Jacquard
Enterprise AI for brand messaging at scale — generate 2,500 variants per brief, predict performance, and optimise across every channel in your brand voice.
Jacquard Review: The Enterprise AI Messaging Platform Built on a Decade of Marketing Language Performance Data
Jacquard was founded as Phrasee in London in 2015 — and spent nine years building something that newer AI marketing tools cannot replicate: a proprietary model trained on a decade of accumulated marketing language performance data across enterprise campaigns. It rebranded as Jacquard in June 2024, expanding its positioning from email subject line optimisation into a comprehensive agentic platform that handles the full lifecycle from brief to deployment across email, SMS, push notifications, web, and app channels. The decade of Phrasee data is the foundational advantage — the model knows what enterprise marketing language performs because it has measured it across billions of sends at brands like TUI Group, eBay, Domino’s, and Virgin Holidays.
Quick Summary
Jacquard is an enterprise agentic AI platform for brand messaging at scale — generating up to 2,500 content variants per brief, predicting performance with 94% accuracy, calibrating each brand’s specific voice and tone, and optimising messaging across email, SMS, push, web, and app in 17 plus languages — used by TUI Group, eBay, Domino’s, and Virgin Holidays across retail, travel, and DTC verticals.
Is it worth using? Yes for enterprise marketing teams running high-volume multi-channel messaging campaigns who want AI that generates on-brand content at the scale and speed that human copywriting teams cannot match, backed by a decade of marketing language performance data.
Who should use it? Enterprise CRM leaders, lifecycle marketing teams, and digital marketing directors at large consumer brands sending millions of marketing messages across email, SMS, and push who want AI-generated content that consistently outperforms human-written alternatives.
Who should avoid it? Mid-market brands whose messaging volume and budget does not justify the $24,000 plus annual investment — Anyword or Klaviyo’s built-in AI testing tools provide more accessible alternatives.
Verdict Summary
Best for
- Enterprise retail, travel, and financial services brands managing multi-channel lifecycle marketing at scale who need AI that generates personalised, on-brand messaging variants at the volume required for meaningful A/B testing without proportionally scaling the copywriting team
- Marketing teams who want performance prediction before send — Jacquard’s 94% prediction accuracy means teams can identify likely winners from generated variants without waiting for full A/B test results
- Global brands who need consistent brand voice across 17 plus languages simultaneously — Jacquard calibrates the brand voice model once and applies it across every language rather than producing generic machine translation
Not for
- Mid-market brands whose email list and send volume does not produce sufficient statistical significance at the scale Jacquard is designed for
- Teams whose primary need is simple email subject line A/B testing that Klaviyo or Braze’s built-in tools handle adequately at no additional cost
- Organisations needing transparent self-serve pricing without an enterprise sales process
Rating
⭐⭐⭐⭐ 4.2 / 5
What Is Jacquard?
Jacquard is an enterprise AI platform for brand messaging founded in London in 2015 as Phrasee — rebranded June 2024 as the company expanded from email optimisation into a comprehensive agentic messaging platform. Its proprietary AI is calibrated by language experts and trained on a decade of marketing performance data, producing content that reflects each brand’s specific voice rather than generic AI output.
The platform’s agentic approach covers the full content lifecycle — receiving a brief, generating up to 2,500 variants calibrated to the brand voice, predicting which variants will perform best using models validated at 94% accuracy, testing in production, and continuously learning from performance outcomes to improve future generation.
How Jacquard Works
- Calibrate your brand voice. Jacquard’s language experts work with the brand to calibrate its specific voice, tone, vocabulary preferences, and compliance rules into the AI model — producing a brand-specific language model rather than generic AI output.
- Input a brief. Brief the AI with the campaign objective, channel, audience segment, and creative direction. Jacquard generates up to 2,500 content variants from the brief — covering the full range of tone, emotional appeal, and structural approaches within brand voice parameters.
- Review performance predictions. Jacquard predicts the performance of generated variants before sending — models validated at 94% accuracy identify likely winners from the variant set without requiring full A/B test results to determine which content will perform best.
- Deploy across channels. Approved variants deploy across email, SMS, push notifications, web, and app — adapting content optimally for each channel’s specific performance characteristics.
- Personalised Campaigns at scale. Jacquard’s Personalised Campaigns product generates individually tailored messaging for each customer segment — moving beyond single winning variants to truly personalised lifecycle messaging at enterprise volume.
- AI learns from every send. Performance data from every campaign feeds back into the brand’s calibrated model — improving generation quality and prediction accuracy with each campaign cycle.
Key Features
- Up to 2,500 content variants generated per brief within the brand’s calibrated voice
- 94% performance prediction accuracy — identifying winning variants before send without waiting for full A/B results
- Brand voice calibration by language experts — brand-specific model, not generic AI output
- Personalised Campaigns product for individually tailored lifecycle messaging at enterprise scale
- Multi-channel coverage — email, SMS, push notifications, web, and app
- 17 plus languages with brand voice consistency maintained across all markets
- Agentic platform covering brief to deployment lifecycle
- ISO 27001 certified — client data not used to train shared AI models
- GDPR compliant through published Data Processing Addendum
- Continuous learning from campaign performance — improving with every send
- Enterprise clients including TUI Group, eBay, Domino’s, and Virgin Holidays
- Built on a decade of proprietary marketing language performance data from the Phrasee era
Real-World Use Cases
- Email variant scale: A retail brand’s email marketing team previously briefed a copywriter for 5 to 10 subject line variants per campaign. Jacquard generates 200 plus variants per brief — covering the full range of emotional appeals and tonal approaches that human copywriting cannot produce at volume — with performance predictions identifying the likely winners before send.
- Lifecycle messaging personalisation: TUI Group uses Jacquard’s Personalised Campaigns to generate individually tailored travel messaging for different customer segments — holiday dreamers, recent bookers, and loyalty programme members each receive messaging calibrated to their specific relationship with the brand at enterprise volume.
- Multi-language campaign consistency: A financial services brand uses Jacquard to generate campaign messaging in English, French, German, Spanish, and Italian — the brand voice model maintains consistent tone and compliance standards across all five languages while optimising independently for each market’s audience response patterns.
- Performance prediction before send: An e-commerce brand uses Jacquard’s performance prediction to identify the top-performing variants from a 500-variant set before committing to a full send — selecting the predicted winners for the main send and using lower-confidence variants for smaller test segments.
Pros and Cons
| Pros | Cons |
|---|---|
| Decade of proprietary marketing language performance data from the Phrasee era — advantage newer tools cannot replicate | Enterprise-only investment — $24,000 plus annually makes it inaccessible below a meaningful scale threshold |
| 94% performance prediction accuracy reduces the time and cost of A/B testing at enterprise scale | Jacquard rebrand from Phrasee still causing confusion — some resources and integrations still reference the old name |
| Up to 2,500 variants per brief covers the range of copy approaches that human teams cannot generate at volume | No self-serve access or transparent pricing — enterprise sales process required for evaluation |
| Brand voice calibrated by language experts — genuine brand compliance, not generic AI output | ISO 27001 certified but SOC 2 Type II not confirmed in published security documentation |
| Multi-channel and 17 plus language coverage from a single brand-calibrated model | Valuation and revenue figures suggest a smaller company than some enterprise buyers prefer for long-term vendor stability |
Pricing & Plans
Jacquard pricing is enterprise and not publicly listed. Based on reported market data, annual contracts start at approximately $24,000 plus and scale with send volume, channels, and number of brand markets. Contact jacquard.com for a custom quote.
Best Alternatives & Comparisons
- Persado — Better for the largest enterprise with the deepest validated performance dataset at JPMorgan Chase and American Express scale
- Anyword — Better for accessible mid-market AI copywriting with performance scoring at transparent self-serve pricing
- Copy.ai — Better for broad AI content generation across all marketing content types at lower pricing
- Jasper — Better for marketing teams wanting AI writing across long-form content alongside short-form copy
Frequently Asked Questions (FAQ)
Jacquard is an enterprise AI brand messaging platform — formerly known as Phrasee, rebranded June 2024 — that generates up to 2,500 content variants per brief, predicts performance with 94% accuracy, and optimises messaging across email, SMS, push, web, and app in 17 plus languages.
Yes — Phrasee officially rebranded as Jacquard on 12 June 2024. Same company, same team, same technology under a new name and expanded platform positioning. phrasee.co redirects to jacquard.com.
Jacquard pricing is enterprise and not publicly listed. Annual contracts are reported to start at approximately $24,000 plus, scaling with send volume and channels. Contact jacquard.com for a quote.
Jacquard’s AI predicts the performance of generated content variants before deployment — identifying likely winners without requiring full A/B test results. The 94% prediction accuracy figure reflects validation of the performance prediction model against actual campaign outcomes.
Jacquard supports 17 plus languages with brand voice consistency maintained across all markets — generating campaign messaging in multiple languages from the same brand-calibrated model rather than producing generic machine translation.
Both are enterprise AI marketing language platforms built on proprietary performance data. Persado has the largest validated dataset at 1 billion plus experiments and the most documented enterprise results at JPMorgan Chase scale. Jacquard has a decade of Phrasee performance data, the highest variant generation volume at 2,500 per brief, and the Personalised Campaigns product for individual lifecycle messaging at scale. Both warrant enterprise evaluation — the choice depends on channel focus and existing platform integrations.
Final Recommendation
Jacquard is the most mature enterprise AI brand messaging platform built specifically on marketing language performance data — and the decade of Phrasee data behind its models is a genuine advantage over platforms built on general AI adapted for marketing use. For enterprise marketing teams whose campaigns run at the scale where small percentage improvements in open, click, and conversion rates produce significant revenue impact, Jacquard delivers the variant volume, performance prediction, and brand voice consistency that makes those improvements systematic rather than occasional.
Next steps
