Quantitative creator scoring tool
Influencer profiles are collected automatically, scored on quantitative metrics only, and sorted into grades. Selection follows defined criteria rather than personal judgment.
Creator selection
Before
Staff compared metrics by hand
After
Four quantitative scores combined into a grade
To protect client information, these screens recreate the same features with mock data.
Background
Creators are usually chosen by instinct. A large following looks promising, and active comments look even better. But what “looks” good differs from person to person.
The problem
Automating collection is only half the job. When criteria exist only in someone’s head, rejections are hard to explain and decisions cannot be repeated consistently. The selection criteria have to be set out on screen so that anyone using the tool reaches the same conclusion.
What we built
- Two platforms in one window — Instagram Reels and YouTube videos. Targets and status for both are managed in one place instead of running each platform separately
- Automated profile collection — Paste profile URLs or upload an Excel file, and the tool builds a list and works through it
- Scoring on quantitative metrics only — Average views, followers, average comments, and comments per view. Each is scored by band, and the total determines the grade. Qualitative factors such as brand image are excluded from scoring entirely. Mixing in factors that cannot be measured puts the criteria back in someone’s head
- Criteria set on screen — Staff adjust the score needed for an A grade or the view count that earns 3 points directly
- Paced collection — The tool waits a random 10 to 30 seconds between tasks. Steady pacing lets the candidate list be completed without interruption
- Clear progress tracking — Waiting, collecting, collected, and needs review are counted separately, and failed items remain visible
Collection uses only accounts the client is authorized to access.
Why a browser extension
The tool runs as an extension in the staff member’s own browser, not on a server. A server would require storing account credentials somewhere, and stored credentials have to be protected. An extension uses the browser that is already signed in. Account access stays with the staff member.
Results are exported to Excel.
Results
The tool was built and delivered in about 1.5 weeks. Enter a candidate list, and the metrics are filled in and each creator is graded against the set criteria.
Select a period, and the creators whose scores rose most during it appear at the top. Account names are hidden.
Related work is listed under Web data collection.
Related work is also listed under Data analysis.
Pricing by collection volume and complexity — Data collection pricing