Product analytics
Asked as “product analytics platform”, and as “product analytics tool”, on behalf of a mid-market B2B software company. 29 first choices recorded across the direct, paraphrase, budget and scale prompts, six models each.
Standing · first-choice share
38%
Contested · Mixpanel 34%
38Amplitude34Mixpanel24PostHog03others
38% of first choices, contested.
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
Show the two products at 0%, ordered by negative rate
06
Google Analytics 4 Lab in the set0%56%16criticized challengernewNew since September 2026: not ranked then, 0% now.
Google Analytics 4 Lab in the set0%56%16criticized challengernewNew since September 2026: not ranked then, 0% now.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 can't be told apart from repeat variation, and both are shown. Each index measures its own floor from its own repeat, over its own categories, so the floor differs from one index to the next. 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.
Google Analytics 4 is made by Google, whose model Gemini 3.1 Pro is in the set. On the four share prompts that model made it the first choice zero times of 4; the other five models zero times of 20. Lab treatment is defined on the method page; the row is marked, not excluded.
All six head-to-head pages: the top four 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 →
01
02
03
04
0506Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative50%
Key
01Amplitude38%
Amplitude38%02Mixpanel34%
Mixpanel34%03PostHog24%
PostHog24%04Pendo3%
Pendo3%05Heap0%
Heap0%06Google Analytics 40%
Google Analytics 40%What they warned about
Two of six models held their first choice under the paraphrase. Claude Opus 4.8, GPT-6 Astra, Gemini 3.1 Pro and Perplexity Sonar Pro changed. 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.
Google Analytics 456%
9 of 16 labels negative · 6 of 6 models · 4 hard negative
“I'd avoid this for actual product analytics (as opposed to marketing attribution). Data sampling and cardinality thresholding make funnel and retention numbers unreliable” Claude Opus 5, negative prompt
Heap27%
6 of 22 labels negative · 5 of 6 models · 1 hard negative
“You should avoid seeking out or signing new contracts ... Heap Analytics: ... It no longer exists as a standalone company” Gemini 3.1 Pro, negative prompt
Amplitude23%
8 of 35 labels negative · 5 of 6 models
“Amplitude only if depth of behavioral analysis is genuinely the bottleneck — otherwise its usage-based pricing tends to outgrow a limited budget fastest” Claude Opus 5, budget prompt
PostHog12%
4 of 34 labels negative · 4 of 6 models · 1 hard negative
“I would avoid treating that deployment as a supported enterprise offering” GPT-6 Astra, negative prompt
What they cite
Six of the six models returned a source list in this edition: six return the pages they cited.
Sites the answers cite
32 of 36 answers in this category came back with a source list, from 6 of 6 models: citations where the model returns them, or the search results it consulted. 579 links across 187 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 · Amplitude27 answers · 46 citations · 5 models
vendor site · PostHog21 answers · 30 citations · 6 models
vendor site · Pendo21 answers · 21 citations · 5 models
vendor site · Mixpanel16 answers · 23 citations · 4 models
vendor site · Userpilot13 answers · 18 citations · 4 models
12 answers · 22 citations · 4 models
12 answers · 13 citations · 3 models
vendor site · Gainsight12 answers · 12 citations · 3 models
11 answers · 12 citations · 3 models
9 answers · 9 citations · 3 models
vendor site · Heap8 answers · 10 citations · 5 models
vendor site · Fullstory8 answers · 8 citations · 3 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.
15 answers · 3 models
12 answers · 3 models
12 answers · 3 models
12 answers · 3 models
10 answers · 3 models
10 answers · 3 models
9 answers · 3 models
9 answers · 2 models
7 answers · 2 models
6 answers · 2 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 product analytics platform”, “product analytics 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 |
|---|---|---|---|---|---|---|---|---|
| Amplitude amplitude.com | 38% | 418 | 74,000 | 5,691 | 373,273 | $207 | 27 | 2 |
| Mixpanel mixpanel.com | 34% | 467 | 12,100 | 458 | 20,284 | $458 | 11 | |
| PostHog posthog.com | 24% | 420 | 33,100 | 271 | 25,936 | 120 | ||
| Pendo pendo.io | 3% | 409 | 12,100 | 761 | 19,185 | $3,609 | 93 | |
| Contentsquare heap.io Heap 0% | 0% | 150 | 40,500 | 11,454 | 16,541 | 32 | 7 | |
| Matomo matomo.org | 0% | 160 | 1,900 | 87 | 35,015 | 9 | ||
| Kubit kubit.ai | 0% | 4 | ||||||
| Mitzu mitzu.io | 0% | 3 | ||||||
| Gainsight PX gainsight.com | 0% | 27 | 260 | 4 | 59,569 | 58 | 9 | |
| Contentsquare contentsquare.com | 0% | 184 | 2,900 | 101 | 149,925 | 86 |
Google's top ten for the two searches hold 16 sites; 7 of them are among the sites the models cited here (amplitude.com, fullstory.com, gainsight.com, heap.io, learn.g2.com, userpilot.com, visionlabs.com). The first result for “best product analytics platform” is domo.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 product analytics platform for a mid-market B2B software company?”Paraphrase“Which product analytics tool would you recommend to a mid-sized B2B software company?”Comparative“What are the top product analytics platforms and how do they differ?”Budget-constrained“What is the best product analytics platform for a company with a limited budget?”Scale-constrained“We are a 500 person company evaluating a product analytics platform. What should we look at?”Negative“Which product analytics platforms should I avoid or be cautious about?”
Claude Opus 5MixpanelMixpanelHeldagainst: Amplitudeno first choicePostHogagainst: AmplitudeAmplitude, Mixpanel, PostHogagainst: Amplitude, Google Analytics 4, Heap, Indicative, Interana, Kissmetrics, Localytics, Mixpanel, Pendo, PostHog, Woopra
Three alternatives
Amplitude, Pendo, PostHogFour alternatives
Gainsight PX, Heap, Pendo, PostHogFour alternatives
Amplitude, Heap, Mixpanel, PostHogOne alternative
MixpanelThree alternatives
Kubit, Mitzu, PendoClaude Opus 4.8Amplitude, MixpanelAmplitudeChangedAmplitude, Mixpanelagainst: Google Analytics 4PostHogno first choiceagainst: Google Analytics 4
Three alternatives
Heap, Pendo, PostHogTwo alternatives
Mixpanel, PostHogTwo alternatives
Heap, PostHogThree alternatives
Amplitude, Google Analytics 4, MixpanelGPT-6 AstraAmplitudeMixpanelChangedAmplitude, MixpanelPostHogAmplitude, Mixpanelagainst: Amplitude, Google Analytics 4, Heap, Mixpanel, Pendo, PostHog
Three alternatives
Mixpanel, Pendo, PostHogThree alternatives
Amplitude, Pendo, PostHogFour alternatives
Fullstory, Heap, Pendo, PostHogOne alternative
AmplitudeThree alternatives
Heap, Pendo, PostHogGPT-5.6 SolAmplitudeAmplitudeHeldAmplitudeagainst: Google Analytics 4, SnowplowPostHogAmplitude, Mixpanelagainst: Adobe Analytics, Amplitude, Fullstory, Google Analytics 4, Heap, Hotjar, Mixpanel, Pendo, PostHog
Three alternatives
Mixpanel, Pendo, PostHogThree alternatives
Mixpanel, Pendo, PostHogFive alternatives
Fullstory, Heap, Mixpanel, Pendo, PostHogThree alternatives
Amplitude, Google Analytics 4, MixpanelThree alternatives
Contentsquare, Pendo, PostHogGemini 3.1 ProAmplitude, MixpanelMixpanelChangedAmplitude, MixpanelPostHogAmplitudeagainst: Heapagainst: Adobe Analytics, Amplitude, Contentsquare, Google Analytics 4, Heap, LogRocket, Statsig, Visual Studio App Center
Three alternatives
Heap, Pendo, PostHogThree alternatives
Amplitude, Pendo, PostHogThree alternatives
Heap, Pendo, PostHogTwo alternatives
Amplitude, MixpanelTwo alternatives
Mixpanel, PostHogPerplexity Sonar ProAmplitudePendoChangedAmplitudeagainst: Google Analytics 4Mixpanel, PostHogagainst: Amplitudeno first choiceagainst: Amplitude, Fullstory, Google Analytics 4, Heap, Hotjar, Microsoft Clarity, Mixpanel, Pendo, PostHog
Two alternatives
Mixpanel, PendoThree alternatives
Amplitude, Mixpanel, PostHogFive alternatives
Fullstory, Heap, Mixpanel, Pendo, PostHogOne alternative
Google Analytics 4Bold 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.
Thirty-six rows: every prompt, every model, every answer.
| Prompt | Model | Version string | Time (UTC) | Searched | Sources | Latency |
|---|---|---|---|---|---|---|
| Direct recommendation | Claude Opus 5 | claude-opus-5 | 2026-09-08 10:15 | yes | 38 | 62 s |
| Direct recommendation | Claude Opus 4.8 | claude-opus-4-8 | 2026-09-08 10:16 | yes | 33 | 50 s |
| Direct recommendation | GPT-6 Astra | gpt-6-astra | 2026-09-08 10:18 | yes | 5 | 36 s |
| Direct recommendation | GPT-5.6 Sol | gpt-5.6-sol | 2026-09-08 10:18 | yes | 5 | 33 s |
| Direct recommendation | Gemini 3.1 Pro | gemini-3.1-pro-preview | 2026-09-08 10:17 | yes | 6 | 45 s |
| Direct recommendation | Perplexity Sonar Pro | sonar-pro | 2026-09-08 10:16 | yes | 20 | 6 s |
| Paraphrase | Claude Opus 5 | claude-opus-5 | 2026-09-08 10:19 | yes | 39 | 72 s |
| Paraphrase | Claude Opus 4.8 | claude-opus-4-8 | 2026-09-08 10:20 | no | 0 | 16 s |
| Paraphrase | GPT-6 Astra | gpt-6-astra | 2026-09-08 10:21 | yes | 6 | 31 s |
| Paraphrase | GPT-5.6 Sol | gpt-5.6-sol | 2026-09-08 10:21 | yes | 5 | 36 s |
| Paraphrase | Gemini 3.1 Pro | gemini-3.1-pro-preview | 2026-09-08 10:20 | no | 0 | 28 s |
| Paraphrase | Perplexity Sonar Pro | sonar-pro | 2026-09-08 10:20 | yes | 17 | 5 s |
| Comparative | Claude Opus 5 | claude-opus-5 | 2026-09-08 10:22 | yes | 28 | 60 s |
| Comparative | Claude Opus 4.8 | claude-opus-4-8 | 2026-09-08 10:23 | yes | 38 | 50 s |
| Comparative | GPT-6 Astra | gpt-6-astra | 2026-09-08 10:25 | yes | 9 | 54 s |
| Comparative | GPT-5.6 Sol | gpt-5.6-sol | 2026-09-08 10:26 | yes | 8 | 68 s |
| Comparative | Gemini 3.1 Pro | gemini-3.1-pro-preview | 2026-09-08 10:24 | yes | 2 | 40 s |
| Comparative | Perplexity Sonar Pro | sonar-pro | 2026-09-08 10:23 | yes | 20 | 10 s |
| Budget constrained | Claude Opus 5 | claude-opus-5 | 2026-09-08 10:27 | yes | 42 | 58 s |
| Budget constrained | Claude Opus 4.8 | claude-opus-4-8 | 2026-09-08 10:28 | yes | 35 | 51 s |
| Budget constrained | GPT-6 Astra | gpt-6-astra | 2026-09-08 10:29 | yes | 3 | 33 s |
| Budget constrained | GPT-5.6 Sol | gpt-5.6-sol | 2026-09-08 10:30 | yes | 5 | 26 s |
| Budget constrained | Gemini 3.1 Pro | gemini-3.1-pro-preview | 2026-09-08 10:29 | yes | 16 | 35 s |
| Budget constrained | Perplexity Sonar Pro | sonar-pro | 2026-09-08 10:28 | yes | 20 | 5 s |
| Scale constrained | Claude Opus 5 | claude-opus-5 | 2026-09-08 10:31 | yes | 31 | 62 s |
| Scale constrained | Claude Opus 4.8 | claude-opus-4-8 | 2026-09-08 10:31 | no | 0 | 29 s |
| Scale constrained | GPT-6 Astra | gpt-6-astra | 2026-09-08 10:33 | yes | 5 | 38 s |
| Scale constrained | GPT-5.6 Sol | gpt-5.6-sol | 2026-09-08 10:34 | yes | 6 | 61 s |
| Scale constrained | Gemini 3.1 Pro | gemini-3.1-pro-preview | 2026-09-08 10:32 | no | 0 | 31 s |
| Scale constrained | Perplexity Sonar Pro | sonar-pro | 2026-09-08 10:31 | yes | 20 | 7 s |
| Negative framing | Claude Opus 5 | claude-opus-5 | 2026-09-08 10:35 | yes | 37 | 84 s |
| Negative framing | Claude Opus 4.8 | claude-opus-4-8 | 2026-09-08 10:36 | yes | 33 | 50 s |
| Negative framing | GPT-6 Astra | gpt-6-astra | 2026-09-08 10:37 | yes | 6 | 37 s |
| Negative framing | GPT-5.6 Sol | gpt-5.6-sol | 2026-09-08 10:39 | yes | 8 | 73 s |
| Negative framing | Gemini 3.1 Pro | gemini-3.1-pro-preview | 2026-09-08 10:37 | yes | 13 | 41 s |
| Negative framing | Perplexity Sonar Pro | sonar-pro | 2026-09-08 10:36 | yes | 20 | 8 s |
Noise floor in this category
Flips between the edition run and its calibration repeat. Six prompts per model is a small sample; the index-wide floor is the number to trust.
ShowHide
Claude Opus 5
2 of 5 flipped
Claude Opus 4.8
4 of 4 flipped
GPT-6 Astra
0 of 5 flipped
GPT-5.6 Sol
1 of 5 flipped
Gemini 3.1 Pro
3 of 6 flipped
Perplexity Sonar Pro
4 of 4 flipped
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
Amplitude read as Amplitude
Unresolved, counted raw
None awaiting human review.
Discontinued, still offered
No shut-down product was recommended here.