Madeleine Patel, co-founder of Speak, talks to Tom Raynel about changing the core of the business mid-development, and her hopes for the future of artificial intelligence and customer interaction. Each Monday, we interview a small business owner, which is now a regular feature of NZME’s editorial campaign On The Up,
Premium
AI start-up Speak eyes $4t market with personalised shopping tech
Subscribe to listen
Access to Herald Premium articles require a Premium subscription. Subscribe now to listen.
Already a subscriber? Sign in here
Listening to articles is free for open-access content—explore other articles or learn more about text-to-speech.
Madeleine and Ankit Patel, co-founders of Speak, a retail AI start-up hoping to build the infrastructure layer for retailers to power hyper-personalised shopping experiences.
AdvertisementAdvertise with NZME.
The business was originally Speak Scents, what changed?
It’s quite a long story but really way back, when I originally had this thought and this idea, I really wanted to change the experience that people had with shopping for perfume, for fragrance. I thought it was a category that was really outdated and hadn’t seen a lot of change.
Around 2017 I was working for a data science agency and becoming far more aware of how predictive technologies were being built and used to drive decision-making. I also met my now husband and co-founder who was a data scientist there.
We got deep into how consumers were making decisions using different technologies, and we decided that fragrance was a modelling playground in the sense that those products were made up of tons of different data points. We could effectively write predictive models to help deliver outcomes for consumers using all of those different data points across those products.
AdvertisementAdvertise with NZME.
We started talking to retailers about 10 or 11 months ago and they said that there’s something super interesting here if we could use this tech to drive decision-making across our categories. That was a kick for us to go, ‘Right, we’ve got this opportunity to make a choice to work out which of these product offerings is more scalable.’ We realised pretty quickly that the tech offering was a much bigger opportunity.
What are the opportunities in consumer-focused artificial intelligence models?
Where we want to play is to power the personalisation layer within the $4 trillion e-commerce market, and, more specifically, to lead the charge in what we call the agent-based shopping market within that.
Agent-based shopping is using machines to power decision-making for consumers. We believe that’s a wave that’s about to hit retail, and retailers are really unprepared for how to surface that to customers and consumers. We’ve got what we believe will become a fast-growing niche within agent-based shopping, and then we’re starting in the fragrance space and the beauty category, and we’ll expand that out.
We’re in the early stages with that, but we’ve done a really successful pilot with Chemist Warehouse. They are the largest-selling fragrance retailer in New Zealand, and that went really well.
What are your plans for the next few years?
We definitely don’t want to stay in the New Zealand market. New Zealand’s been a great test ground and will continue to be, but it’s just not a big enough market for us at the moment. In the next 12 months we definitely want to be focusing on the Australian market with enterprise and the US market with small-to-medium business customers.
It’s building out into other categories within the next 12 months as well. The next ones we want to focus on are skincare and supplements. Then down the track, we also want to build into more emotionally driven categories outside the beauty space like wine, for example, which is quite hard for consumers to articulate in terms of what they want, but has a lot of variables that go into how you make that decision around that product.
AdvertisementAdvertise with NZME.
What would be your advice to other budding entrepreneurs wanting to start a business?
This might sound cliche, but just start and figure out what you can learn or what you can understand from not much. Start to test your ideas or your problems that you think you can solve, and I think just starting to understand the context and the landscape that you’re working in is a great start.
Tom Raynel is a multimedia business journalist for the Herald, covering small business, retail and tourism.