Bungad AEO: frequently asked questions

A monthly subscription that measures how AI assistants answer the questions your buyers ask, diagnoses why your brand is not in those answers, and fixes it. Each month we ask the engines we hold a key for your tracked questions, record which brands were named and which pages were cited, and publish a score with the evidence behind it. It is Answer Engine Optimization sold as a measured, repeating service rather than a one-off project.

Bungad is Tagalog for the front, or the first thing you see. It is the goal stated as a name: when someone asks an assistant about your market, your brand should be the first thing they see.

Both, and the split is worth knowing before you buy. The software asks the questions, parses the answers, scores the run and shows you everything in the app. A person sets up your structured data, writes the monthly content fixes on Growth and above, does the source-building work on Scale, and reviews by hand the two answer surfaces that have no API. A provider selling only the first half is selling a tracker, and a tracker is a cheaper purchase than this one.

Perplexity, and only Perplexity. It is the single engine we hold an API key for, so it is the only one where we can promise a reproducible, auditable run. ChatGPT and Claude adapters are written but hold no key, and no run has ever queried either. Gemini is not built at all. Google AI Overviews and Microsoft Copilot expose no public API, so they are reviewed by hand for Scale clients rather than automated.

No, and any wording suggesting otherwise would be false today. Every plan gets whatever we have keyed at the time, which today means one engine on all three. What the higher plans buy is more brands, a larger question set, more manual checks and more work done by a person — not more engines.

When keys exist for them. The adapters are written and tested; what is missing is an account and a key for each, which is a purchase rather than a build. We are not going to publish a date for something whose only remaining step is outside the code, and we will not list either engine as covered until a run has actually queried it.

Three plans, in US dollars, month to month. Starter is $99: one brand, twenty-five tracked questions, four manual checks a month on top of the scheduled one, and structured data set up for you. Growth is $299: three brands, one hundred questions, twenty manual checks, competitor citation share, monthly content fixes written for you and a monthly call. Scale is $799: twenty brands, five hundred questions, one hundred manual checks, source building, and a named strategist you work with directly. No setup fee and no annual contract.

There is a free check, which asks three questions on Perplexity and needs no payment and no account. A domain checked in the last week is served from that stored run rather than asked again. There is also a free account tier that tracks one brand and five questions with a single check, which exists so you can look around rather than as a way to run the service without paying. A free account gets no scheduled monthly check: the scheduler skips it.

None of the three. Every plan is month to month, in US dollars, with no setup fee and no annual lock-in, and you can cancel from the billing page without talking to anyone.

Cancellation takes effect at the end of the period you have already paid for, not immediately, and a started period is not refunded pro rata. Card details were never held by us in the first place: Stripe holds those, and we keep only a customer reference and the subscription status.

In the app, after every check: the visibility score with its three inputs shown separately, every tracked question with the answer it produced, which brands were named instead of you, every source domain the engine credited, and the recommendations that follow from that list. The citation list is the part we hand over first, because the score tells you where you stand and the cited domains tell you where the work is.

No. There is no email provider configured on this product and no message has ever been sent to a customer. The monthly report is written and tested but dormant, so the report you read is the one in the app. There is no newsletter and no mailing list either.

We write them and you publish them. That means the passage, the heading and the structured data arrive as text you can hand to a developer. We never need access to your website in order to measure you, because the measurement reads what the engine says rather than what your site contains. If you would rather we published the change, that is arranged directly rather than through the app.

No, and nobody can. What a model writes is not under any vendor’s control, there is no paid inclusion, and a provider who promises placement is describing a product that does not exist. What is controllable is whether a clear, quotable answer to the question exists on your own domain and whether the sources the engine already trusts describe you accurately. We commit to the measurement and the work, never to the outcome.

Honestly: unknown for your brand, and any vendor quoting you a number for it has invented it. What we can say is what the cadence supports. A published fix has to be crawled before it can be cited, and the engine is non-deterministic run to run, so a single month’s move is not evidence on its own. The interval we measure on is a month, and we publish our own number either way.

Zero. On the eight buyer questions we track on our own behalf, Perplexity named Bungad in 0 of 8 answers: mention rate 0%, citation share 0.0% with 0 of 90 cited sources pointing at our domain, no average position because there was no mention to rank, and a visibility score of 0 out of 100. It is published in full, with the questions, the competitors named instead and the domains the engine cited, and it is re-measured monthly whichever way it moves.

It is the fair question and it deserves the unflattering answer: our own score is the worst one available, and we are the newest name in the category. What the number proves is narrower and still worth something — that the measurement is real rather than a marketing graphic, that we will show you a report including the parts that sting, and that if the scoring flattered anyone it would have flattered us first. Nobody else in this category publishes their own audited number. If the only evidence you will accept is a vendor’s own high score, buy from someone who has one.

Four cases, and each one costs us a sale. If you only want the data and have someone in-house to act on it, a self-serve tracker such as Otterly or Peec AI is a cheaper purchase than our Starter plan. If you need a dozen markets and languages now, an enterprise platform such as Profound or AthenaHQ covers that breadth and we do not. If you need ChatGPT or Copilot numbers this month, buy a tracker that already has them, or open the Bing Webmaster Tools AI Performance report, which is free and first-party for Copilot. And if your site is under about twenty pages, spend the money on content first: measuring an empty site tells you what you already know.

Query more engines automatically, which is the big one: we hold a key for a single engine and several rivals hold more. Report by country, language or market, none of which exists anywhere in our run, our schema or our reports. Offer an enterprise procurement track — security review, a negotiated data-processing agreement, multi-year terms — which we have never done. And show you a track record of customer results, which a company publishing its first audit cannot.

The sample. One engine, a fixed question set, English phrasing, one answer per question per month. That is enough to be certain about an absence and not enough to read meaning into a small move, which is why every figure we publish names its stored run and why we say plainly that per-domain citation counts are estimates rather than measurements. The score and the brand tallies are the measurement; the individual counts around them are not.

No, and treat anyone who offers to as a warning. A visibility score is a weighted composite of a particular question set on particular engines with a particular matcher, so two vendors’ numbers are not the same unit. Ours is comparable against itself over time and against the competitors measured inside the same run, which is the only comparison that means anything.

Per account: your email address, name, a hashed password, the organisation name, the brands and domains you track, the questions you track, and the engine answers and citations returned for them. Tracked question text goes to Perplexity, because that is the engine being asked. Payments go to Stripe, storage to a managed PostgreSQL database, hosting and request logs to Vercel, and typeface requests to Google Fonts. No other third party is called, nothing is sold, and there are no analytics, advertising or tracking scripts anywhere on this site.

No, and the clash is real enough to change the words we use. In the Philippines, AEO usually means Authorized Economic Operator, a Bureau of Customs accreditation programme for importers and exporters, and it owns the acronym locally. When the audience is not already fluent in marketing vocabulary we say AI search visibility instead, because competing for that acronym locally means competing with the Bureau of Customs.