Start free

How it works

How Bungad works

Every week we ask your buyers’ questions, record who was named and cited, and write the fixes. Today on one engine: Perplexity, the only one we hold an API key for.

Updated 20 October 2026 · 4 minute read

Dodong, the Bungad mascot, on a navy studio table among sketchbooks and fabric swatches while designers talk at desks behind him

The loop

Six steps, every week. The last feeds the first.

Same questions each time, so you see whether your change moved the answer. Weekly on paid plans; a Free account checks by hand.

Kalye / Weekly checkPerplexity
“best CRM for small agencies”Pipewise #1, Kalye #3 among tracked brandsNamed #3
Sources Perplexity citedg2.com/categories/crm/agency lists Pipewise and LedgerlineDiagnose
Suggested fixA plain page for agencies on kalye.io, with prices in the textOpen
Next checkSame question, so you see what movedProve
Example for a fictional brand, Kalye: one pass of the loop.
  1. 01

    Ask

    Your buyers’ questions, drafted from one sentence about what you sell.

  2. 02

    Run

    One call per question to the keyed engine. The full answer is stored.

  3. 03

    Measure

    Named? Where among the brands? Was your site cited?

  4. 04

    Diagnose

    Who was named instead, which domains were cited, can AI crawlers read your homepage.

  5. 05

    Fix

    Specific fixes, each with the answer that prompted it. You publish them.

  6. 06

    Prove

    The next check asks the same questions and shows what moved.

One question

Question in, score out.

Steps two and three of the loop, close up: what happens to each tracked question on the one engine we query today, Perplexity.

One question, start to finishFrom a buyer’s question to a scoreWhat a check does with every tracked question. The brands are made up; the steps are the product’s own.
  1. Ask

    What’s the best invoicing tool for small agencies?

    1 API call45 s timeout1 retry

    Your question, word for word, sent to the engine we hold a key for.

  2. Engine answers

    Perplexity · live

    Fernhub is the usual pick 1.

    Kalye is a strong alternative 2,

    and Tallybook suits freelancers 3.

    A failed call is stored as a failure and left out, never as “not named”.

    The full answer and every source it cites are stored.

  3. Names and citations found

    • Named: KalyeWhole word, links blanked first
    • 2nd of 3 brands namedPosition score 0.75
    • 1 of 3 sources yourskalye.io, counted apart from the name
    • Tone readShown, not scored

    Rivals are found the same way, so you see who was named instead.

  4. Scored

    75
    • Named, 50%50
    • Cited, 30%10
    • Rank, 20%15

    This answer alone scores 75. A run averages every answer it analysed.

Perplexity, and only Perplexity, todaySame questions, same arithmetic, every check.

Setting up

Setting up, in the order it happens

Nothing is fixed before it is measured. The arithmetic is in how we measure; the idea in what is AEO.

Dodong, the Bungad mascot, beside a navy clipboard holding a checklist, some boxes ticked in gold, and a gold pen
01

Name the brand and the domain

The name is matched as a whole word outside links; a link is a citation, counted separately. Ordinary-word names are harder to measure, and we say so.

02

Agree the question set

Twenty-five on Starter, one hundred on Growth, five hundred on Scale, per brand. Questions naming your brand are shown but kept out of the score. Which to track.

03

Run the baseline check

The zero point, not a report card. Ours: named in 0 of 8 answers, mention rate 0%, citation share 0.0%, score 0 out of 100. The whole run.

04

Fix what kept you out of the answer

No page answers it, the answer is buried or locked in an image, PDF or form, markup and page disagree, or trusted sources miss you. Why you are not in AI answers.

05

Repeat on the same questions, every week

Same questions, engine and arithmetic, published either way. Models update and rivals publish, so one measurement describes a day, not a position.

Dodong, the Bungad mascot, standing beside a glowing gold speech bubble on a navy background

A run

What a run is, mechanically

One API call per question per engine, read for four signals. A failed call is stored as a failure and excluded, never as a “not named”.

More detail

Asked twice a minute apart, our headline figures matched while per-domain citation counts moved.

Swipe sideways to see every column

SignalHow it is takenWhat it feeds
Named or notWhole-word match on the brand name, with links blanked out firstMention rate
PositionWhere your first mention falls among all the brands the answer namesThe position component
SentimentCue phrases around the mention, read as positive, neutral or negativeShown, not scored
CitationsEvery source domain the answer credited, and which of them are yoursCitation share, and the fix list
Anatomy of a scored answerWhat we read in one answerThe same made-up answer, taken apart. Every mark below is something the code records.

AnswerPerplexity

What’s the best invoicing tool for small agencies?

For small agencies, Fernhub is the usual pick for recurring retainers 1. Kalye is a strong alternative with simpler pricing 2, and Tallybook suits solo freelancers 3.

  1. 1fernhub.com
  2. 2kalye.ioYours
  3. 3reviewsite.example

This answer alone75(1 × 0.5 + 1/3 × 0.3 + 0.75 × 0.2) × 100

  1. The question, word for wordAsked the same way every week and stored with the result, so the number keeps its meaning.
  2. A rival, found the same wayCompetitors are matched as whole words too. That is how you see who was named instead.
  3. Named, and whereA whole-word match outside any link counts as named. Kalye is the 2nd brand named: 0.75.
  4. Tone: shown, not scoredThe sentence around the name is read as positive, neutral or negative, and kept out of the score.
  5. A link is a citationEvery source domain is recorded. kalye.io is the brand’s own, so 1 of these 3 sources is yours.

What you receive

What you receive, and where

In the app after every check: the score and its three inputs, each answer, who was named instead, every cited domain, and the fixes.

Nothing is emailed. No email provider is configured and no message has ever been sent to a customer. The monthly report is written and tested but dormant.

Dodong, the Bungad mascot, on a marble bakery counter among trays of fresh bread rolls while the baker kneads dough behind him
Try the arithmeticHow named, cited and rank add upEight answers analysed. Set where the brand was named in each, move the sources, and watch the score. The same arithmetic as the code.
Visibility score

26Occasionally surfaced

  • Named · mention rate3 ÷ 8 = 0.375 × 5018.75
  • Cited · citation share4 ÷ 60 = 0.0667 × 302
  • Rank · position component2 ÷ 8 = 0.25 × 205

Score = round(25.75) = 26, clamped to 0–100. Position scores: 1st 1.0, 2nd 0.75, 3rd 0.5, 4th 0.25, 5th or later 0, averaged over all eight answers.

Dodong, the Bungad mascot, on a white clinic reception desk while a smiling receptionist sits behind it

Who does what

Software, people, and your part

Code runs on a schedule; a person’s work differs by plan. Weekly check on every paid plan, plus 4, 20 or 100 manual checks. What AEO costs.

Swipe sideways to see every column

Software, on every planBy hand, and on which plan
Asks the tracked questions on the keyed engineStarter, $99 — structured data set up for you
Parses each answer for mentions, position, sentiment and citationsGrowth, $299 — Starter’s work, plus monthly content fixes written for you, and a monthly call
Scores the run and stores it wholeScale, $799 — Growth’s work, plus source building, and a named strategist you work with directly
Shows the score, sources and recommendations in the appNothing else, on any plan

Your part

  • Publish the fixes. We write them; you ship them.
  • Tell us when the product changes. Prices, markets, features.
  • Leave the questions alone between changes. Tune for the market, not a bad number.
  • Read the citation list, not just the score. That is where the work is.

What does not happen

  • No coverage beyond the keyed engine. ChatGPT, Claude and Gemini are built and unkeyed; AI Overviews and Copilot have no public consumer API, and we do not track them on any plan.
  • No guaranteed mention, no training-data edits. We change what the engine finds: correcting what AI gets wrong.
  • No market, country or language segmentation, and no email.

Cadence

Why weekly, and daily only for key questions

Our run-to-run variation means a daily line would mostly plot noise, so the score is weekly. On Growth and Scale, key questions you choose are also checked daily, listed on their own. How often to check.

Dodong, the Bungad mascot, on a marble reception counter in a bright clinic, with a receptionist smiling in the background

FAQ

Common questions

More, including the ones against us, in the FAQ.

How does Bungad actually work?

Each week we send your tracked questions to the engines we hold an API key for, one call per question per engine, and parse every answer for four things: whether your brand is named as a whole word outside any link, the order the named brands appear in, the tone of the sentence naming you, and every source domain the answer cited. Those signals become a score, a list of who was named instead, and a list of the domains the engine drew on. The diagnosis and the fixes are written against that list, and the whole thing repeats on the same questions the following month.

Which engines does Bungad check today?

Perplexity, and only Perplexity. It is the one engine we hold an API key for, so it is the only one a run can query. The ChatGPT, Claude and Gemini adapters are written but hold no key, and no check asks any of them. Google AI Overviews and Microsoft Copilot have no public consumer API, and we do not track them on any plan. Free and Starter use one engine. Growth and Scale use every engine we hold a key for. Today we hold one key, so every plan uses one engine, Perplexity.

Do you need access to my website?

No. The measurement reads nothing of yours: it reads what the engine says about you. For the fixes we write the passage and the markup and hand it over, and you publish it. If you would rather we published it, that needs whatever access your own stack requires, and it is arranged directly rather than through the app. Nothing in the product has ever held a credential for a customer’s website.

What happens if an engine call fails mid-run?

It is recorded as a failed call and left out of the arithmetic. It is never counted as an answer that failed to name you, because those two things look identical in a table and mean opposite things — one is a measurement, the other is a missing measurement. The run reports how many calls it completed so a thin month cannot pass as a bad month.

Why is the question set fixed rather than rewritten every month?

Because a score is only comparable against itself. Change the questions and the new number measures a different thing, so a rise cannot be told apart from an easier question set. You can change tracked questions whenever you like, and the change is worth making when your market moves — but the month it changes is a new starting point, not an improvement: the report compares only the questions that were asked both times.

Start with the baseline

3 questions on Perplexity. Were you named, was your site cited, and can AI crawlers read your homepage? Free, no account.

Or start free with 1 brand and 5 tracked questions. A domain checked in the last week shows its stored run; the daily free allowance resets at midnight UTC.

Dodong, the Bungad mascot, waving, a hand on his hip, against a gold background