B2B Review Battlecard - Software Reviews To Sales Ammunition

Turn B2B software reviews into a sales battlecard: why buyers switch, named alternatives, objections, and ready-to-use talk tracks. Pay per report.

What this Actor does

It collects public source data (see Evidence sources), then uses a language model to turn that raw material into one structured report. One run produces one report, and you pay only when a report is delivered.

Input

Field	Type	Required	Description
productUrl	string	yes	G2 review page
maxRecords	integer	no	Reviews to analyze
outputLanguage	string	no	Report language
sampleData	boolean	no	Sample data (free, no data fetched)

Advanced

Field	Type	Required	Description
llmApiKey	string	no	Optional. Leave empty to use the service's built-in analysis model. Supply your own OpenAI-compatible key to bill model usage to your account. Stored encrypted.
llmApiBase	string	no	OpenAI-compatible base URL. Empty means the service default.
llmModel	string	no	Model identifier. Empty means the service default.

Output

One dataset item per run. Report fields:

Field	Type	Description
alternativesNamed	array	Alternatives named
winThemes	array	Win themes
lossThemes	array	Loss themes
buyerProfiles	array	Buyer profiles
objectionsToArm	array	Objections to arm
battlecardLines	array	Battlecard lines
customerQuotes	array	Customer quotes

Every report also carries:

Field	Type	Description
scope	object	Inputs and filters this report was produced for.
sourceRecordCount	integer	Records collected from the source.
recordsAnalyzed	integer	Records actually included in the analysis.
recordsDroppedPastBudget	integer	Records excluded once the analysis budget was reached.
chunksAnalyzed	integer	Number of analysis passes merged into the report.
sources	array	Upstream dataset identifiers used as evidence.
limitations	array	Known limits of this analysis.
demoMode	boolean	true when no model key was configured.
generatedAt	string	ISO-8601 timestamp.

Evidence source

Upstream Actor: zen-studio~g2-reviews-scraper (G2 software reviews). Data is collected at run time from publicly visible pages.

Pricing

Pay per event: $2.50 per delivered report (battlecard-report). For comparison, Crayon and Klue-style competitive-intelligence platforms cost $1,000+/mo.

Example input

{
  "productUrl": "https://www.g2.com/products/notion/reviews",
  "sampleData": true
}

Demo mode

If no model key is configured, the Actor still runs end to end and returns an extractive evidence report (record counts, frequency-ranked terms, verbatim samples) instead of model synthesis, with demoMode: true. This lets you evaluate a run before connecting a model.

Limitations

• Analysis is derived solely from publicly visible data collected at run time; it reflects the sampled window, not the full population.
• Results reflect the material collected in this run, not the entire population of available data.
• Analysis quality depends on how much public text the source exposes.

Independent third-party analysis service. Not affiliated with, endorsed by, or sponsored by the platforms whose public data is analyzed.