Measure Public Search Share of Voice

Share of voice needs a clear denominator. Search results can give you a repeatable directional measure, but they cannot count every conversation on the internet.

This workflow calculates each brand’s share of returned mentions under one fixed method.

SocialListeningAPI searches public posts on LinkedIn, X, Reddit, Facebook, Hacker News, TikTok, and YouTube and Google results for the phrases you choose. It returns normalized records with source URLs.

Step 1: Freeze the comparison method

Create a SocialListeningAPI account and claim the 100 free credits. Write down:

• Brands and products included
• Exact query list
• Sources included
• Collection date and time
• Deduplication rule
• Rule for deciding which brand a result mentions

Do not change these rules midway because one brand has weak results.

Step 2: Collect share of voice results

for endpoint in \
  linkedin/search-posts \
  x/search-posts \
  reddit/search-posts \
  facebook/search-posts \
  hackernews/search-posts \
  tiktok/search-posts \
  youtube/search-posts \
  google/search
do
  curl --get "https://api.sociallisteningapi.com/api/v1/$endpoint" \
    --data-urlencode 'query=brand name category' \
    --header 'x-api-key: YOUR_API_KEY'
done

One brand query across eight sources costs 15 credits. The Facebook search API returns one page with no next cursor, so treat its count as returned results, not total Facebook activity. Claude or ChatGPT can collect sources through the MCP.

Step 3: Deduplicate and label records

Store query, brand, platform, source URL, date, and inclusion status in an external database. Remove duplicate URLs and copied records. Have a person review ambiguous brand names.

SocialListeningAPI performs search. Storage, brand labeling, and counting happen outside it.

The measurement does not automate participation. A human must read Reddit community rules and approve any reply.

Step 4: Calculate the directional share

For each brand, divide its unique included URLs by all included URLs for the compared brands, then multiply by 100. Show the raw count beside every percentage.

Break results down by platform when useful. Do not combine available engagement fields into the same measure unless you document a separate method.

Step 5: Publish the limits

Name the metric share of returned search results. Include queries, source list, dates, raw counts, deduplication, and missing coverage. Repeat the same method later to compare direction.

Conclusion

Use this measure to compare like-for-like search runs, not to claim a brand owns a percentage of all online discussion.

If someone cannot reproduce the percentage from the raw counts and method, the report needs more detail.

Continue with Build a Voice-of-Customer Message Bank.

Written by ShashFounder at sociallisteningapi, focused on product and growth