Connect Adjust to ChatGPT

Windsor MCP gives ChatGPT access to a comprehensive set of Adjust metrics, including:

• Install and attribution: installs, reattributions, reinstalls, assisted installs, eCPI, eCPI, click conversion rate, impression conversion rate.
• Engagement and activity: sessions, base sessions, average DAUs, WAUs, MAUs, uninstalls, first uninstalls, reinstalls.
• Revenue metrics: ad revenue, IAP revenue, all revenue, ARPDAU, revenue events, cohort revenue, and cohort all revenue.
• Ad spend and ROI: ad spend, click cost, impression cost, install cost, ROAS, gross profit, ROI, revenue-to-cost ratio.
• Fraud detection: rejected installs (total, click injection, anonymous IP, distribution outlier), rejected install rate, click injection rate, anonymous IP rate.
• SKAdNetwork: SKAN installs, reinstalls, total installs, SKAN ROAS (min, max, average), SKAN ROI, conversion values 0-63, valid conversions, invalid payloads.

The data is live. Each time you run a query on ChatGPT, Windsor.ai fetches the latest available data directly from the Adjust API, so your analysis is always up-to-date.

Yes. ChatGPT can compare organic installs and non-organic installs across sessions, revenue, and ARPDAU, giving you a clear picture of how much growth is driven by paid acquisition versus users finding your app on their own.

Yes. You can ask ChatGPT to calculate your total rejected install rate, break down fraud by type and network, and identify which campaigns or partners are delivering the highest share of fraudulent traffic.

Yes. Prompt ChatGPT to break down cumulative revenue by cohort, network, or campaign, and see which acquisition sources are actually worth their cost over time.

For deeper insights, try more advanced prompts such as:

• Compare cohort revenue for users acquired in the last 3 months, broken down by network and campaign. Which cohorts are generating the highest cumulative revenue relative to their acquisition cost, and which ones looked efficient on eCPI but underdelivered on long-term revenue?

• Analyze my fraud data for the last 90 days across all networks. Rank networks by rejected install rate and break down each network's rejection volume by fraud type (click injection, anonymous IP, distribution outlier, and too many engagements). Flag any network where the rejected install rate exceeds 10% of total installs.

• For each campaign, compare Adjust-attributed installs against network-reported installs for the last 60 days. Calculate the install discrepancy rate for each network and rank them by the size of the gap.

• Analyze my SKAdNetwork data for the last 90 days. Show SKAN installs, valid conversions, invalid payloads, and SKAN ROAS across minimum, maximum, and average ranges by campaign. Which campaigns show the widest gap between SKAN minimum and maximum ROAS, and what does that suggest about conversion value distribution?

Windsor.ai’s pricing for ChatGPT data integration varies depending on the number of connected accounts and data sources. We offer absolutely transparent pricing plans for diversified business needs and budgets, along with a 30-day free trial.

With Windsor.ai’s intuitive connector, the integration process takes just a few minutes. No technical expertise is required; connecting Adjust to ChatGPT is simple, straightforward, and completely code-free.

Yes, we have helpful resources to help you get started. You can explore our official documentation and tutorials for seamless Adjust to ChatGPT integration with Windsor.ai:

• • Adjust Field Reference
  • ChatGPT Integration Documentation