generated through automated keyword analysis and presented in the keyword optimization pipeline.
3. Product research results
   1. AI-assisted research that gathers information from external sources.
   2. Used to enrich and expand content where details are missing or incomplete.

How conflicts are resolved

• Formatting conflicts → Template rules win.
• Content conflicts between prompts and product data → Prompts win.
• Conflicts between product data and research → Product data wins unless overridden by prompts.

Why it’s designed this way

This priority structure ensures you remain in control of your content, while still making the most of available data and research when needed.

1. Programmatic/template formatting rules are at the very top because they guarantee visual and structural consistency across your entire catalog. Formatting is handled programmatically — not left up to the AI to insert HTML tags or apply styles — because the AI can be inconsistent in applying formatting.
   1. This ensures that settings like bullet lists, heading levels, and bold text are always honored exactly as configured, no matter what your prompt says.
   2. Because formatting is applied after AI generation, you can rely on it to be predictable, uniform, and SEO-friendly across all products.
   3. Exception: If you have a custom section in the template with a custom prompt and set its Text layout to Prompt defined, you can describe your ideal formatting or layout in the prompt itself. For example: “Format as a bulleted list where the first word of each bullet is bold and the rest of the text is regular. Use bullet points in color #367C1A.”
   4. If you choose Prompt defined and don’t describe any formatting in the prompt, the text will be output as a plain paragraph with no special formatting applied.
2. User-provided prompts or instructions come next because they reflect your explicit creative direction. If you give the system a clear prompt, it will follow it — even if that means overriding product data or research results. This gives you total control when you need it.
   1. If your product data are accurate, you can confidently rely on research results, and you don’t want or need to add other details to better influence the result, you can skip special instructions and custom prompts entirely. They’re optional — but if they are present, they will be followed.
   2. Be careful not to add contradictory prompts that intentionally or unintentionally confuse the model. For example, if your template is set to output a bulleted list but your instruction says “write in paragraph form,” the programmatic formatting will still apply, and your prompt will be ignored in that case.
3. Product data, image analysis, and target keywords are prioritized next because they are structured, factual data that usually comes directly from your store. They fill in accurate details when no overriding prompt is present.
4. Product research results sit at the bottom of the priority list. While research can uncover useful competitive insights and missing details, it’s considered less authoritative than your own data or instructions. It’s used to supplement, not replace, your existing information.

In short:

• Formatting rules guarantee consistent structure and styling because they’re applied programmatically, not generated by AI.
• Prompts give you complete creative control when needed.
• Product data keep details accurate when you don’t override them.
• Research fills in the gaps when other inputs aren’t available.

This design allows you to scale content production confidently, knowing exactly which source of truth will win in any conflict.

How conflicts are resolved

If conflicting information is found between these sources, the higher-priority source wins.

Example 1 — Feature inclusion

• Product research: Finds “includes a leather strap.”
• Your special instruction: “Do not mention leather or animal materials.”
• Outcome: Leather is omitted from the content.

Example 2 — Keyword usage

• Target keyword: “luxury silk scarf.”
• Custom prompt: “Focus on cotton blend, not silk.”
• Outcome: Content emphasizes cotton blend.

Best practices for avoiding conflicts

• Check template settings first before adding formatting instructions.
• Write clear and specific instructions to reduce ambiguity.
• Use Special instructions for tone, prohibited terms, and must-mention features.
• Let product data and research handle enrichment unless you have a reason to override them.

By understanding how WriteText.ai prioritizes inputs, you can scale content creation for one product or your whole catalog — while ensuring the AI never goes against your explicit directions.