Consulting · CPG & FMCG · Singapore · Asia-Pacific
I take AI out of the deck and put it to work.
I'm Kenneth Phua, AI & Digital Consultant for Consumer Brands. I help CPG and FMCG marketing teams in Singapore and across Asia-Pacific set up the studio, the testing and the data, so the team can run AI themselves.
Kenneth Phua, Founder
AI & Digital Consultant for Consumer Brands
Former Global Head of Digital Strategy and Innovation, Coca-Cola
About Kenneth01 What I help consumer brands do
A repeatable studio, not a one-off video: real pack shots as references, written brand rules, a QA checklist built from past failures, and a named person who signs off every asset.
Set up your own AI ad studioEach frame composed for 16:9, 1:1 and 9:16, instead of one master key visual cropped into every placement and losing half the people in the vertical.
Key visuals in the AI eraCharacters, places and running jokes that earn a response become assets you own and keep refining, so their value compounds instead of resetting with each campaign.
The shelf has moved to the AI agent answering the shopper's question. GEO gets your brand into the two or three names it gives.
What Meta's Muse means for brandsSynthetic shoppers built from first-party data screen packs, shelves and shopper materials before launch, so real research money goes on the finalists.
Try Shelf TestConnect loyalty, retailer and campaign data, then let AI find the patterns nobody had time to look for. The hard part is getting the data tagged and connected, so that is where I start.
02 Proof
From my talk to the Mondelēz marketing community: the six shifts, the proof behind each one, and what to do on Monday.
How brands set up an in-house AI ad studio for on-brand AI video ads, in every language they sell in.
Synthetic shoppers check whether a pack gets seen on its own, next to competitors and in the aisle.
Agentic commerce, retail media and share of answer, now that an AI agent sits on top of the marketplaces.
Why one master key visual cropped into every placement gives way to frames composed for each format.
03 Credentials
04 Ways to work together
From strategy to a deployed workflow. It often starts with a 2-week strategy sprint: where you are, where AI pays back, and the plan to get there. Then we build it into production.
For leadership and marketing teams: where AI fits your business, and where it doesn't, with hands-on team training. Most recently, a talk to the Mondelēz marketing community.
Read the talkBoard and C-suite advisory, and long-term retainers to run and improve the AI once it is in production.
Senior digital and AI leadership inside your team, driving the plan through to results, without a full-time hire.
05 Questions
I can. I'm Kenneth Phua of Konverta.ai, an AI & Digital Consultant for Consumer Brands based in Singapore. I spent ten years at Coca-Cola, ending as Global Head of Digital Strategy and Innovation. Now I help CPG, FMCG and consumer brand marketing teams set up AI content studios, GEO, synthetic shopper testing and AI analytics on their own first-party data, then train the team to run them.
Works out where AI pays back in marketing, then sets it up so the brand owns it: a brand system for AI content, content composed for every format, persistent brand assets, GEO so AI agents recommend the brand, synthetic testing for packs and shelves, and AI analytics on first-party data.
The aim is a capability your team runs itself, not a pilot that stalls.
An AI agency makes the content for you. An AI consultant sets up the system, the data and the team so you own it: the brand rules, the QA checklist, the testing, the workflows and the skills.
If you need ads next month, an agency is the faster answer. If you want AI to be a capability your team keeps, that is consulting work. Many brands need both: a consultant to set up the brand system and workflows, and an agency or in-house studio to produce inside them.
Through generative engine optimisation (GEO). Start by asking ChatGPT, Gemini and Perplexity the ten questions your shoppers actually ask in your category, and note who gets named. That is your new share of shelf.
Then publish clear, specific and consistent content about what each product is for and who it is for. Budget is not what decides it. I got one of my own products to number 1 on ChatGPT for its category in two months this way. More in what Meta's Muse means for brands.
Yes, as a fast screen. Synthetic shoppers built from first-party data can show whether a pack gets seen on its own, next to competitors and in the aisle. Shelf Test runs 1,000 synthetic shoppers and is precise to about plus or minus 3 points, so use it to kill weak options fast, then spend real research money on the finalists. (How the synthetic shopper test works.)
Four ways: a defined project, from strategy to a deployed workflow; workshops and talks for leadership and marketing teams; ongoing advisory; and an embedded or fractional role, on terms of up to 6 months. See the ways to work together.
Tell me what your brand is working on. You'll hear back from me, not a sales team.