<!-- Source of truth: app/samples/mailchimp-lite-2026-10-15-v2.md and app/samples/mailchimp-lite-2026-10-15-v2.json (re-derived 2026-10-15T16:22:41Z from the stored answers of the run created 2026-10-15T16:18:34.811Z = 12:18 PM ET). Every number below re-verified against v2. Comments that say ".md" or ".json" mean the -v2 files. -->
What share of voice in AI answers looks like for a well-known brand
Share of voice in AI answers is the part of all brand mentions, across a set of buyer questions, that goes to you. When we asked AI answer engines about email marketing platforms, Mailchimp, one of the best-known names in the category, was named in 71% of the answers to questions that didn't use its name, and got 25% of the share of voice.
So even a category leader is left out of some answers, and it shares the rest with four or five rivals. Below is what we measured, how to read it, and what it means for a smaller business.
About this example. This is an independent sample we made with public API answers on Oct 15, 2026. Answer Checkup is not affiliated with Mailchimp or with any other brand named in this post, and none of them asked for or reviewed it.
What we asked
We used 25 buyer-style questions about email marketing platforms. 19 didn't name Mailchimp, like "What is the best email marketing platform for a small business?" or "Which email marketing platform is best for automation?" 6 did, like "What is Mailchimp?" and "How much does Mailchimp cost?"
<!-- source: .json methodology[0]: "25 buyer-style prompts (19 that do not name the brand, 6 that do)"; prompt text from .md "Prompts and answers" rows 1, 27, 39, 47 -->
Each question was asked once on the OpenAI API with web search and once on the Perplexity API. That's 50 answers, 38 of them from the questions that don't name the brand.
<!-- source: .json module3.answers = 50; module3.shareOfVoice.unbranded.answers = 38 -->
We tracked five competitors: Klaviyo, Constant Contact, Brevo, HubSpot and ActiveCampaign.
<!-- source: .md "Competitor source: buyer-provided ... Tracked: Klaviyo, Constant Contact, Brevo, HubSpot, ActiveCampaign" -->
Presence and share of voice are two different numbers
Presence is how often you're named at all. Mailchimp was named in 27 of the 38 unbranded answers, so its presence was 71%.
Share of voice compares you with the others. Across the 38 unbranded answers, the six tracked names were mentioned 110 times in total (one count per answer per name). Mailchimp had 27 of those 110, or 25%.
<!-- source: .md Summary "named in 27 of 38 answers ... (71% presence, 25% share of voice among tracked names)"; .json rows answers 27+22+18+18+18+7 = 110; Mailchimp shareOfVoice 0.245, presenceRate 0.711 -->
Presence tells you whether you're in the conversation. Share of voice tells you how crowded it is.
| Name | Unbranded answers naming it | Share of voice | Presence | Named first |
|---|---|---|---|---|
| Mailchimp | 27 | 25% | 71% | 11 |
| Brevo | 22 | 20% | 58% | 8 |
| Klaviyo | 18 | 16% | 47% | 6 |
| HubSpot | 18 | 16% | 47% | 3 |
| ActiveCampaign | 18 | 16% | 47% | 5 |
| Constant Contact | 7 | 6% | 18% | 1 |
<!-- source: .md table "Share of voice, unbranded prompts" (38 answers); .json module3.shareOfVoice.unbranded.rows (answers, shareOfVoice, presenceRate, firstPlace) -->
"Named first" counts the answers where that name came before the other tracked names.
What stands out
A leader still misses answers. Mailchimp wasn't named in 11 of the 38 unbranded answers. A question like "What should I look for in an email marketing platform?" got advice, not brand names: in both answers to it, none of the six tracked names came up.
<!-- source: 38 - 27 = 11; .md "Prompts and answers" rows 9 and 10: brand no, competitors named "-" -->
The gap to second place is small. Brevo was named in 22 answers, five fewer than Mailchimp. Three more brands were tied at 18. In AI answers, a well-known brand often shares the list with several others rather than being the only pick.
<!-- source: .md share of voice table: Brevo 22, Klaviyo/HubSpot/ActiveCampaign 18 -->
Being named and being named first are not the same. Mailchimp was named first in 11 answers. Its average position when named was 2.4, so it was often the second or third tracked name in the list.
<!-- source: .json Mailchimp firstPlace 11, avgPosition 2.4 -->
Branded questions are a different test. When the question used the name, Mailchimp was named in 100% of the answers. That tells you the engines know the brand. It doesn't tell you whether they suggest it to someone who doesn't know it yet. Unbranded questions are where new buyers come from.
<!-- source: .json module3.brandPresence.branded = 1; .md "Brand presence: 71% of unbranded answers, 100% of branded answers, 78% overall (50 answers)" -->
Names nobody was tracking
Share of voice only counts the names you choose to track. The answers also named other email tools that weren't on the list. These ones each appeared in 3 or more answers:
| Name | Answers naming it |
|---|---|
| Kit | 11 |
| MailerLite | 11 |
| Omnisend | 5 |
| Sender | 4 |
| GetResponse | 3 |
| Zoho Campaigns | 3 |
<!-- source: mailchimp-lite-2026-10-15-v2.md "Other names that appeared in 2+ answers (heuristic extraction, not in share of voice): Kit (11), MailerLite (11), Omnisend (5), Sender (4), GetResponse (3), Zoho Campaigns (3)"; v2 .json module3.untrackedNames identical. The files do not say whether these counts cover all 50 answers or only the 38 unbranded ones, so the copy does not compare them with the tracked table. Builder says Kit 11 includes ConvertKit mentions; the files do not state that, so it is not claimed. -->
Kit and MailerLite were each named in 11 answers. These extra names come from a heuristic extraction, so treat them as a lead, not an exact count, and don't compare them one to one with the tracked table above. The lesson is practical: if you only track the competitors you already know about, you can miss names the engines bring up on their own. Read a sample of answers before you settle on your list.
What this means for a smaller business
If a brand this well known is named in 71% of unbranded answers, a smaller business should not expect to be named in most of them. A more useful goal is to move from never named to sometimes named, for the questions closest to what you sell.
Three things to take from the example:
- Track unbranded questions. Questions with your name in them will flatter you. The ones without it show whether an engine would suggest you to a stranger.
- Pick your competitors deliberately. Share of voice depends on who you count. Track the four or five names your buyers actually compare you with.
- Measure a rate, not a single answer. One answer can go either way. Look at how often you're named across many questions, and re-check over time with the same questions.
To check this by hand, our guide on whether ChatGPT recommends your business walks through the steps. For what tends to help you get named, see AI SEO and GEO.
Limits of this example
- One run is a snapshot. Each question was asked once per engine. Run it again and some answers will change.
- The API is not the app. We measured through the OpenAI API with web search and the Perplexity API. Those approximate the consumer apps but are not identical.
- Matching is heuristic. A name counts only when the answer uses it with matching capitalization and doesn't just repeat the question. Position is the order in which tracked names first appear.
- Only tracked names count. Share of voice here covers Mailchimp and the five tracked competitors. Other tools also appeared in answers and are not in these numbers (see below).
- This is not a ranking of products. It shows what the answers said on one day, not which tool is better.
<!-- source: .json methodology[1]-[3] (snapshot, API vs app, matching and position rules) -->
FAQ
- What is share of voice in AI answers?
- It's your share of all the brand mentions across a set of buyer questions. If the tracked names were mentioned 110 times and you were named 27 times, your share of voice is 25%.
- What's the difference between presence and share of voice?
- Presence is how often you're named at all, out of all answers. Share of voice is how your mentions compare with everyone else's.
- What's a good share of voice?
- There's no standard. It depends on how many competitors you track and how crowded your category is. Compare yourself with your own past results and with the names your buyers consider.
- Why use questions that don't name my brand?
- Because new buyers don't know your name yet. Questions that include it mostly test whether the engines recognize you.
- Will my share of voice be the same next week?
- Probably not exactly. Answers vary from run to run, so treat any single result as a reading, not a fixed score.
Sources
- Answer Checkup Lite sample for mailchimp.com, run Oct 15, 2026 at 12:18 PM ET: 25 questions, 50 answers through the OpenAI API with web search and the Perplexity API. An independent sample, not affiliated with or reviewed by Mailchimp.
Check your own site
Answer Checkup is not affiliated with OpenAI, Perplexity, Mailchimp, Brevo, Klaviyo, HubSpot, ActiveCampaign, Constant Contact, Kit, MailerLite, Omnisend, Sender, GetResponse or Zoho Campaigns.
Questions: [email protected]