Matt Celuszak is a remarkable guy. At age 33, he is one of the world’s foremost experts on mapping the landscape of human emotions and making it accessible to machines.
The way he runs his company CrowdEmotion is no less striking. He introduced a measure that would be unthinkable for almost every other CEO: flat, transparent salaries for everyone.
We sat down with him to understand the why and how.
Matt Celuszak has long been fascinated by measuring the human condition. Growing up an outdoors guy in Victoria, a beautiful town on the Canadian west coast, he first channeled his interest into studying human kinetics - or the science of body movement - at the University of British Columbia.
But soon he moved on, from the body to the mind, trying to, as he puts it, “understand what drives people to ultimately fulfil their own lives”. After building a tool to measure the effectiveness of sport sponsorships landed him a job at IMG, a spell at a Canadian market research company cemented his passion for measuring human sentiment. There was a problem with how it was being done: “I found that what we use for human measurement is still very human dependent. And humans aren’t very good at quantifying the qualitative”, says Matt.
Fast-forward to 2017, and he has built CrowdEmotion to track all the individual components of the human emotional footprint. The company sells emotional analysis and interpretation software to clients like the BBC, CNN, and Omnicom, straddling the fields of neuroscience, psychology, and artificial intelligence. Fully bootstrapped, CrowdEmotion employs 9 people, and is about to reach breakeven.
Impressive as this achievement is, we were most curious about a groundbreaking policy Matt implemented at CrowdEmotion: flat, transparent salaries. A 40-year-old Director of Artificial Intelligence earns the same as a salesperson, age 22, who six months ago was studying for her final university exam.
It seems an extension of the more popularised trend of openly publishing salaries. Buffer have their elaborate formula that emphasises their objective criteria for determining who gets paid what, and Seattle’s Gravity Payments aim to reduce inequality by setting a minimum wage of $70,000 for every employee.
But Matt didn’t start CrowdEmotion planning to implement a flat salary structure. Instead, the policy was an outcome of his pragmatic, goal-oriented questions: What are the needs of this organisation? And based on these needs, what types of decisions will we have to make?
Matt’s reasoning for flat salaries goes like this...
- CrowdEmotion is unique in two ways:
It’s venturing into unchartered territory. Mapping the emotional landscape is a pioneering endeavour.
Analysing emotions is a horizontal technology, meaning that it can be applied whenever a human interacts with a machine. Its potential use cases range from crime prevention to the effectiveness of online advertising.
In such a scenario, you don’t know where the killer commercial application for this product will come from, and because you are combining so many different fields (neuroscience, psychology, artificial intelligence, business), there’s no one who’s “been there, done that”. This means there’s no reason to pay a premium for experience.
Therefore, it’s essential that the company...
foster a diversity of perspectives. This doesn’t just mean race and gender. But rather different ways of thinking and seeing the world. In CrowdEmotion, there’s a friendly rivalry between a Trump fan and an Obama fan - Matt says this is a very good thing.
give everyone an equal seat at the table. Great ideas can come from anyone. You have to remove any barriers that keep people from speaking up. Salaries create the perceived notion that those who earn more warrant greater attention.
motivate good decision-making. In a startup that depends on the whole team, where literally anyone could come up with the idea that will lead the company to huge success, you need to tie individual success to team success to company success.
Alternatively, if you knew exactly where the killer application will be, and you were racing against competitors to get there first, you’d go very vertical and hierarchical: CEO says XYZ, go and execute. But in this case, it’s not a speed challenge - it’s a discovery challenge, so you benefit from equal salaries immediately putting everyone on a level playing field. As a result, everyone feels entitled to be candid, question everything, and not mistake a fancy title or a high salary as indicators for more wisdom.
Put differently, Matt’s reason for rejecting salary differences is that by assigning diverging monetary values to people, you give them different amounts of power relative to each other. This will, at least subconsciously, bias their interactions. Lower-paid people will defer to those with more money. As a result, you will get decisions slanted towards the views of the highest-paid team members. And that’s a bad outcome because it’s not the best argument that wins the day.
So how does it actually work? It’s as simple as it sounds: CrowdEmotion pays everyone the same base salary. Employees can sacrifice salary in favour of stock options. This fact means that the effective salary does vary at the end of the year.
What were the results of this revolutionary new approach? “I get challenged a lot”, Matt says emphatically. And while there is no counterfactual, Matt believes that the flat salaries have heavily contributed to this level of openness and willingness to argue against the CEO, or any other older, more experienced person. Matt believes that flat salaries offer the best chance to surface the best data - because it minimises the bias in favour of or against the messenger. Politics in CrowdEmotion are very rare. Flat salaries remove many incentives to become political.
What makes Matt think that this model has been successful? “We’re winning”, he says tersely. “We’re breaking even soon, and we’re winning more and better clients than our competitors.”
But, of course, the long-term effects remain to be seen. CrowdEmotion is still a small pre-profit startup. Matt admits that only half of his board of directors agree with him - the other half just play along.
So what’s the actual number? What’s the CrowdEmotion salary? Matt laughs “I won’t tell you that. That’s just for internal consumption.” Matt believes that the practice of some startups to publish their salary data is silly: “It doesn’t really add any value and it harms your company by telling your competitors how to poach your staff.”
But doesn’t the difference in experience between a senior person and a graduate mean that the former’s opinions and decisions will, on average, be of higher value? Is there really no value in experience any more?
Matt takes the view that since CrowdEmotion is operating in such a groundbreaking technological environment, “experience” doesn’t matter nearly as much as conventional wisdom might make you think. Overall, the business world is still struggling to assign value to experience. We cannot simply default to the assumption that having racked up years of work automatically results in a commensurate increase in value to the employer, Matt says.
He also believes that experience comes with baggage. Good inventions come from gaps in knowledge, and good solutions come from looking at the problem with fresh eyes.
Once flat salaries, always flat salaries? “No,” Matt says quickly, “before our business makes money, everybody is of equal value. They have to be! Because it doesn’t matter if you have twenty years of experience. You can build the best AI system on the planet, but if you’re not listening to the salesperson’s problems, we’re never going to make money. So I don’t care. It’s not of value.”
Once you reach sustainable profitability, however, you are increasingly better able to distinguish individual contributor’s value and can start paying differently. Also, you have reached a point where you can afford to buy in talent to scale your operations.
But even then, Matt believes, experience is still not a good yardstick for differentiating salaries. Experience means you’re looking backwards and not forwards. Instead, salaries should be differentiated in contribution value and by the amount of business risk the person is sharing.
CrowdEmotion will remain fiercely meritocratic, Matt insists, while offering a degree of differentiation: “We’ll move to the flat tiers and a contribution/risk variable, but not much more. The principle is important that people are pushed to help each other achieve. We will still be discovering so long as humans and machines are interacting, and therefore we need to keep some element of equality and transparency at the fore”, he says, adding: “the Diamond Shreddie didn't come from the CEO of Nestle, it was an intern at the agency.”
But surely there must be downsides? “Yeah”, Matt bursts out without hesitation, “cashflow management sucks!”, he says, laughing bitterly. There’s also the pain watching some older team members struggle with family costs, while younger team members enjoy a significant premium over their peers.
Another challenge comes from an unexpected place: Matt is worried that some of his younger team members are exhausting themselves too much. The relatively high salary they receive places a burden on them, making them work too hard. “I have to keep reminding them that you’re allowed to fail, too.”, says Matt.
But at the end of the day, there must be something profoundly good about the CrowdEmotion company culture. When Matt speaks about his team, his eyes get misty: “At the last company party, I was brought to tears. My team is just amazing. They’re hearty, they’re smart… they get it. They get it better than I do.”
So - if it works for Matt - should you do it too? It seems that companies who wonder if they should try this approach ought to ask themselves the following questions:
Are you breaking new technological ground and don’t know yet what the killer application will be?
Are you in exploration mode? If you’re past the exploration stage and just want to execute, then you need an efficient hierarchy. And a hierarchy is unlikely to work without salary differences.
Are you in it for the long run? If you want to get acquired as fast as possible, then speed matters. You’ll want to raise a ton of money, hire the most experienced people (so buy in talent with high salaries), and run as fast as you can. This won’t work with relatively low, flat salaries.
Does this fit with your personality and the company’s DNA? Matt openly admits that he wouldn’t be a good CEO for a company that replaces existing use cases or existing technologies. He’s the right leader for a groundbreaking new technology solution. Equally, he doesn’t revel in command and control-style leadership. He’s happy taking on the role of the facilitator, which is what an early stage technology business needs: ensuring that ideas get surfaced, prioritised correctly, and acted upon systematically. This even means that he is fine with some of his decisions being challenged to the point of being effectively vetoed by anyone on the team.
If the answer to all of the above questions is “yes”, then you’re cut out to follow in Matt’s footsteps. His line of reasoning is sound: no salary model is perfect, and even his comes with downsides. But arguably, for a trailblazing company like CrowdEmotion, it sounds like the right solution.
A conversation with Matt feels like a full reframe on the subject of how to reward and motivate team members. In true Elon Musk fashion, he has gone back to first principles and found a powerful contrarian answer that, so far, seems to be working out for his business.
That is a real achievement - not less impressive than taking the first steps towards mapping the human emotional landscape.