The Clariti MCP is most useful when you point it at a job you were going to do anyway. Here are three jobs creators in our early-access group pointed it at, with the prompts behind them and the results that got them most excited.
Build a working test group out of your own library
One creator needed mid-traffic recipes that already had video, to split into an A/B test. Condensed, the prompt was:
"Using Clariti, give me a list of 40 recipes that would be considered mid-range in traffic. Exclude the most and least popular posts, and only include posts that currently contain a video."
Clariti supplied traffic performance for every recipe plus which posts already contained video, and returned 40 qualifying posts, ready to split into two test groups of 20. No exports, no manual data matching.
Plan your next newsletter from two tools at once
Here the goal was working out which newsletter format actually performs, then lining up the posts to support the next send. Because the MCP runs inside Claude or ChatGPT, other connected sources are available in the same chat. Condensed:
"Compare my last 20 newsletters by subject and click rate and find the patterns. Then check Clariti for which pasta salad recipes I already have for the roundup I'm planning."
Newsletter opens and clicks came from the email platform, site performance from Clariti, in one conversation. Back came single-theme roundups (zucchini 5.23%, tomato 5.13% click rate) clearly beating older digest-style sends (1.1–1.7%), plus 9 existing pasta salad posts ranked by traffic, and the recommendation to publish "10 Easy Pasta Salad Recipes" rather than padding the list out to 15.
Decide what to publish next, and what to rule out
This one was about picking the strongest roundup ideas for the month ahead without duplicating what already exists. Condensed:
"Using Clariti for my site, recommend the two strongest August roundup ideas based on seasonality, existing recipes, search demand, content gaps, and internal linking. Exclude topics I already have roundups for unless a replacement would outperform. Rank the two and explain why."
Clariti supplied years of August performance, the existing recipe library, and every roundup already published. What came back was two ranked ideas, cucumber (9/10, with 13 existing recipes ready to include) and bell pepper (8.5/10, with 12), plus 7 duplicate topics ruled out and one deferred until a few more supporting recipes exist.
Three prompts you can reuse
Copy them as written or swap in your own site, post, timeframe, and criteria.
Build a working list: "Using Clariti, give me a list of 40 recipes that would be considered mid-range in traffic. Exclude the most and least popular posts, and only include posts that currently contain a video."
Plan your next send: "Using my newsletter and Clariti data together, compare my recent newsletters by click rate, find the patterns, and check which posts I already have to support the next send."
Choose what to publish: "Using Clariti data for my site, recommend the two strongest roundup posts to publish in August based on seasonality, my existing recipes and gaps against my current roundups. Rank them with your reasoning."