Dataset Card for marketing-instruct-4k

Dataset Details

Dataset Description

A curated instruction-tuning dataset of ~4,300 marketing copywriting examples across five task types, built for the AutoScientist Challenge 2026 (Marketing category). Used to fine-tune Marketing-Mixtral-8x7B.

Key finding: this carefully curated dataset at its natural size outperformed a 12,000-row version expanded via automated augmentation (80% vs 58% win rate against the base model), demonstrating that data quality and balance matter more than volume for this task.

• Curated by: Sue Huynh
• Language(s): English
• License: Apache 2.0

Dataset Sources

• Model trained on this data: suehuynh/Marketing-Mixtral-8x7B
• Challenge: AutoScientist Challenge 2026 — adaptionlabs.ai

Uses

Direct Use

Instruction fine-tuning for marketing copy generation models across: ad/social copy, email marketing, product descriptions, brand voice rewriting, and campaign data-to-insight summarization.

Out-of-Scope Use

• Not designed for marketing strategy, competitive analysis, or market research generation tasks — these were explicitly excluded during curation to keep the dataset focused on copy generation
• Not suitable for non-English marketing content without translation and re-validation
• Not intended as a source of factual product information; product attributes used in examples are illustrative, not verified real product specifications

Dataset Structure

Each example follows a unified instruction-tuning schema:

{
  "instruction": "Write a product description for an online store listing...",
  "input": "",
  "output": "The Power Bank Model-39 packs 20000mAh..."
}

Field	Type	Description
instruction	string	The task brief, including format and constraints
input	string	Optional structured context (e.g. attribute list, table); empty string if not applicable
output	string	The target marketing copy

Task Type Distribution

Task Type	Approx. Examples	Description
Ad/Social Copy	~1,200	Platform-specific ad and social post copy
Email Marketing	~900	Campaign and lifecycle emails (subject/preview/body)
Product Description	~800	Attribute-grounded product descriptions
Brand Voice Rewriting	~700	Neutral copy rewritten to match a specified voice
Table-to-Insight	~500	Campaign performance data summarized into insights

(Counts are approximate; see the dataset viewer for exact splits.)

Dataset Creation

Curation Rationale

Built to fine-tune an open marketing copywriting model covering the core deliverables a marketing team produces day to day, while explicitly excluding strategy/analysis tasks (e.g. SWOT analysis, competitor comparison tables, dashboards) to keep the model focused on copy generation rather than business consulting.

A specific design goal was attribute faithfulness: product description examples were filtered so that output claims are traceable to the provided input attributes, reducing the tendency of LLMs to hallucinate product features or specifications.

Source Data

Data Collection and Processing

The dataset combines three sources:

1. Public marketing datasets (HuggingFace), adapted and quality-filtered via Adaption Labs' AdaptData platform, including ad copy and product description sources
2. Synthetic examples, hand-designed across 40+ brand voice styles, 15+ industries, and varied table/insight scenarios, generated to fill gaps where public data was sparse or low quality (email marketing, brand voice, table-to-insight)
3. Filtered real promotional email data, selected from a public email classification dataset, filtered to the "promotions" category only and length-thresholded to exclude transactional/spam-style content

Processing steps:

• Deduplication on output text hash
• Removal of examples under 10 words (insufficient training signal)
• Removal of placeholder-variable content (e.g. unfilled {time}, {items} tokens)
• Removal of broken-encoding and non-UTF8 artifacts
• Balanced sampling across task types to avoid over-representation of any single task

Who are the source data producers?

A mix of original dataset authors on HuggingFace (see individual source dataset cards for attribution), AdaptData's automated adaptation system, and manual synthetic data authored for this project.

Personal and Sensitive Information

This dataset does not contain personal, sensitive, or private information. All business names referenced in examples (e.g. local business types, product names) are either genericized, fictional, or based on publicly available product listings with manufacturer specifications only.

Bias, Risks, and Limitations

• Skews toward Western, English-language marketing conventions and consumer/B2B SaaS contexts common in US-based marketing
• Brand voice examples, while diverse in style, are not exhaustive of global marketing tone conventions
• Synthetic examples may carry stylistic patterns from the language model used to assist in their creation
• Product attribute data in examples is illustrative; do not treat as verified real-world product information

Recommendations

Models trained on this dataset should have their outputs reviewed by a human marketer before publication, particularly for factual product claims. Users extending this dataset to other languages or markets should validate cultural and linguistic appropriateness of brand voice and tone examples.

Citation

BibTeX:

@misc{huynh2026marketinginstruct,
  author = {Huynh, Sue},
  title = {marketing-instruct-4k: A Curated Marketing Copywriting Instruction Dataset},
  year = {2026},
  publisher = {HuggingFace},
  howpublished = {\url{https://huggingface.co/datasets/suehuynh/marketing-instruct-4k}},
  note = {AutoScientist Challenge 2026 — Marketing Category}
}

Dataset Card Authors

Sue Huynh — Brown University MSc Data Science

Dataset Card Contact

HuggingFace: suehuynh