What problem are you solving?
A small AI customer-service vendor sends a rough lead sample with duplicate companies, missing websites, inconsistent country names, and mixed demand signals. They want to see whether you can clean a small sample before handling a larger list.
Use these assumptions
Assume the final output will go into Google Sheets, Feishu Sheets, or Excel. Do not access a real CRM or invent missing real-world data. Work only from the supplied sample.
Base your work only on this material
company,country,website,source_note,signal BrightPets,US,brightpets.com,Shopify store says support replies are slow,needs faster customer replies Bright Pets,USA,,same company as BrightPets?,duplicate maybe OceanLamp,UK,oceanlamp.co,has live chat but no FAQ bot,customer support automation fit NovaCase,United States,novacase.io,job post mentions support backlog,possible AI support need GreenTrail,DE,,only Instagram found,missing website Ocean Lamp,United Kingdom,oceanlamp.co,duplicate record from directory,duplicate MiniDesk,CA,minidesk.ca,small team uses Gmail for support,maybe low budget HuaMao Shop,CN,huamaoshop.cn,asks about WhatsApp auto reply in forum,strong fit
What to submit
- Cleaned lead table
- Dedupe explanation
- High, medium, or low priority for each lead
- Rationale for high-priority leads
- Missing information list
- Short client-facing delivery note
What makes a result pass review
- Use only the supplied data
- Identify BrightPets/Bright Pets and OceanLamp/Ocean Lamp as duplicate pairs
- Do not invent websites, emails, or contacts
- Priority labels must be evidence-based
- Mark missing fields and whether client confirmation is needed
Recommended format
- Submit the cleaned result as Markdown table or CSV
- After the table, add dedupe notes, priority rules, and the client delivery note
- You may use AI or spreadsheets, but the final result must be checked by you