In-app messaging and customer communications
Asked as “in-app messaging platform”, and as “customer communications tool for product messaging”, on behalf of a mid-market B2B software company. 58 first choices recorded across the direct, paraphrase, budget and scale prompts, six models each.
Standing · first-choice share
21%
Contested · OneSignal 17%
21Intercom17OneSignal16Userpilot47others
21% of first choices, contested.
Since September 2026▲+2Since September 2026: 20% → 22%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.Intercom held the lead, +2 points on 20%, inside the 12-point floor.
The standing
Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all six models, for a mid-market B2B software company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrantSince September 2026
01
Intercom21%17%47accepted challenger▲+2Since September 2026: 20% → 22%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.20% → 22%
Intercom21%17%47accepted challenger▲+2Since September 2026: 20% → 22%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.20% → 22%02
OneSignal17%3%34accepted challenger▲+7Since September 2026: 11% → 18%, +7 points. Inside the 12-point floor: within noise. Read over the models both editions asked.11% → 18%
OneSignal17%3%34accepted challenger▲+7Since September 2026: 11% → 18%, +7 points. Inside the 12-point floor: within noise. Read over the models both editions asked.11% → 18%03
Userpilot16%7%27accepted challenger▼−2Since September 2026: 16% → 14%, −2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.16% → 14%
Userpilot16%7%27accepted challenger▼−2Since September 2026: 16% → 14%, −2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.16% → 14%04
Appcues9%9%35accepted challenger▲+6Since September 2026: 2% → 8%, +6 points. Inside the 12-point floor: within noise. Read over the models both editions asked.2% → 8%
Appcues9%9%35accepted challenger▲+6Since September 2026: 2% → 8%, +6 points. Inside the 12-point floor: within noise. Read over the models both editions asked.2% → 8%05
Pendo3%11%27accepted challenger▲+2Since September 2026: 2% → 4%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.2% → 4%
Pendo3%11%27accepted challenger▲+2Since September 2026: 2% → 4%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.2% → 4%06
Knock3%0%11accepted challenger▲+4Since September 2026: 0% → 4%, +4 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 4%
Knock3%0%11accepted challenger▲+4Since September 2026: 0% → 4%, +4 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 4%07
Braze2%12%25accepted challenger=heldSince September 2026: 2% → 2%, ±0 points. Inside the 12-point floor: within noise. Read over the models both editions asked.2% → 2%
Braze2%12%25accepted challenger=heldSince September 2026: 2% → 2%, ±0 points. Inside the 12-point floor: within noise. Read over the models both editions asked.2% → 2%08
CometChat2%6%16accepted challenger▲+2Since September 2026: 0% → 2%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 2%
CometChat2%6%16accepted challenger▲+2Since September 2026: 0% → 2%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 2%09
Firebase2%18%11accepted challenger▲+2Since September 2026: 0% → 2%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 2%
Firebase2%18%11accepted challenger▲+2Since September 2026: 0% → 2%, +2 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 2%Show the six products at 0%, ordered by negative rate
15
Telegram0%91%11criticized challenger=heldSince September 2026: 0% → 0%, ±0 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 0%
Telegram0%91%11criticized challenger=heldSince September 2026: 0% → 0%, ±0 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 0%13
Sendbird0%19%16accepted challenger=heldSince September 2026: 0% → 0%, ±0 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 0%
Sendbird0%19%16accepted challenger=heldSince September 2026: 0% → 0%, ±0 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 0%14
Product Fruits0%10%10accepted challenger▼−5Since September 2026: 5% → 0%, −5 points. Inside the 12-point floor: within noise. Read over the models both editions asked.5% → 0%
Product Fruits0%10%10accepted challenger▼−5Since September 2026: 5% → 0%, −5 points. Inside the 12-point floor: within noise. Read over the models both editions asked.5% → 0%11
CleverTap0%7%14accepted challenger=heldSince September 2026: 0% → 0%, ±0 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 0%
CleverTap0%7%14accepted challenger=heldSince September 2026: 0% → 0%, ±0 points. Inside the 12-point floor: within noise. Read over the models both editions asked.0% → 0%12
Stream Chat0%7%14accepted challenger▼−5Since September 2026: 5% → 0%, −5 points. Inside the 12-point floor: within noise. Read over the models both editions asked.5% → 0%
Stream Chat0%7%14accepted challenger▼−5Since September 2026: 5% → 0%, −5 points. Inside the 12-point floor: within noise. Read over the models both editions asked.5% → 0%10
Customer.io0%6%17accepted challenger▼−5Since September 2026: 5% → 0%, −5 points. Inside the 12-point floor: within noise. Read over the models both editions asked.5% → 0%
Customer.io0%6%17accepted challenger▼−5Since September 2026: 5% → 0%, −5 points. Inside the 12-point floor: within noise. Read over the models both editions asked.5% → 0%The floor is 12 points of share, measured: how far the models move a leader on their own when the same questions are asked twice with nothing changed. A larger change is movement; a smaller one is noise, and both are shown. How the floor is measured
Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its product page.
All fifteen head-to-head pages: the top six products, each against each
Recommended versus criticized
Every product with at least 10 labels here, on both axes. The 30% line names a quadrant, not the verdict above: that one needs more than 40%.
Criticized challengerCriticized default
Negative label rate →
010203040506070809101112131415Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative40%
Key
01Intercom21%
Intercom21%02OneSignal17%
OneSignal17%03Userpilot16%
Userpilot16%04Appcues9%
Appcues9%05Pendo3%
Pendo3%06Knock3%
Knock3%07Braze2%
Braze2%08CometChat2%
CometChat2%09Firebase2%
Firebase2%10Customer.io0%
Customer.io0%11CleverTap0%
CleverTap0%12Stream Chat0%
Stream Chat0%13Sendbird0%
Sendbird0%14Product Fruits0%
Product Fruits0%15Telegram0%
Telegram0%What they warned about
A high negative share on a product with few labels is a warning. A low share on a product with many labels is salience, not sentiment.
Telegram91%
10 of 11 labels negative · 10 of 6 models · 4 hard negative
“For sensitive communication, avoid using: Telegram (possible connections to Russia, with allegations of collaboration with the FSB)” Claude Haiku 4.5, negative prompt
WhatsApp89%
8 of 9 labels negative · 8 of 6 models · 4 hard negative
“Platforms to **avoid or use extreme caution with**... collects metadata (contacts, device info) shared with Meta” Grok 4.20, negative prompt
Facebook Messenger100%
7 of 7 labels negative · 7 of 6 models · 5 hard negative
“Apps specifically called out as more concerning in the supplied sources include **Facebook Messenger**, which is described as collecting substantial data and not offering end-to-end encryption by default” Perplexity Sonar, negative prompt
Intercom17%
8 of 47 labels negative · 7 of 6 models · 3 hard negative
“What I'd Avoid for Mid-Market B2B: - **Intercom**: Pricing tied to MAUs/contacts scales unpredictably” Kimi K2, paraphrase prompt
What they cite
Citations exist only for the models that return a source list: fourteen of the six in this edition.
Sites the answers cite
79 of 84 answers in this category came back with a source list, from 14 of 6 models: citations where the model returns them, or the search results it consulted. 1034 links across 328 sites, every framing counted. Ranked by the number of answers carrying the site or page. None of the 36 answers across every segment cited this index's own page for the category.
vendor site · Guideflow43 answers · 48 citations · 11 models
vendor site · Userpilot39 answers · 47 citations · 13 models
vendor site · CometChat27 answers · 33 citations · 11 models
vendor site · Knock25 answers · 33 citations · 12 models
23 answers · 23 citations · 7 models
23 answers · 23 citations · 9 models
vendor site · Product Fruits20 answers · 20 citations · 9 models
vendor site · OneSignal14 answers · 21 citations · 10 models
vendor site · Respond.io13 answers · 15 citations · 7 models
vendor site · Intercom12 answers · 21 citations · 6 models
vendor site · Pushwoosh12 answers · 12 citations · 7 models
11 answers · 16 citations · 7 models
Pages the answers cite
The ten pages named in the most answers, by full address. A page here is one the models returned with a recommendation, not one the index endorses.
36 answers · 11 models
35 answers · 11 models
23 answers · 7 models
23 answers · 9 models
21 answers · 8 models
20 answers · 11 models
19 answers · 9 models
12 answers · 7 models
11 answers · 8 models
11 answers · 8 models
Search against answers
Each company's standing in the answers beside its site's footprint in Google search, one row a site: the products the models named on it with their shares, and the share they add up to; monthly searches on Google, and DataForSEO's estimate of AI search demand (modeled from search signals, directional, not a count of queries to any assistant), for the most-searched of the company's and its products' names (the name is in each row's hover text); estimated monthly organic visits to the site; and its best position in Google's top ten for “best in-app messaging platform”, “in-app messaging platform”. US estimates from DataForSEO and Google's Ads Transparency Center. A small company's site, or a mid-sized company's site for its flagship, is marked company; a product on a large parent's site (Google, Microsoft) has no site figures. A column with no figures for this category is left out, and an empty cell means none were seen, not none exist. Two measurements side by side: neither is read as the cause of the other.
| Company and products | Share | Own site cited | Name searches, Google | AI search demand, est. | Organic visits | Paid search a month, est. | Google ads, last 30 days | Best Google position |
|---|---|---|---|---|---|---|---|---|
| Intercom intercom.com | 21% | 441 | 33,100 | 7,165 | 64,348 | $3,304 | 120 | |
| OneSignal onesignal.com | 17% | 330 | 4,400 | 133 | 29,726 | 38 | 6 | |
| Userpilot userpilot.com | 16% | 251 | 1,600 | 10 | 62,084 | 6 | 1 | |
| Appcues appcues.com | 9% | 226 | 1,900 | 26 | 18,506 | 2 | ||
| Crisp crisp.chat | 5% | 72 | 60,500 | 36,276 | 29,685 | 2 | ||
| Firebase In-App Messaging firebase.google.com | 5% | 13 | 91,247 | 1 | ||||
| Pendo pendo.io | 3% | 420 | 12,100 | 761 | 19,185 | $3,609 | 93 | |
| Knock knock.app | 3% | 54 | 33,100 | 94,585 | 7,075 | 0 | ||
| Braze braze.com | 2% | 505 | 27,100 | 6,395 | 55,552 | $13,112 | 59 | |
| CometChat cometchat.com | 2% | 41 | 720 | 12 | 9,329 | 7 |
Google's top ten for the two searches hold 15 sites; 3 of them are among the sites the models cited here (onesignal.com, pushwoosh.com, userpilot.com). The first result for “best in-app messaging platform” is userpilot.com.
How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.ShowHide
How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.
ShowHideModelDirect“What is the best in-app messaging platform for a mid-market B2B company?”Paraphrase“Which customer communications tool for product messaging would you recommend to a mid-sized B2B company?”Comparative“What are the top in-app messaging platforms and how do they differ?”Budget-constrained“What is the best in-app messaging platform for a company with a limited budget?”Scale-constrained“We are a 500 person company evaluating an in-app messaging platform. What should we look at?”Negative“Which in-app messaging platforms should I avoid or be cautious about?”
Claude Opus 5no first choiceno first choiceno first choiceno first choiceno first choicenothing named
Claude Opus 4.8no first choiceno first choiceno first choiceno first choiceno first choicenothing named
GPT-6 Astrano first choiceno first choiceno first choiceno first choiceno first choicenothing named
GPT-5.6 Solno first choiceno first choiceno first choiceno first choiceno first choicenothing named
Gemini 3.1 Prono first choiceno first choiceno first choiceno first choiceno first choicenothing named
Perplexity Sonar Prono first choiceno first choiceno first choiceno first choiceno first choicenothing named
Bold is the first choiceAlternatives are counted; the count opens them.What the answer argued against
The record
One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency. Full answer text is in the free responses file. Download the record
Zero rows: every prompt, every model, every answer.
| Prompt | Model | Version string | Time (UTC) | Searched | Sources | Latency |
|---|
Normalization in this category
Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.
ShowHide
Category-scoped readings
GetStream read as Stream Chat
GetStream (Stream Chat) read as Stream Chat
Oracle read as Oracle Service Cloud
Stream read as Stream Chat
Stream (GetStream) read as Stream Chat
Stream (GetStream.io) read as Stream Chat
Twilio read as Twilio Conversations
Unresolved, counted raw
Apple iMessage
Chat-SDK
Chat21
Crescendo
Ejabberd (XMPP)
EngageLab EngageSDK
Google Hangouts
LinkedIn
Meta Messenger / Facebook Messenger
Pusher Channels
Rakuten Viber Messenger
SMS
SMS / Standard SMS-Based Messaging
SMS/MMS (Standard Texting)
Salesforce
Standard SMS/MMS/RCS
TapTalk.io
Twilio Inc.
Twilio SDK
Viber
Discontinued, still offered
No shut-down product was recommended here.