PLUGIN SETUP
Start using The Trade Desk in ChatGPT
Connect The Trade Desk to Adzviser, install Adzviser in ChatGPT, and start asking questions about your data.
Connect The Trade Desk to Adzviser
Connect The Trade Desk using the API login and password provisioned for Adzviser.
Install Adzviser in ChatGPT
Open Adzviser in ChatGPT's plugin directory, install it, and sign in to your Adzviser account when asked.
Open Adzviser in ChatGPT ↗Ask your first The Trade Desk question
Start a new ChatGPT conversation and ask Adzviser to list your workspaces and show the advertiser connected to each. Choose a workspace, ask about performance or trends, and continue with follow-up questions.
CONVERSATION FLOW
Turn The Trade Desk data into clear next steps in ChatGPT
Start with the results, focus on the decision in front of you, and keep asking follow-up questions in the same conversation.
Review recent performance
Review recent campaign results and the metrics behind the headline. Ask for a compact table and make the reporting period clear.
See the “Weekly Campaign Performance” promptFocus on a decision
Optimize programmatic bids. Ask ChatGPT to compare the options, supporting evidence, and possible downsides.
See the “Bid Optimization” promptAsk a follow-up
Analyze cross-channel impact. Ask what may explain the result, what evidence is missing, and what to examine next.
See the “Cross-Channel Attribution Analysis” promptCHATGPT STARTERS
Start with a quick check, then analyze The Trade Desk
First, confirm ChatGPT can access the right data. Then ask a detailed question and follow up on the recommendation that matters most.
List my connected The Trade Desk advertiser accounts and show the Adzviser workspace for each. After I choose one, return "Click conversion revenue" for the last 7 complete days.
Optimize programmatic bids.
Review The Trade Desk bid performance for the workspace I choose during the last 30 complete days. For each ad group: Bid Strategy, Win Rate, eCPM, Conversions, and CPA. Recommend bid adjustments for optimal performance.
Keep the same data source and time period. Which recommendation should I prioritize, what results support it, and what should I review before acting on it?
