An Interview with Dan Haight of Darkhorse Analytics

Founded in 2008, Darkhorse Analytics is a diverse group of scientists and information designers who help businesses make integral, evidence-based decisions.  Their work allows organizations to discover value, engage their stakeholders, and convince their audiences using data. Darkhorse Analytics grew out of founders Dan Haight and Joey Cherdarchuk’s work at the University of Alberta’s Centre for Excellence in Operations.

Darkhorse’s expertise in predictive modeling has been employed in the Canadian Air Transport Security Authority, at the NHL , international construction equipment company Finning, and for the Alberta Gaming and Liquor Commission.

Darkhorse’s president, Dan Haight, shares his experiences in this interview.

Who is Darkhorse Analytics? And what kind of companies do you work with?

Darkhorse is a consulting and software company that helps people make better decisions. We analyze and visualize data in a way that is both clear and engaging. We are industry non-specific, so we work with large organizations, with government departments, with non-profits, and any company that has data.  We help reduce the uncertainty in decision-making.

Can you give us some examples of some the bigger companies you’ve worked with or some interesting problems you’ve worked on?

We worked with the Edmonton Oilers for several years to develop some analytical capabilities in-house. We did a kind of Moneyball-style planning, player acquisition, and drafting that was a very interesting, fun, and rewarding, yet also a complex project to work on. On the other end of the spectrum, we’ve worked with Procter and Gamble helping them do some data visualization work to impress their customers. So, as I said, we do very non-sector specific work.  We look for opportunities where the work is both interesting and has the potential to move the dial.

Do most people know what their problems are? Or is this something that they’re trying to get you to help them figure out in order to optimize their situation?

There’s really two sides, some people have a very specific question in mind such as “We want to forecast the future number of calls, or amount of money, or tax revenues that we are going to bring in based on what’s going on in our province or our company,” or “We want to improve our transportation logistics by optimizing our routes, we know there are some improvements we can make there”.  Both of these have a clear direction. But then we also get the other side of the coin which is more, “We know that moving to an evidence-based approach will give us better results, but we don’t know how to get there and we don’t know where to start.” We help both of these types – the experienced and the beginners – to unlock the value in their data.

How did the company start?

We were founded in 2008. My partner Joey and I, we were both working at a research center at the U of A, had been doing similar work but in a very academic environment where timelines were not that important but the level of precision was very important. We had been doing a lot of projects in emergency services (ambulance and fire). It was actually one of the Ambulance Chiefs we worked with who said, “You know, what you’ve developed here is really, really useful and really powerful. It shouldn’t just stay in the research realm; it should get out there”. And that was kind of the kick in the pants we needed to get started. We founded Darkhorse and began pursuing customers. We were really driven by what we found interesting, where we thought we could move the dial, and where people would pay us.

What have been some of your biggest wins?

The Edmonton Oilers project was a big win for our company. I’d actually set that as a personal goal. I had read an article in Sports Illustrated while waiting in a hospital room at 2:00 in the morning with a broken ankle. The article was about the use of analytics in basketball and baseball, and I thought, “Man, there’s got to be a way to do this in hockey.” And within five years from saying, “Hey, that would be cool to do!” we were actually at the draft helping the scouts and the General Manager refine their picks and their list.

We’ve also had several projects that we can’t really talk about, where the impact we’ve made is large and we’re quite proud of that (in the millions of dollars) and so that’s always rewarding, but what’s really important to us is when we can move the dial, where it actually influences decisions as opposed to just creating a report that sits on somebody’s desk somewhere.

We particularly like the work we’ve done in ambulance and health care. We tell our analytics staff, “If you make a mistake, people could die. So be really careful.” But kidding aside, it’s been great to hear back from some of them: “Hey, what you predicted five years ago, you were within one percent now that we’ve actually tested it,” or “We would never have hit our response targets as an ambulance service until we implemented the recommendations you gave us.” It really all comes down to when we can make an impact on projects.

Can you recall some of the scariest times you’ve had?

We’ve had a couple of periods in the company where projects didn’t come through on time, or didn’t come through at all, or fell through at the last moment, and where we were depending on them from a revenue point of view, that that put us in a financial bind. We had one where it got so bad that my partner was like, “You know, I’m just not sure that I can keep doing this all the time. What are we going to do?” We were at the point of deciding, “Do we keep going?” This was right before Christmas, and  then, on a Friday afternoon before the holidays, we got a call from a potential client, and he said, “You know what, I want to hire you guys.” It was the largest project by four times what we’d ever had before, and it was a done deal. It kind of resurrected the company right from the (potential) ashes at that moment. But it also ended up being a bit of an impetus because we changed what Joey, my partner, was doing.

He had been in charge of company operations and a lot of the project management and planning, and we kind of just moved him out and said, “Do what interests you”, and it turned out that what interested him was data visualization. He began to do research in the area, experimenting, and writing on our blog about that to the point now, six years later, his work is world-renowned and some of the biggest names in data visualization know him. His blog posts, some of them, have been downloaded well over a million times and have actually become part of the curriculum in courses on data visualization. Our company itself has moved to the point where in some years, data visualization is the majority of the work that we do. So although it was a bad time, and it was a stressful time, it ended up being the seed for a whole new line for us, and it’s something that really differentiates us.

What are your biggest challenges today for the company?

Scaling from our current size, where we have about 15 people full time, to that level where not everything flows through the founders, is probably our biggest challenge. Learning how to scale beyond where we currently are.

The secondary challenge is learning, as a company, how to move from being just a services and consulting company, to being a software company – or at least partially a software company. Discovering how the scales, and the marketing, and the operations need to change in response to that. Right now, probably about 10 percent of our market is coming from that, and we’d like to grow that in the future.

The third is protecting the culture and the values that have given us the success that we’ve had to this point.

What are your biggest insights into growing a business?

I won’t claim that it’s the biggest insight, but an insight is that culture is extremely important. If you can create a culture where your team takes initiatives, and where leaders don’t have to come up with all the great ideas, nor make all the right and the best decisions, but where the ideas are driven by a whole group that has skin in the game. That’s a huge part of what we’re trying to do and what has made us successful thus far.

I gave the example of data visualization earlier. We, from the top, had decided on high that data visualization was not going to be in our strategic plan because we’d never made any money doing data visualization –  it was not going to be part of our strategy. But Joey ignored us, and as I mentioned, within four years it was 75% our business. We allow people the freedom to pursue things, and it has paid off handsomely for us.

What are three things you would tell your 20-year-old self?

Number one: Ask for help. There are lots of people, particularly in this community, that are able and willing to help, and actually derive great pleasure from helping out. We operated in a bit of isolation early on, not knowing that there was a community and that we could benefit from that community.

Secondly, when there isn’t a community, there is the opportunity to build the community and do things that connect people. We have a “Lunch-A-Lytics” group that’s building the analytics community in Edmonton, and we have a number of initiatives that we do that just attempt to give back to either students or new grads to help provide them with the things that we wish we’d had at the time.

Another is to take risks early and often. You never know what serendipitous meeting will result in a complete change and completely new opportunity that you didn’t see. It’s very easy to get into the mode of “I have a plan. I got to stick to the plan,” but in reality, your life really comes down to how you choose between the serendipitous and the planned, in the moment when the serendipitous comes along. Which route of trails do you follow that will turn into the yellow brick road? And which ones are dead ends? You really don’t know, and those are the most important decisions you’ll make, but the more opportunities for serendipity that you create, the better the chances are that you’re going to hit a homerun.