Skip to content
Jasper logo

Jasper

Agentic AI marketing platform for brand-governed campaigns, content, and workflows.

WritingMarketingAgents

How it works / How to use

Agentic AI marketing platform for brand-governed campaigns, content, and workflows. Set up Brand Voice and Knowledge, then draft in Canvas or Chat with marketing guardrails. Purpose-built marketing agents and Grid automate campaign workflows on higher plans; advanced and custom agents are on Business. Public API access is a Business-plan capability, not Pro.

  1. Configure Brand Voice, Knowledge, and any campaign context Jasper should treat as authoritative.
  2. Open Canvas for structured drafts or Chat for iterative campaign work; pick a marketing agent when one fits the job.
  3. Provide audience, channel, offer, and constraints—then edit against brand rules and verified facts.
  4. For repeat campaigns, use Grid on supported plans; use the API on Business for programmatic generation.

Brand-governed drafts in Canvas

Open Canvas, state the asset type first (landing hero, ad set, email), then paste source facts under a labeled block. Ask Jasper to follow Brand Voice and flag anything not supported by Knowledge.

What to provide

  • Campaign brief: audience, offer, channel, and goal
  • Brand Voice and Knowledge already configured in Jasper
  • Required claims, dates, and words that must not be invented

Details that improve the result

  • Name the channel constraints (character limits, CTA, forbidden claims)
  • Say which facts are fixed vs flexible
  • Request options when you want variations, not one long draft

Example prompt

Canvas task: landing page hero for a B2B analytics trial. Audience: ops managers at 50–500 person companies. Tone: confident, plain language. Must include '14-day trial' and 'no credit card'. Do not invent customer counts or awards. Output: headline, subhead, primary CTA, 3 supporting bullets.

If the first output is not good

  • “Keep headline; shorten subhead by 20%.”
  • “Regenerate bullets only; more concrete, less jargon.”
  • “Add two alternate CTAs under 20 characters.”

Common mistakes

  • Skipping Brand Voice setup and expecting on-brand output
  • Asking for metrics or testimonials not in Knowledge
  • Mixing email, ads, and blog sections in one unfocused Canvas run

Campaign iteration in Chat

Paste the existing copy under “Current draft:”, then give one clear edit instruction. Chat works well for iterative marketing rewrites with Brand Voice applied.

What to provide

  • The current draft or outline
  • What changed in the campaign (offer, audience, channel)
  • Edit goal: shorten, localize, change tone, or add variants

Details that improve the result

  • Reference Brand Voice explicitly on sensitive copy
  • Ask for diff-style changes when the draft is close
  • Separate facts (paste) from instructions (above)

Example prompt

Using our Brand Voice, revise this launch email for re-engaged free users. Keep pricing facts exactly as in SOURCE. Make the opening more specific to inactive users. Length under 220 words.

SOURCE:
14-day Pro trial, ends March 31.

Current draft:
[paste email]

If the first output is not good

  • “Rewrite subject lines only; 3 options under 45 characters.”
  • “More direct CTA; do not change paragraph 3.”
  • “List assumptions you made about the audience.”

Common mistakes

  • Restarting from scratch when a small section needs change
  • Letting Jasper rewrite verified pricing or dates
  • Asking for every channel at once in one chat turn

Ads and paid social copy

Lead with platform and limits, then offer and audience. Ask for numbered variants so you can pick and test.

What to provide

  • Platform (Meta, LinkedIn, Google, etc.) and ad format
  • Offer, audience, and single primary action
  • Character or word limits and disallowed claims

Details that improve the result

  • One angle per batch (pain, outcome, social proof) yields cleaner variants
  • Include exact promo code or date rules when they must appear
  • Say whether emojis, questions, or urgency are on-brand

Example prompt

LinkedIn single-image ad copy. Audience: finance directors at mid-market SaaS. Offer: webinar on cash-flow forecasting. Limits: intro text 150 chars, headline 70, description 100. No guaranteed ROI claims. Return 3 variant sets labeled A/B/C.

If the first output is not good

  • “Variant B is closest; give 3 new headlines only.”
  • “Remove urgency language; keep factual tone.”
  • “Add one variant focused on time savings, not cost.”

Common mistakes

  • Ignoring platform character limits until after generation
  • Multiple offers in one ad set prompt
  • Invented performance stats or client logos

Email sequences and nurture flows

Outline the sequence table first (email #, goal, send timing), then generate one email at a time or ask for the full outline before body copy.

What to provide

  • Sequence length and trigger (signup, trial day 3, cart abandon)
  • Facts for each email: offer, deadline, links, product names
  • Tone progression across the sequence

Details that improve the result

  • Keep product and pricing language consistent with Knowledge
  • Say whether emails should stand alone or reference prior messages
  • Request subject + preview + body for each step

Example prompt

3-email trial nurture. Day 0 welcome, Day 3 feature highlight, Day 10 trial ending reminder. Facts: product name 'Northwind Analytics', trial length 14 days, upgrade link /pricing. Brand Voice: helpful, not pushy. Output table: Day, Subject, Preview, Body outline; then write Email 1 fully.

If the first output is not good

  • “Write Email 2 full body; keep subject lines under 50 chars.”
  • “Email 3 is too salesy; soften CTA, keep deadline fact.”
  • “Add one plain-text alternative CTA line.”

Common mistakes

  • Contradictory offers across emails in one sequence
  • Missing send-day context so CTAs repeat awkwardly
  • Fabricated case studies or uplift percentages

Blog and long-form marketing content

Ask for an outline with H2/H3 first, confirm structure, then generate section by section with Knowledge cited in your brief—not invented by Jasper.

What to provide

  • Target keyword or topic angle (not a vague theme)
  • Outline approval before full draft, if possible
  • Sources, product facts, and internal links to include

Details that improve the result

  • Provide an audience level (beginner vs practitioner)
  • Say what must not be claimed (medical, legal, performance guarantees)
  • Keep brand terminology consistent with Knowledge

Example prompt

Outline a 1,200-word blog for marketing ops leads: 'Reduce reporting time without new headcount'. Use ONLY facts in KNOWLEDGE. No customer quotes unless listed. Return outline first; wait for approval before drafting.

KNOWLEDGE:
[paste product capabilities and approved stats]

If the first output is not good

  • “Expand section 2 with one concrete workflow example.”
  • “Shorten intro; keep H2s unchanged.”
  • “Flag any claim not grounded in KNOWLEDGE.”

Common mistakes

  • Full draft in one shot without outline review
  • SEO stuffing that breaks Brand Voice
  • Statistics not present in provided Knowledge

Purpose-built marketing agents

Pick a marketing agent when Jasper offers one for the job. State goal, inputs, constraints, and output template. Advanced and custom agents are on Business—confirm plan before relying on them.

What to provide

  • The agent that matches the task (content, campaign, research-style marketing prep)
  • Inputs the agent expects (brief, URL, file, brand context)
  • Deliverable format and approval gates

Details that improve the result

  • Agents work best with a single campaign outcome
  • Attach Knowledge sources the agent should not override
  • Review agent plans before publishing or scheduling

Example prompt

Use the campaign agent to turn this brief into a launch checklist plus three channel-specific message maps (email, LinkedIn, sales talk track). Constraints: no pricing changes, no new product names, mark anything not in Knowledge as 'needs verification'.

If the first output is not good

  • “Regenerate only the sales talk track.”
  • “Add a risk section for claims that need legal review.”
  • “Shorter checklist; owners as TBD placeholders only.”

Common mistakes

  • Expecting Business-only custom agents on Pro
  • Multi-campaign goals in one agent run
  • Skipping human review before external publish

Grid (campaign automation)

Grid on Business automates bulk campaign assets with Workflow Agent columns. Define columns/fields, map them to templates, and validate one row before scaling.

What to provide

  • Repeatable campaign pattern (weekly newsletter, product drop, regional variants)
  • Input table or source fields per row
  • Fixed brand elements vs row-specific variables

Details that improve the result

  • Lock boilerplate disclaimers and legal lines
  • Use Brand Voice at the template level
  • Review first batch output before generating hundreds of rows

Example prompt

Grid: 20 regional webinar invites. Columns: city, date, speaker_name, registration_url. Fixed closing paragraph from Legal. Generate subject + body per row; do not invent speakers or dates not in the sheet.

If the first output is not good

  • “Regenerate rows 4 and 11 only; speaker names were wrong.”
  • “Add a column for preview text; max 90 chars.”
  • “Pause pipeline until Brand Voice v2 is applied.”

Common mistakes

  • Batch generating before single-row proof
  • Variable fields that overflow email templates
  • Automating publish without human spot-check

API and programmatic copy (Business)

Public API access is a Business-plan capability, not Pro. Send structured briefs (audience, channel, facts JSON), receive drafts, then run your own brand and legal checks before shipping.

What to provide

  • Business plan with public API access
  • Template or prompt contract per asset type
  • Brand and compliance validation on your side

Details that improve the result

  • Keep Knowledge facts in the API payload, not only in the UI
  • Version prompts/templates when campaigns change
  • Log outputs for regulated industries

Example prompt

API request: generate meta_description for URL slug provided. Payload includes { product_name, primary_keyword, forbidden_terms[], max_length: 155 }. Return one string; no HTML.

If the first output is not good

  • “Tighten prompt to forbid superlatives.”
  • “Add server-side check for forbidden_terms.”
  • “Fallback template if API times out.”

Common mistakes

  • Building on API while on Pro
  • No post-generation validation of dates, prices, or claims
  • Changing prompt shape without versioning consumers

How to prompt

Jasper is built for marketing workflows, not general chat. Lead with asset type, audience, channel, and constraints; ground facts in Knowledge; invoke Brand Voice explicitly. Iterate in Chat or Canvas, use agents and Grid for scaled campaign work, and reserve API generation for Business integrations.

Asset: homepage hero.
Audience: IT managers at mid-market firms.
Brand Voice: on.
Facts (must appear): 'SOC 2 Type II', '14-day trial'.
Forbidden: ROI guarantees, competitor names.
Output: headline (max 8 words), subhead (max 20 words), CTA button label.

Configure Brand Voice and Knowledge first

Jasper’s marketing drafts improve when brand rules and approved facts live in Brand Voice and Knowledge—not only in each prompt.

Example

Before drafting, upload approved product sheet to Knowledge. Prompt: 'Draft ad copy using Brand Voice; cite only KNOWLEDGE for product claims.'

Name the channel and limits up front

Character counts, CTA rules, and forbidden claims prevent rework on ads, email, and social.

Example

Google RSA: 15 headlines ≤30 chars, 4 descriptions ≤90 chars. Offer: free trial. No superlatives.

Separate SOURCE from instructions

Paste pricing, dates, and legal lines under a labeled block; tell Jasper not to rewrite them.

Example

Use SOURCE verbatim for pricing.

SOURCE:
$69/mo Pro, billed monthly.

Task: write FAQ answer about billing.

Work outline → section → polish

For long-form marketing content, approve structure before full prose to keep Brand Voice consistent.

Example

Step 1: H2 outline only. Step 2: expand section 2. Step 3: tighten intro after body is approved.

Use agents and Grid for scale, Chat for edits

Agents and Grid suit repeatable campaign patterns; Chat and Canvas suit refinement and one-off assets.

Example

Grid generates 50 localized invites; Chat rewrites row 12 tone; agent builds launch checklist once.

Best output tips

Invest in Brand Voice and Knowledge

Jasper’s value is brand-aware marketing drafts. Load voice guidelines, product facts, and approved claims into Brand Voice and Knowledge so Canvas, Chat, and agents share the same guardrails.

Prompt like a creative brief

State audience, channel, offer, objective, length, and exclusions. Marketing output quality tracks brief clarity more than clever phrasing.

Never invent campaign facts

Dates, prices, statistics, awards, and customer quotes must come from your brief or Knowledge. Ask Jasper to flag unsupported claims.

Pick the right Jasper surface

Canvas for structured assets, Chat for iterative edits, marketing agents for guided workflows, and Grid for repeatable generation at scale on Business.

Know plan gates

Purpose-built marketing agents include advanced and custom options on Business. Public API access is Business-only, not Pro. Confirm plan before building automations.

Generate variants deliberately

For ads and subject lines, request labeled sets (A/B/C) with platform limits. Change one angle per batch for cleaner testing.

Outline long-form before drafting

Approve H2/H3 structure for blogs and landing pages, then expand section by section. This keeps Brand Voice steady and reduces off-brand tangents.

Automate with proof rows

Grid excels at repeated campaign assets on Business. Validate one output, then scale—especially when rows include names, dates, or URLs.

Iterate with surgical follow-ups

When a draft is close, ask to change one section, tone dimension, or length constraint instead of restarting the entire asset.

Review before publish

Treat Jasper output as a strong first draft. Human review is essential for compliance language, competitor mentions, and numeric accuracy.

  • Set up Brand Voice and Knowledge before drafting—Jasper is built for brand-governed marketing copy.
  • Open each request with asset type, audience, channel, length, and forbidden claims.
  • Paste approved facts under a labeled SOURCE block; tell Jasper not to invent metrics or testimonials.
  • Use Canvas for structured assets; Chat for iterative rewrites; marketing agents when one matches the job.
  • Advanced and custom agents and public API access are Business capabilities—not Pro.
  • Grid on Business automates repeat campaign patterns; proof one row before batching.
  • Ask for numbered variants on ads and subject lines so you can test systematically.
  • Iterate with targeted follow-ups instead of regenerating entire campaigns from scratch.
  • Mark anything not in Knowledge as ‘needs verification’ before publishing.
  • Run brand and legal review on final copy—especially pricing, comparisons, and performance claims.

Try this AI

Try Jasper

Product Details

Pricing, features, limits and latest updates

Jasper

Agentic AI marketing platform for brand-governed campaigns, content, and workflows.

Paid · $69 / month

Pricing Plans

Pro

$69/ Monthly

Pro

$708/ Yearly

Business

Custom

Key Features

Chat

Brand-governed drafts in Canvas and Chat using Brand Voice, Knowledge, and Audiences; Grid bulk workflows and API are Business capabilities.

API

Public API access is a Business-plan capability, not Pro.

Automation

Grid bulk workflows on Business automate campaign assets; Workflow Agent columns map templates to spreadsheet rows.

Agents

Purpose-built marketing agents; advanced and custom agents are on Business.

Limits

  • Public API: Public API access is a Business-plan capability, not Pro (Jasper developer docs).

Ideas / Prompt experiences

Share a prompt that worked for you. Username and email are shown with your submission. External links are not allowed.

Example prompt

Canvas brief with Brand Voice

Prompt

Project: Spring launch email series. Audience: existing SMB customers. Asset: 3 subject lines + 180-word email body. SOURCE (approved facts only): [paste product facts] Constraints: no invented metrics; formal but plain tone.

Short explanation

Jasper IQ (Brand Voice, Knowledge, Audiences) should be set before drafting; separate SOURCE from instructions.

Example prompt

Chat iteration on one paragraph

Prompt

Keep paragraphs 1 and 3. Rewrite paragraph 2 to mention migration downtime in plain language. Do not add pricing or customer quotes not in Knowledge.

Variation

Give me two CTA button labels under 20 characters.

Short explanation

Chat requires a Project; iterate with targeted follow-ups instead of regenerating entire campaigns.

Share your experience

Required fields are marked. Variation, result, and explanation are optional.

ChatGPT logo

ChatGPT

Explore

Conversational AI for writing, coding, research, files, images, and automation. Capabilities and limits vary by plan.

WritingFree / Paid

Claude logo

Claude

Explore

AI assistant for writing, analysis, coding, and collaborative problem-solving.

WritingFree / Paid

Perplexity logo

Perplexity

Explore

An answer engine that searches the web and cites its sources.

WritingFree / Paid