Skip to content
Get Ranked On TopIllustrative case studyContent Engine plan

A fictional business. The figures are a modelled scenario, not a client result.

Sumpit Studio · Podcast network, 14 shows in Filipino and English · Quezon City

How a Quezon City podcast network turned 860 episodes into chapters that AI search cites

0 → 412 AI-answer citations a month

12 months · Asia-Pacific · Content Engine plan · Chaptered transcripts · Question-led chapters · PodcastEpisode & Clip schema · Bilingual topic hubs

Organic visits a month

1,600 → 19,800

Questions in Google's top 10

2,340

On air → on the page14 shows · 860 episodes · 4,120 chapters

The challenge

Forty-one thousand minutes of talk. Zero words to search.

Fourteen Filipino and English shows lived as embedded players with no transcripts, so search couldn't read them and AI assistants never cited them.

Sumpit's hosts go deep: an hour on MP2 savings, forty minutes on why adobo improves overnight, a whole episode on night-differential pay. Listeners loved them. Search engines and AI assistants never noticed, because every episode page was a third-party player and a three-line blurb. When people asked those exact questions, the answers came from forum threads and news explainers that covered the topic in a paragraph.

An episode page, as a crawler read it

Indexable words
41
Audio
00:52:40
Timestamps
0
Questions answered
0

41,280

minutes of conversation in the catalogue, and not one line of it in text.

Diagnosis

What the audit heard on the tape.

Episode pages

860

Every one of the 860 episodes was a third-party player embed with an average of 41 words of show notes. Crawlers saw a title, a date and an iframe.

Finding 1 of 4

Audio

41,280 min

41,280 minutes of interviews and explainers in Filipino, Taglish and English, none of it in text that a search engine or an AI model could read or quote.

Finding 2 of 4

AI answers

0 / 120

We put 120 questions the shows answer in depth to four AI assistants. Sumpit was cited zero times; forums, news explainers and bank blogs were cited instead.

Finding 3 of 4

Language

61%

61% of episodes are spoken in Filipino or Taglish, yet titles and notes were written in English only, so Filipino searches had nothing to match.

Finding 4 of 4

The catalogue · episodes per showFilipino & TaglishEnglish

521 of 860 episodes are in Filipino or Taglish.

The strategy

Cut the tape into chapters. Let every chapter answer a question.

Built on our Content Engine plan: content and AI-search work for businesses whose expertise is real but not yet readable by machines.

  1. 01

    Chaptered transcripts

    Speech-to-text tuned for Taglish, then a human editor per show. Each episode is cut into four to six chapters with in and out timecodes, so the page carries the whole conversation as clean, quotable text.

  2. 02

    Question-led chapters

    Chapter titles are the questions listeners actually type, in the language they type them: Filipino for Pera Usapan, English for Night Shift Notes. Each chapter opens with a two-sentence answer before the transcript.

  3. 03

    PodcastEpisode & Clip schema

    Every episode page declares PodcastEpisode with inLanguage, and each chapter a Clip with startOffset and endOffset, so search and AI assistants can link straight to the minute that answers.

  4. 04

    Bilingual topic hubs

    Thirty-two hubs (personal finance, home cooking, shift work, local history) gather chapters from different shows, with Filipino and English versions linked by hreflang.

On the edit timeline

Scrub the waveform. Watch a stretch of talk become a chapter search can cite.

Pick an episode and drag the playhead. The stretch under it snaps to its chapter: timecodes, transcript, schema and, if AI search cites it, the questions it answers.

Drag across the waveform, or focus it and use ← → to move 15 seconds, Page Up / Page Down to jump a chapter, Home / End.

Chapters

Chapter 1 selected: How big should an emergency fund be on a ₱20,000 salary?

Chapter 1 of 4 · Pera Usapan #112

How big should an emergency fund be on a ₱20,000 salary?

In
00:00:00
Out
00:10:10

Transcript

[00:00]Start with one month of fixed costs, not six. On twenty thousand that's about eleven thousand pesos, built five hundred at a time, every payday.

Cited 14 times in AI answers last month

Questions this chapter answers

  • How much emergency fund do I need on a minimum wage?
  • Where should I keep an emergency fund in the Philippines?

Structured data (JSON-LD)

{
  "@type": "Clip",
  "name": "How big should an emergency fund be on a ₱20,000 salary?",
  "startOffset": 0,
  "endOffset": 610,
  "inLanguage": "fil",
  "url": "https://sumpitstudio.example/pera-usapan/112#t=0"
}

These three episodes earned 87 of the network's 412 AI-answer citations in month 12.

Month by month

12 months on one edit timeline.

  1. Months 1–2

    Audit & question map

    Catalogue and crawl audit, the 120-question AI test, and a map of 1,900 questions the shows already answer.

  2. Months 2–4

    Transcript pipeline

    Taglish speech-to-text tuned on 40 hours of the network's own audio, an editors' style guide, the chapter template and schema.

  3. Months 4–11

    Backlist chaptering

    All 860 episodes transcribed and chaptered, the most-searched shows first; 4,120 chapters live by month 12.

  4. Months 6–10

    Topic hubs

    32 bilingual hubs built from the question map, linking chapters across shows and both languages.

  5. Months 9–12

    New episodes & tracking

    Chapters published on release day; AI citations and Google positions tracked per chapter every week.

Results

Month twelve: the shows get quoted by the minute.

AI-answer citations a month

0 → 412

Tracked weekly across four AI assistants

Organic visits a month

1,600 → 19,800

+1,138% in twelve months

Timestamped chapters published

4,120

From 860 episodes, about 4.8 per episode

Questions in Google's top 10

2,340

Matched to chapter titles, Filipino and English

Modelled scenario for a fictional business — not a client result. Rankings, traffic and AI citations are never guaranteed; see our Terms.

AI-answer citations a month
0200400M1M4M7M10M12412
Organic visits a month
010,00020,000M1M4M7M10M1219,8001,600

Questions, as people asked them

Question as typedChapter that answersGoogle, month 1Google, month 12AI-cited
magkano dapat ang emergency fundPera Usapan #112 · 00:00:00not in top 100#3Yes
pag-ibig mp2 o time depositPera Usapan #112 · 00:10:10not in top 100#5Yes
bakit mas masarap ang adobo kinabukasanLutong Bahay Lab #87 · 00:00:00#64#2Yes
anong suka ang pang-paksiwLutong Bahay Lab #87 · 00:09:00not in top 100#4Yes
how to sleep after night shiftNight Shift Notes #203 · 00:00:00#88#7Yes
night differential pay philippinesNight Shift Notes #203 · 00:36:00not in top 100#4Yes
e-wallet cash out feesPera Usapan #112 · 00:23:05not in top 100#9Not yet

Questions are shown exactly as typed, in Filipino or English.

Get a plan like this

Sitting on hours of expert audio or video that search can't read? Start with the Content Engine plan, or tell us about your catalogue.

← All case studies