Enter Juno, a company that its founder Michelle Gilmore says aims to use artificial intelligence to build “the listening engine for the world.”
They say that the purest form of love is attention, and Juno aims to pay better attention to what people think and feel – and why they think and feel that way – in ways that help users as varied as governments, global corporations to childcare centres, and RSLs truly understand what their customers need.
Juno is Michelle Gilmore’s fifth business, and the first time she’s taken on outside capital to help her build it: $2.8 million of venture capital in a round led by Nick Crocker at Blackbird.
We caught up with Michelle Gilmore in Barangaroo, Sydney. We got nerdy, we got deep, we got a very good insight into a woman who pushed things as far as she could in one industry (human-led behavioural research) and is now embarking on a journey to take those experiences to be at the forefront of building another.
Here’s Michelle.
Natasha Gillezeau: So, this is a question that I love to ask everyone that we interview for the series. We have a contributor called Cara Davies, she built and sold a Gen Z fitness app called Steppen. And she recommended a book to us called The Unfair Advantage. The thesis of the book argues that all entrepreneurs have some kind of unfair advantage. It could be connections, intelligence, where they were born, or having a tough upbringing and having to weather the challenges of that. So my first question to you is, Michelle, what do you feel is your unfair advantage?
Michelle Gilmore: I’m really uncomfortable with comfort and certainty. It genuinely freaks me out. Even the thought of it, I’m a bit wriggly. Therefore, all of the things that we talk about in terms of the successful entrepreneurial mindset, I naturally lean that way. If anything, I seek it out. I think that’s probably a little of an advantage. My advantage with Juno is that I did the job that Juno does for fifteen years. So I have a competitive moat that you can’t buy. And in the AI race, that’s incredibly valuable. Priceless, if you will.
Love it. Do you feel with the AI space – it’s getting tacked onto lots of different startups, it’s quite buzzy – do you feel well positioned to discern what’s real AI and not?
I think I’ve got a reasonably tuned read on bullshit versus not. And there’s a lot of spin like there is in every peak until it levels out. And when you scratch under the surface of some of the products, services and systems that are spinning up because of AI there’s not a lot of substance under there. That doesn’t mean they won’t be successful, which I think is an important distinction to make.
For me, a great use of AI for humanity is purposeful, and considered. And targeted use cases that are going to do great things for the world. That’s how I filter and gauge, I think. I’m learning like anyone else. But as an industrial designer and systems engineer, my whole 25 years has been about trying to translate or augment or identify behaviours and then build them into systems. So, AI is a space that I feel reasonably comfortable in. Thinking about it as an organism feels quite natural to me. I’m not going from thinking in 2D ways to trying to think living, breathing organic mode, which I do think some practitioners are really struggling with.
Can you tell me a little bit about that? So it’s like, it sounds like there is a machine way of thinking about it, and living systems way of thinking about it?
Yeah. The unlocking that’s going to happen when AI is embedded into systems and services in very real ways, and not just slapped on the outside. And to embed and integrate into the system in very real ways, you have to think about the system. And so, the quick short term slap ons, they may work, but they’re not necessarily what gets me excited about AI.
We’re in the present, and we’re talking about technology and AI. But can you take me back to your upbringing and how we got to this moment in time?
Sure. I grew up in small town Australia, with a lower class working family. My Dad was a security guard, and my Mum worked in a supermarket. We didn’t have much money. And I didn’t really know we didn’t have much money. Very loving, values-led parents. We had to contribute, it was all about ethics, and working in the family and around the family. I’m really proud of some of the generational cycles that my parents broke. I think that’s a hard thing to do.
I went to public school. I did well because I wanted to, not because it was easy to, I got a scholarship to go to university, I worked from the time I was 15. And have worked for myself since I was 21. I’m a really bad employee; I wanted to see the whole board, run the plays myself and build something from the ground up. There’s always been this deep sense of wanting to push, wanting to do more, constantly progressing is important to me. I’m deeply triggered by injustice, I’m incredibly passionate about systems, products and services that are trying to do good.
What do you think you notice about injustice that other people sometimes don’t notice or take into account?
I think for me it’s the little things that accumulate over time, and we need to get practical. Yesterday, I was talking to someone about potential candidates, and they described one of them as being “accented”. And I said “I don’t know what that means”. And he said “he’s got a really strong accent”.
And I stopped him and zoomed out and said “we all have accents. That is not okay at Juno. What you’ve just brought into my business will not be accepted here. Now let’s talk about why you think that’s okay, and why you think as a citizen in Australia it’s okay to label someone that doesn’t sound like you like that.” That’s just a very real time example that happened yesterday that I was very triggered by.
I think in Western-dominated culture we often prioritise doing what is ‘polite’ over what we know is right. And we’ll tell our friends after work, or talk about it behind closed doors, but not to the person direct. Which I think is cowardly, and doesn’t create any change.
Let’s get into Juno, and what you’re actually building.
We are building the listening engine for the world. If you think about how most people are using AI, they are asking the language model to generate something, pull information. We’re inverting that. Juno listens, and gathers information, context, experiences from humans so that our customers can make better decisions.
Childcare services are using it so they don’t have to deal with parent feedback. A local RSL used it to change their menu. It’s a way to quickly hear people, and understand the reasons why they’re saying the things that they’re expressing and feeling. And then acting on that.
After the US election, we spoke to 700 Australians in 24 hours straight after it was called. And we now have a dataset, which is 120 hours of conversation across multiple languages that happened across 24 hours. If I had my old human-led team on this, maybe I could have thrown 10 people at it, and maybe we could do 100 conversations, and we’d need 3 months to synthesise that data and a large budget.
Juno is doing behavioural investigation into humans at unprecedented scale and speed, and at a quality I’m proud of. It tries to qualify subjective terms, it pursues the why behind human responses, it will soon place responses in context, it considers demographic differences and how they might be changing or not changing responses. It’s an incredible scaling of something that I’ve been working on for 20 years. I’ve always been frustrated with the ceilings we’d hit with a human-led team. We had offices in New York, London, Sydney and it was a really great, successful business. But not many humans are good at it, and by that, I mean asking non-leading questions in a semi-structured interview style without putting their own bias and opinion into it.
Very hard.
They’re very bad at asking for examples. In some cultures, it’s seen as disruptive. I tried to launch a whole training company to teach humans how to do it, and it failed. And it failed because my hypothesis was wrong, I thought if more people could do it, we could scale and the world could do more investigation into human behaviour. That was the wrong question to ask. The question to ask was how can we augment our unique IP and experience with technology to scale this capability so that more people can access high quality behavioural data and more people can participate in giving that data, on their own terms.
Are you at the phase of having a customer? What are you learning from these initial customers?
We started working with customers straight away. We bootstrapped for our first 6-9 months, my co-founder Josh built an incredible product himself. We were able to onboard a sweep of early testing partners, and by that I mean businesses that were willing to experiment with us. EY, Innocean Australia, which is a creative/media agency, NAB, Leukamia Foundation, an RSL club, a local childcare centre, someone used it for their wedding to talk to their guest list. That’s a cut-through of examples that speaks to broad use cases and a nice set of early customer pull signals.
Tell me about the wedding example.
A customer used it to talk to their guests about their dietary requirements, accommodation needs, it was a destination wedding. And they used Juno to inform their seating plan.
And it can do all that? How?
So based on what they told Juno they wanted to find out, Juno went and had a contextual conversation with everyone on the guest list. Just like a wedding planner would. I assume that a wedding planner would take care of that, Juno is taking care of that.
The childcare centre is hearing everyday from parents that are giving feedback. They don’t have the operational ability to gather, filter, synthesise – they cannot wade through texts, emails, and face to face conversations. Juno is now facilitating those conversations for them via a simple, known interface.
And is having a one to one conversation with each of these parents and caregivers, and feeding the outcomes of that straight to the care centre owner. And they can act, or not act, or just see what happens. Is this a pattern, are more people saying this, is this an outlier, etc.
It’s very interesting when we think about use cases for organisations, businesses, that have not been able to stand up a capability like this, or have been trying to filter through and make sense of data without the teams to do it, and Juno does all of that for them.
What are some of the most interesting conversations that you’ve had with these customers that have unlocked something for the build of Juno?
I think for me, the things that give me goosebumps are the insights that led to change actually happening. So, an RSL club changing their menu. That may sound small, but I think that’s very cool. I think it’s cool that they were using feedback forms, and then they asked members to have a conversation with Juno, and they realised their menu was not working for x reasons, and they changed their menu, and in the following weeks saw better revenue. That is a very successful customer journey for us, and for me, and one I’m proud of because the manager of that club was able to do something they hadn’t been able to do before without spending more time and money.
How does Juno compare to Typeform or Dovetail?
Interesting, I was at Dovetail recently. They are a remarkable company, they are an insights company and Benjamin and his team have done such inspiring things.
[MP reader note: highly technical details cut for brevity, but in essence Dovetail is all about housing different data sources and presenting largely internal UX/UI research design teams. Juno is the product that goes and retrieves insights from humans in a qualitative way. We also spoke about how Typeform, which is a conversational survey is more ‘shallow’, Juno is intentionally deep.]
Not to talk so much about Missing Perspectives, but that’s a good example right? Because that’s not necessarily a capability we have today, to deeply understand the effects on a reader.
You don’t need the capability, because we’ve got it! So at the end of this article, we should just put a Juno link, and anyone who reads or listens to it, can talk about it. So it can not only carry the conversation forward, which is very important if you’re thinking about creating content that lives beyond the piece that media is missing. Juno can keep the conversation going, which means you get more engagement, but help you to understand more about how that piece is going in relation to your objectives as a publication or network.
One media company that is good at this is Disney. They get a lot of value out of the assets. Your co-founder Josh Davey made a post about the speed at which Juno could compile shopper research after an ACCC case against Coles and Woolies. Would you say that speed is one of the main benefits that you offer?
Absolutely. Within 24 hours, we were able to give Danielle Long at The Australian real time sentiment on how shoppers felt about the case.
Never before have journalists been able to have near real time sentiment analysis, that they can then spot check via transcripts, pull quotes, look at context around it, and have automated pattern matching as it relates to how people felt en masse or not, would it change shopper behaviour, would it not, did people know about shopper behaviour. The speed is one component, but speed alone would not work… it’s the combination of speed, quality, and actual human insight.
What does your role as CEO involve?
Today or generally?
Maybe let’s go over the last week.
Good time frame device use there! It’s nice to talk to someone who is good at interviewing. I want to point out that what you just did there is something that Juno would do.
Haha, it’s getting very meta up in here.
But it’s really important, right? I think that using examples helps. What you did is gave me a time frame that would make me not going into long term memory, but use short term memory, which by and large is more reliable. And make it real enough that everyone listening would actually get some real examples. That’s what Juno does.
So there’s something about being the custodian of the mission and the purpose and having zero tolerance on it being anything other than the purest version of that. And I’m not saying I do that well every day, but I’m saying that I’m conscious of that being my responsibility, not only when I’m talking externally, but internally too. There’s no room to move on that, there’s no room to move on the pace at which we need to run, there’s no room to move on quality, there’s no room to move on ensuring that we are reinforcing our technology in ethical ways.
It’s my job to make sure that my team are hitting the output they need to and to make sure that the product is heading towards the mission, and using a combination of evidence and intuition to do that every day.
In that guardian role in the business, is there anything that has challenged that? A conversation with an investor? Or a customer has asked you to do something and you’re like “no way”?
Yeah, absolutely. Someone asked me last night if they could use Juno to prove that everyone loves their brand. And I said no. You could use Juno to prove or disprove if everyone loves your brand… I got this all the time as a practitioner in an agency. Particularly commercial organisations that had agendas would try and use qualitative research to back up an opinion they already had. Often, these were not outlier cases. My opinion was always, if you’re willing to lead with us and find the real answers, then we are the business for you. If you’re trying to skew results, that’s not going to work.
The great thing is we’ve now built technology that is the custodian of that, and I no longer have to do that as a human, which gives me goosebumps. We have built that ethical framework into this technology. It’s not perfect, but the intent is we will not enable humans to skew results or data for the betterment of their own agenda.
There are some really interesting ethical considerations in the building of this startup that I’m glad you’re thinking about.
I think about it a lot, and the behaviours that we are endorsing or not. Anyone who is building a product or service should do that, but particularly when we’re talking about a technology that is going to scale to this pace, everyone has their own moral compass, but I believe that we need to be responsible and accountable for the decisions that we’re making that will be acted out by this technology. Juno is built to listen to humans. It doesn’t have an opinion on good or bad, love or hate, it’s about how do I listen and ask well to understand you as a human.
You could probably bring this into couples as well.
How does Juno become the Esther Perele of AI? Ha! It would be interesting to hear and think about how she’d be able to dive deeper into the thoughts and feelings of a couple before a session. The use cases are infinite.
What’s Blackbird’s role in the company?
They are our lead investor. They led our first funding round of $2.8 million. We are incredibly excited to have them on board and we think of them as partners.
So that was a good meshing. I guess something that maybe not all people would realise is that different forms of investment comes with different terms and expectations.
I think we need to increase (particularly early-stage) investing education, and the options there. For us, it’s an incredible raise, it’s our first raise, it gives us an excellent amount of proportional capital to hit the milestones that we need to hit in the timeframe we want to hit them. But Venture is one way to raise, there are many others and knowing who you are attaching your company too is important, not only at the obvious levels (like legal) but who are you actually jumping into this permanent arrangement with and is their character something that aligns with you and your business.
Well congratulations, there are many parts to building a company but it’s a huge accomplishment and I wish you the best of luck to your team. Anything you want to add that I’ve missed?
You get 30 interviews for free via Juno, so people can try the product out. Get in there and use it.
Thank you so much Michelle and good luck!
Visit https://heyjuno.co to learn more about what Michelle and her team are building!
