First-party research articles
First-party research articles. The blog of Shoryaku AI, a marketing service that has AI read ads, social, and analytics together. It records what we looked up and decided while building the service.
-
Why automatic competitor-site comparison was removed
The feature was removed after no comparison completed in production. The search API was unavailable, outputs were cut off, and the effect could not be measured.
-
Is a 14-day free trial the norm? Counted on 8 companies' official pages
Of 8 marketing-analytics SaaS sites, 4 stated a trial length. All 4 said 14 days with no card. The other 4 did not state a length.
-
Manage AI SaaS cost from both actual usage and the monthly cap
A single model price per call is not enough. This records how Gemini usage was converted into credits and how the AI cost ceiling was set.
-
All 11 AI SaaS listings on an M&A site reported zero monthly revenue
The prices were asking prices at listing time, not closed deals, and not a statistic for the whole AI SaaS market.
-
On 8 M&A listings, page views alone did not explain monthly revenue
Among listings that disclosed both, one had 1.6 million page views and ¥10,000 in monthly revenue, and another had 475 page views and ¥1.3 million. This is not a correlation study.
-
Asking prices are not enough to know what an indie SaaS sells for
An earlier pass called about 160 listings actual sales. This correction separates asking prices, ended listings, and official closed examples.
-
Pricing for 8 marketing-analytics SaaS products could not be compared by monthly fee
Public prices for products such as AgencyAnalytics, Databox, and Supermetrics bill by customers, data sources, or locations, not by one monthly number.
-
A 7-year SaaS listed for sale because of a server migration
The seller reported 47 paying companies and growth, and listed the business over operating capacity and a server move. One listing is not a general rule.