Maria Siasat.

Senior marketer. Curious builder. Relentless experimenter.

Why this site exists

I build projects to understand how AI, automation, data and emerging technology are changing the way marketing gets done.

The habit

After years leading enterprise marketing programs, I still believe the best way to learn something is to build it, test it and find out where it breaks.

Explore the experiments View my professional background

Latest finding1 Aug 2026Ten AI-drafted hooks cost about twenty minutes. Choosing the three worth filming still takes judgement, and that judgement is now the scarce skill.

Enterprise experience across Salesforce, Optus and high-growth technology businesses

$95M+

Pipeline and revenue contribution

764

C-suite leaders engaged

10+ yrs

Enterprise and complex B2B marketing

ABM

Growth, executive programs, AI experimentation

My résumé shows what I have delivered. This site shows how I think.

The log

What I found this week, before I found a way to make it sound impressive.

A running record of the experiments: what I built, what it cost me in time, and the mistakes worth repeating out loud so nobody else pays for them twice.

1 Aug 2026Creative

Ten AI-drafted hooks cost about twenty minutes. Choosing the three worth filming still takes judgement, and that judgement is now the scarce skill.

18 Jul 2026Integration

Review Optimiser front end: two hours. Wiring the outbound agent into Instantly and testing the sequencing: four to five days. The gap is the whole story.

2 Jul 2026Voice AI

Bramwell AI held a convincing coaching conversation in week one. Making the score mean the same thing twice took months.

14 Jun 2026Outbound

Signal-led outreach ran three to five times standard SDR reply rates. The lift came from the research step, not the copywriting step.

30 May 2026Research

An AI research agent writes a clean report in minutes. Verifying every claim in it is where the days go. Accuracy is a sourcing problem.

Why I build

Why I keep building things

I have spent much of my career creating enterprise marketing programs, account-based campaigns and executive experiences. But reading about emerging technology is not enough for me.

I want to understand what happens when you connect the landing page, the AI agent, the data, the payment system, the follow-up and the actual customer experience. So I build things.

Some ideas work quickly. Some become far more complicated than expected. Some reveal that the technology is not yet ready. Every project teaches me something I can bring back into my professional work.

This is not a startup portfolio or a collection of polished success stories. It is a record of curiosity in action.

The experiments

Projects built to test a question, not to look finished.

These are projects I have built to explore a question, test an assumption or learn a new capability. They are not all finished businesses. Some are prototypes. Some are live experiments. Some have changed direction. Some have exposed how difficult it is to turn a clever idea into a reliable customer experience. That is the point.

Experiment 01

Review Optimiser

Can automation help small businesses build trust without adding more work?

Current phaseMarket and workflow testing

The question

Could reputation management, customer communication and business data be combined into a simple system that helps small businesses improve their online trust signals?

Why I became curious

Many small businesses depend on customer reviews but do not have the time or systems to consistently request, monitor and respond to them.

What I built

The positioning
Customer journey
Website
Review audit concept
Outbound nurture sequence
Payment flow
Reputation-management workflow
Supporting automation

What I used

Lovable
Lovable Cloud (Postgres, auth, storage)
Zapier webhooks
Instantly nurture sequences
Claude (SEO and AEO agent)
OpenClaw / Helena by Enrich Labs
Google Business Profile reviews
Resend email

What surprised me

The visible website and messaging were the easy part. The harder work was connecting the systems, validating the data, creating reliable handoffs and making the full journey feel simple for the customer.

What became harder than expected

The front end took two hours. Connecting the outbound agent to Instantly, and testing the sequencing, took four to five days.

What I would do differently

Design the data and handoffs first, then build the front end. I built the story before I built the plumbing, and the plumbing set the pace.

Key learning

Automation only creates value when the customer experience still feels coherent, trustworthy and human.

www.reviewoptimiser.com Keep me posted

Experiment 02

Bramwell AI

Can an AI voice coach help people stop rambling and communicate with greater confidence?

Current phaseIntegration and testing

The question

Could a voice-based AI coach give people useful, immediate feedback before an interview, presentation, performance review or difficult professional conversation?

Why I became curious

Most people know what they want to say, but struggle to structure it clearly under pressure. I wanted to explore whether AI could provide realistic practice, personalised feedback and a measurable readiness score.

What I built

A voice-coaching concept using conversational AI, structured feedback, practice scenarios and a readiness score.

What I used

Realtime voice AI
Speech to text transcription
Conversation design
Scoring and feedback logic
Lovable Cloud (Postgres, auth, storage)
Stripe subscriptions and paid tiers
Edge functions for session scoring

What surprised me

Creating a convincing conversation was faster than building a reliable end-to-end product. The difficult work was connecting the voice interaction with meaningful scoring, customer data, payments, reporting and a consistent user experience.

What became harder than expected

The first voice experience was quick to prototype, but reliable scoring, payments, customer data and follow-up required much more operational design.

What I would do differently

Treat the scoring model as the product from day one, instead of treating it as something to add after the demo felt good.

Key learning

A strong demonstration is not the same as a dependable product.

www.bramwellai.com Keep me posted

Experiment 03

Business Trust Council

Can public trust signals become useful commercial intelligence?

Current phaseConcept and research framework

The question

Could review volume, recency, rating quality and other public signals be used to create meaningful trust benchmarks for different industries?

Why I became curious

Businesses are constantly told that trust matters, but most have no clear way to understand how they compare with competitors or which signals are weakening customer confidence.

What I explored

Industry benchmarking
Research-led marketing
Trust scoring
Data storytelling
Category creation
Lead generation
Research as a commercial asset

What I used

Claude research agent
Structured source verification
Automated report generation
Lovable Cloud (Postgres, auth, storage)
Resend email

What surprised me

Accuracy is a sourcing problem, not a writing problem. A clean draft appears in minutes. Verifying every claim is where the days go.

What I would do differently

Fix the methodology and the sampling rules before generating a single report.

Key learning

Research becomes commercially useful when it leads to a clear decision or action, not when it simply produces more information.

An experimental research and benchmarking concept, not an established authority.

www.businesstrustcouncil.com Keep me posted

Experiment 04

Signal-led B2B growth experiments

Testing whether account signals, AI research and highly personalised outreach can outperform generic demand generation.

Current phaseLive experiment

The question

Does relevance beat volume? If a message is tied to a real trigger inside a real account, does the economics of outbound change?

What I tested

Identifying accounts showing live buying signals
Mapping signals to an ideal customer profile
Prioritising accounts by urgency
Creating one-company, one-trigger outreach
Building executive roundtables
Testing benchmark reports as lead-generation assets
Connecting marketing signals to weekly sales actions

What happened

Reply and conversion rates running three to five times standard SDR outreach
Executive forums generating approximately four to six pipeline conversations per program

What surprised me

The research step, not the writing step, is what made outreach land. Once the trigger was right, the message almost wrote itself.

What I would do differently

Build the weekly signal-to-sales ritual earlier. Signals decay quickly, and a good insight delivered late is just trivia.

Key learning

Personalisation is judgement, not merge fields. Relevance comes from the trigger, the audience and the timing.

Specific client and account details have been withheld for confidentiality.

Keep me posted

Experiment 05

Creative-led content engine

Can one person plus an AI team run a short-form content engine that actually performs, not just posts?

Current phaseLive experiment

The question

Short-form video is now where attention and buying research start, including in B2B. Can a single operator use AI to research, script, edit and iterate at creator speed, and still keep taste and a human voice?

Why I became curious

Most B2B teams still treat creative as decoration on top of targeting. On short-form platforms the creative is the targeting. I wanted to feel that difference myself rather than read about it.

What I built

Hook library scored by performance, not opinion
AI research agent for angle discovery
Script drafting with a fixed voice guide
Batch shooting and cut-down workflow
Weekly test-and-learn review
Repurposing into long form and email

What I used

Claude for research and scripting
AI video editing and captioning tools
Analytics exports for hook and retention scoring
Lovable Cloud for the tracking sheet

What surprised me

Volume is easy now. Taste is the constraint. AI produced ten usable scripts in the time it used to take to write one, and the deciding skill became knowing which three were worth making.

What I would do differently

Score hooks before producing anything. I made polished videos on weak premises for the first two weeks.

Key learning

Creative is the performance lever, not the wrapper. The team that iterates on hooks fastest wins the auction.

Run as a personal test of creative-led performance thinking, not as client work.

Keep me posted

What I'm learning

What building these projects is teaching me

The prototype is rarely the hard part

AI can make the first version look impressive very quickly. Reliability, integration and customer trust take far longer.

Automation exposes bad processes

Connecting more tools does not fix a confused customer journey. It usually makes the confusion move faster.

Data without a decision is decoration

Research and signals are only useful when they help someone decide what to do next.

Personalisation requires judgement

Adding a company name to a message is not personalisation. Relevance comes from understanding the business trigger, audience and timing.

The final 30% is where trust is built

Payments, reporting, support, failure handling and handoffs are less exciting than the original idea, but they determine whether the product can be trusted.

Creative is the new targeting

On short-form platforms the hook does the work the audience settings used to do. Iterating on creative beats refining the segment.

Building changes how I lead marketing

Hands-on experimentation makes me better at assessing technology claims, briefing specialists, identifying operational risks and asking better questions.

About

A working notebook, not a portfolio.

I am Maria Siasat, a senior marketer based in Sydney.

My career has taken me through enterprise technology, telecommunications, hospitality, account-based marketing, executive programs and growth strategy.

Outside my formal roles, I build projects. Not because every idea needs to become a company, but because building is how I learn. I enjoy moving from a question to a prototype, discovering what I underestimated, and translating the lesson into better marketing strategy and execution.

Particularly interested in

Marketing
Customer experience
AI
Automation
Data
Trust
Human behaviour

Curiosity is part of how I work

I am usually exploring a question.

Can AI improve the quality of a difficult conversation?
Can public trust signals reveal a commercial opportunity?
Can buyer intent make B2B outreach more relevant?
Can a small team create an enterprise-quality customer journey using modern tools?

The projects change, but the motivation is consistent. I want to understand what is genuinely useful, what is mostly hype and what becomes possible when strategy, creativity, data and technology are connected properly.

This curiosity is not separate from my professional experience. It is one of the reasons I continue to grow as a marketer.

Supporting credibility

The professional background behind the experiments

I am a senior B2B and enterprise marketer with experience across Salesforce, Optus and high-growth technology environments. My professional work has included account-based marketing, executive engagement, field marketing, integrated campaigns, go-to-market strategy and close partnership with enterprise sales teams.

At Salesforce, I led enterprise ABM and executive marketing programs across ANZ, including the World Tour Executive Experience, private executive briefings and persona-led programs for complex buying groups. At Optus Business, I led demand generation supporting enterprise mobile accounts and worked closely with sales teams and agency partners, including Saatchi & Saatchi, to move from business objectives through to campaign execution.

I have also co-founded and helped scale Made in Italy from one location to multiple venues across Sydney, giving me direct experience of brand building, customer experience and business growth.

For full employment history and role details, visit my LinkedIn profile or download my résumé.

View LinkedIn Download Résumé

Contact

Have an interesting problem, idea or experiment?

I am always interested in conversations about marketing, AI, customer experience, enterprise growth and the practical reality of turning new ideas into something useful.

Connect on LinkedIn Email Maria View Résumé