Online OptimizationHoldings Pte. Ltd.

Incorporated in Singapore · Private limited company

Visibility
inside the
answer.

Search used to return a list of documents. Generative engines return a finished answer, and increasingly a software agent acts on it. Online Optimization Holdings builds the infrastructure that determines whether a business is present at that moment — across six disciplines, operated as one chain.

JurisdictionSingapore
Entity typePte. Ltd.
UEN202632024G
Incorporated2026
StagePre-revenue
Disciplines06
Engines tracked06

01 — Thesis

The link
is no longer
the destination.

For twenty-five years, discovery meant a ranked list of documents and a click. Generative engines collapse that into a synthesised answer. The commercial question changes with it: not where a page ranks, but whether it is drawn into the answer at all.

A second shift is now building on top of the first. Software agents are beginning to read, compare and transact on a buyer's behalf. An agent does not browse a site, weigh a brand story, or respond to a banner. It reads structured data and acts on it. A business that is not legible to that reader is not considered by it.

Both shifts point at one requirement: a business must be readable by the machines that summarise, and by the machines that decide. That is what this company builds and operates.

02 — Disciplines

Six links,
one chain.

These are not six services sold separately. Each link produces the input the next one depends on: an unindexed page cannot be structured, an unstructured page is rarely cited, and an uncited business is invisible to an agent. Each link is defined below, with the measure that governs it.

L01
SEOSearch engine optimisation
Make the source retrievable. Technical health, crawl access, index coverage, site architecture and page authority. The established discipline, and still the substrate: an engine cannot cite what it cannot reach, and strong generative visibility is generally built on sound conventional foundations rather than instead of them.
Governed byIndex coverage · Crawl integrity · Ranked position
L02
AEOAnswer engine optimisation
Make the source answerable. Structuring material so a machine can extract a discrete, quotable answer to a specific question: schema markup, entity definition, question-level formatting, factual density. AEO governs the shape of the material; GEO governs where that material lands.
Governed byExtractability · Entity resolution · Answer coverage
L03
GEOGenerative engine optimisation
Make the source cited. Earning inclusion in the answer a generative engine composes, where a response typically draws on only a short list of sources. Measured as citation share across engines, not as position on a page. This is the discipline the visibility console instruments.
Governed byCitation rate · Share of model · Attribution integrity
L04
SEMSearch engine marketing
Buy the position the chain has not yet earned. Paid placement against commercial intent. Its role here is temporal: it holds presence while the earned layers mature, and it prices demand precisely enough to tell the rest of the chain which queries are worth pursuing at all.
Governed byCost per acquisition · Impression share · Intent coverage
L05
SMMSocial media marketing
Generate the corroboration engines look for. Beyond audience reach, social and community surfaces are themselves retrieval material. Consistent third-party mention is part of how a model resolves whether an entity is real, current and credible enough to cite.
Governed byMention volume · Sentiment consistency · Entity corroboration
L06
AGENTICMachine readability
Make the business transactable by software. Product, price, availability and policy data published in the structured forms autonomous agents consume. The newest link, and the one with the least settled practice — treated here as an infrastructure requirement rather than a campaign.
Governed byFeed completeness · Schema validity · Agent retrievability
A note on terminology SEO, SEM and SMM are settled categories with stable definitions. AEO and GEO are recent, and the industry uses them inconsistently — some practitioners treat them as one discipline. This company separates them because they are governed by different measures: AEO by whether a machine can extract an answer, GEO by whether an engine cites it. Where a client uses the terms differently, scope is defined by the measure, not the label.

03 — Agentic layer

When the
buyer is
software.

Agentic systems delegate a purchase decision to software: a person sets intent and limits, and an agent handles discovery, comparison and — increasingly — the transaction itself.

An interoperability layer is forming around this, with competing protocols published by major platform and payment operators. Those protocols remain unsettled. The requirement they impose does not: a business that cannot be read by a machine cannot be selected by one.

What changes
The reader. Persuasion, layout and brand atmosphere are addressed to a person. An agent evaluates structured attributes — specification, price, availability, terms, return policy — and discards what it cannot parse.
What it needs
Published, verifiable data. Complete product and service feeds, valid schema, machine-readable terms, and an entity definition that stays consistent across every surface an agent may consult.
What we operate
The readability layer. Auditing what an agent can currently retrieve about a business, correcting the gaps, and maintaining that data as protocols change. Infrastructure work, delivered on a retained basis.
Not claimed
Influence over agent selection. No supplier can guarantee that an autonomous system will choose a given business. The controllable variable is eligibility: whether the business is legible, complete and current enough to be considered at all.

04 — Instrumentation

Visibility
console.

The same query, run against six generative engines, with citation state reported per engine. Rows that turn yellow are engines that cited the source. Enter a query to run it.

Visibility console Idle
01ChatGPTNot run
02Google AI OverviewsNot run
03PerplexityNot run
04ClaudeNot run
05GeminiNot run
06CopilotNot run

Illustrative demonstration. States are simulated for presentation and do not represent live engine responses.

05 — Corporate record

Structure
and standing.

Particulars as filed with the Accounting and Corporate Regulatory Authority. Stated as at the date of this page, and not a substitute for a business profile extract, which is available on request.

Registered name
Online Optimization Holdings Pte. Ltd.
UEN
202632024G
Incorporated
July 2026
Entity type
Private company limited by shares
Jurisdiction
Republic of Singapore
Registered office
Level 39, Marina Bay Financial Centre Tower 2
10 Marina Boulevard, Singapore 018983
Shareholding
Wholly held by a single shareholder
Financial year end
31 December 2027 (first financial period)
Operating status
Incorporated; commencing operations

06 — Operating model

How the
business runs.

Three functions, sequenced. Measurement precedes intervention, and intervention is only ever billed against a measured baseline.

01

Measure

Establish a baseline across all six links for a defined query and product set, and report it as a fixed-scope engagement.

02

Operate

Work the links the baseline identifies as weakest, under a retained monthly mandate with a named scope and a named measure.

03

Report

Re-run the baseline on a fixed cycle and report movement against it. The measurement is the deliverable, not a by-product.