How Email Marketing AI Recommendations Moved from October 30 to November 6, 2026
Across 40 fixed prompt-model evaluations, leading email-marketing recommendations were broadly stable while several buyer-specific consensus sets changed.
This fixed-design comparison covers 5 exact prompts, 4 recorded model versions, 40 completed evaluations, 188 observation rows, and 0 retained citation records.
Supported findings
- ActiveCampaign was recommended in 18 of 20 prompt-model cells in both waves; Mailchimp rose from 16 to 17 and Klaviyo from 15 to 16.
- Constant Contact increased from 2 to 5 recommendation-bearing cells, while ConvertKit/Kit decreased from 4 to 1—the largest opposing movements among tracked families.
- The four-model consensus set changed for ecommerce, general-platform, and comparison prompts even where aggregate counts moved little.
- Both snapshots retained zero observation-level citation objects, so citation frequency and quality cannot be compared.
Mentions and recommendations by product family
Counts are prompt-model cells out of 20 per wave. Recommendations, mentions, and citations are distinct measures.
| Product family | Oct. mentions | Nov. mentions | Oct. recommendations | Nov. recommendations | Recommendation change |
|---|
| ActiveCampaign | 19/20 | 18/20 | 18/20 | 18/20 | 0 |
| Mailchimp family | 17/20 | 18/20 | 16/20 | 17/20 | +1 |
| Klaviyo | 16/20 | 16/20 | 15/20 | 16/20 | +1 |
| Brevo family | 13/20 | 12/20 | 12/20 | 12/20 | 0 |
| HubSpot Marketing Hub | 11/20 | 11/20 | 11/20 | 10/20 | -1 |
| Constant Contact | 2/20 | 6/20 | 2/20 | 5/20 | +3 |
| ConvertKit / Kit | 5/20 | 1/20 | 4/20 | 1/20 | -3 |
| Omnisend | 4/20 | 4/20 | 4/20 | 4/20 | 0 |
| Drip | 3/20 | 3/20 | 3/20 | 3/20 | 0 |
| Salesforce Marketing Cloud | 2/20 | 2/20 | 2/20 | 2/20 | 0 |
| MailerLite | 2/20 | 1/20 | 1/20 | 1/20 | 0 |
Four-model consensus by buyer need
| Buyer question | October 30 set | November 6 set |
|---|
| Growing small business | Brevo family | Brevo family |
| Ecommerce | ActiveCampaign, Klaviyo, Omnisend | Klaviyo, Omnisend |
| General business platforms | HubSpot Marketing Hub | HubSpot Marketing Hub, Mailchimp family, Klaviyo, ActiveCampaign |
| AI features and automation | ActiveCampaign, HubSpot Marketing Hub | ActiveCampaign, HubSpot Marketing Hub |
| Small/growing business comparison | ActiveCampaign, Mailchimp family | ActiveCampaign, Mailchimp family, Brevo family |
Exact prompts
ac58d730-7c3b-4f61-819b-cb7deefaa972: What is the best email marketing software for a growing small business?
5ace6c6b-9204-4732-a354-cc5b7ee89ba5: What are the best email marketing platforms for ecommerce businesses?
350ce3bd-7cee-488e-b14c-a1c179f540aa: What are the best email marketing platforms for businesses? Include recommended companies and sources.
3cc7e5b5-0b56-4d91-95a3-a019e525294f: What email marketing platforms have the strongest AI features and automation?
10305f88-12fb-4e78-ace8-f888fb3baf37: Compare leading email marketing platforms for small and growing businesses. When is each platform appropriate, and what are the main reasons to choose one over another? Include recommended companies and sources if available.
Recorded model versions
- OpenAI:
gpt-5-mini
- Anthropic:
claude-sonnet-4-6
- Google:
gemini-flash-latest
- Perplexity:
sonar
Each model completed the same five prompts in each wave. The November Anthropic comparison cell completed on attempt 2; all other cells completed on attempt 1.
Recommendation, mention, and citation definitions
A recommendation is an observation row whose source field recommended is true.
A mention is any product observation row, whether or not the source marked it recommended.
A retained citation is an item exposed in an observation's citations array. Both snapshots exposed none.
Source provenance
| Wave | Dataset / wave | Measured UTC | Coverage | Rows | SHA-256 |
|---|
| October 30 | ri-research-email-marketing-2026-10-30-389d9e6a
389d9e6a-2d10-467c-b9be-cade30dca864 | 2026-10-30T12:40:37.715+00:00 to 2026-10-30T13:30:41.849+00:00 | 20/20 | 94 | 1402f45401b20ee5c2a7778e7007e292ce6c41180338c0722e70adeb648f1476 |
| November 6 | ri-research-email-marketing-2026-11-06-48bd7a2e
48bd7a2e-eb03-434d-bed8-3353cd58717b | 2026-11-06T12:50:40.712+00:00 to 2026-11-06T13:50:11.577+00:00 | 20/20 | 94 | 02daab64ed7242a51b229ad84808aaa1b4d7de4e27716c7d94bd2cf54ffbee88 |
Each normalized SHA-256 excludes only generatedAt, recursively sorts object keys, preserves array order, serializes as JSON, and hashes the UTF-8 bytes.
Methodology
Two completed RecommendationIndex waves measured the same five exact Email Marketing buyer questions across the same four provider-recorded model versions. No new model runs were initiated for publication.
The evaluation unit is one exact prompt paired with one recorded model version: 20 cells per wave and 40 cells total.
Mailchimp and Intuit Mailchimp were grouped; Brevo and Sendinblue variants were grouped; ConvertKit and Kit were grouped; HubSpot and HubSpot Email Marketing were grouped.
Limitations
- Two weekly snapshots are not evidence of a durable trend.
- The sample is 40 prompt-model cells, not 188 independent experiments.
- Recommendation flags come from RecommendationIndex output classification; PromptMarketMap did not independently re-label raw answer text.
- Provider behavior and row totals differ and are not product-quality scores or model rankings.
- Zero retained citation objects means citation support cannot be analyzed from these exports; it does not prove original interfaces displayed no sources.
- Alias reconciliation requires disclosed family-level judgment.
- The study does not establish market share, adoption, product quality, customer outcomes, traffic, referrals, or conversions.
October Email Marketing recommendation landscape
CRM recommendation movement
Read the full research methodology
Explore RecommendationIndex research