Research on software costs,
measurement and owned infrastructure.

Key findings

• [CALCULATED] 69 measurement-relevant platform changes from ten platforms took effect between 1 January 2025 and 25 September 2026, 32 of them in 2025 and 37 in 2026 up to the cutoff.
• [CALCULATED] The median notice period is 6 days across all 65 dated changes and 63 days across the 34 changes with advance notice.

Key findings

• [OBSERVED] The GA4 Measurement Protocol is intended to supplement automatic collection rather than replace it; a fully server-side implementation can retain only partial reporting. [8]
• [OBSERVED] A Measurement Protocol request can return HTTP 2xx even when its payload is malformed, incorrect, or not processed. [9][10]

Key findings

• OBSERVED The 2025 METR RCT reported 19 percent longer completion time in its specific setting when AI was permitted, even though participants subjectively perceived a speedup.
• OBSERVED METR's 2026 update provided weak indications of a possible speedup for newer tools, but names selection and participation problems that prevent a strong interpretation.

Key findings

• OBSERVED Several infrastructure-relevant projects changed their license or added new license options in recent years; this gave rise to forks, new foundation models and altered procurement decisions.
• CALCULATED In the disclosed model, CockroachDB 24.3+ reaches 72/100, MongoDB Community Server 63/100, Terraform/HashiCorp BSL 45/100, the MariaDB BSL reference model 42/100, Redis 38/100 and Elasticsearch 26/100. Higher means more modeled procurement risk, not worse software.

Key findings

• OBSERVED The EU Data Act has applied since 12 September 2025; its cloud switching rules address technical, contractual, commercial and organizational obstacles.
• OBSERVED From 12 January 2027, switching charges including the corresponding data egress fees are to be removed entirely. This rule must not be equated with generally free ongoing internet egress.

Key findings

• CALCULATED In the baseline model, 10,000 identical lead events generate 90,000 Zapier Tasks, 100,000 Make Credits, 10,000 n8n Executions or one active Activepieces flow - before retry and failure paths.
• CALCULATED Across the twelve extended archetypes, the assumed retry multiplier varies from 1.05 to 1.35; billable technical work therefore rises without additional business value.

Key findings

• CALCULATED Across the five base scenarios, the nominal break-even lies between 21.3 and 33.3 months.
• CALCULATED Two of the five scenarios show nominally positive savings after 36 months but a negative present value at eight percent. Nominal savings and economic advantage are therefore not identical.

Key findings

• CALCULATED The lock-in risk score of the ten-platform sample ranges from 17 to 63 points; the median is 55.
• CALCULATED The export completeness module reaches a mean of 72.3 and a median of 73.8 points.

Key findings

• CALCULATED The median of the pricing transparency index is 80 out of 100 points; 13 of 20 products reach at least 80 points.
• CALCULATED In the ten-vendor sample, the renewal/Exit module reaches a mean of 79.0 and a median of 80.0 points.