Domain, the property brand of Fairfax Media, is seen as the growth engine for the media organisation, figures for 2015 include 37% digital advertising revenue growth and 38% growth of revenue from Domain.com.au. In this interview with Domain CTO Mark Cohen, we explore how he addresses the challenges in and around successfully introducing strategic innovation within the organisation and the insights he can share with other business leaders.
HI: Fairfax/Domain have a history of collaborating with universities to innovate, what were the outcomes and key learnings gained through these collaborations?
MC: For more than 5 years Fairfax (via various groups including Domain, Drive, RSVP) collaborated on a range of projects with a number of leading universities via the Smart Services CRC. [The CRC – Cooperative Research Centre – Programme, established in 1990, supports collaborations between industry, researchers and the community]
The CRC model can be a good way to share the cost and reduce the risk of innovation; but it also poses a number of challenges, primarily cultural and time based.
Our RSVP project, the reciprocal recommender that improved our ability to match people to people made our whole investment in the CRC break even. However, the lessons learnt probably proved most valuable as they highlighted the need to involve all the stakeholders across the business. This should be done early in any projects and should include everyone from the development team (who will eventually implement changes into the product) to the senior leadership (who need to be reminded of earlier agreed expectations to avoid accidental or arbitrary shutting down of promising projects).
Another major lesson was to make it someone’s job to ensure new technology succeeds; in this case extracting value through implementation of the IP created. It’s equally important to make sure that this person is supported across all relevant parts of the organisation – if that doesn’t happen, you are setting them up to fail.
Implementing innovation is not something you can contain. It impacts and permeates the whole organisation, which should be a core objective when looking to maximise return.
HI: Now that the CRC funding rounds have opened and a condition of funding is they must be industry led, what advice would you give to your peers in industry considering investment in a CRC?
- R&D is experimental by nature; out of ten projects expect one to be a hit, and three more to break even.
- Don’t silo the CRC work away from the rest of your innovation programs; involve everyone across the organisation, be inclusive and collaborative.
- Resource the partnership to succeed; make this a shared mission across the business, not the remit of a single poor soul.
- Manage expectations at the senior level; keep reminding them this is not like funding the purchase of an existing, proven piece of software.
- Remember to value the experiment even if the outcome is a no go; the cost to learn this and avoid mistakes is significantly less via a CRC than going it alone.
Innovation projects by nature are not predictable and so convincing a CFO to fund/resource them can be difficult. Though, we’ve found a way around that.
HI: Funding innovation projects can certainly make a CFO nervous. How has this changed at Domain?
MC: Domain has grown significantly over the last two years and we now have one of the best development teams in corporate Australia. Our primary focus is product innovation and we resource that internally.
We don’t produce copious amounts of business case documents selling the virtues of innovation and preparing to run a gauntlet involving many levels of approval. Our approach, ratified by our CFO and CEO, is to put forward a budget for resources at the beginning of the year and then focus those resources around projects that improve domain.com.au or our other brands and deliver the greatest return to the business.
As long as we perform we can negotiate more resources year on year. The best bit is we can just get on with the job, no distractions.
HI: It sounds like you have a forward thinking CEO. How did he convince the Fairfax Board to support this model?
MC: Having a CEO with a proven track record in entrepreneurship and driving innovation and who has ‘skin in the game’ is a huge advantage. He has allowed us to reshape the way we think and work at Domain. Our culture of innovation is palpable; everyone is encouraged and motivated to keep thinking of new ways to improve the product and the user experience.
HI: So innovation is now firmly on the agenda. What was involved in the journey to get it there?
MC: Before I took on my role as GM of Product Technology for Fairfax I had the Head of Technology role in the Marketplaces division. This entailed looking after classifieds, person-to-person, and business-to-person, as opposed to media.
In that role, when we workshopped the values of the business, I actually spoke against listing innovation as a value. The reason I stated was, “You’re including it because it’s something aspirational, not something we actually do.” We bashed it backwards and forwards and, eventually, the CEO said he wanted it as a value, because it’s what he thought we should be doing. Getting CEO support to draw up the innovation agenda was the pivotal point for me.
I also explained to people that there is a difference between entrepreneurialism and innovation. When people said, “We want to be innovative” I’d say, “Okay; this is what you have to do to put that value into action.” That’s when we began to get traction.
We started off really simply. We did hackathons, which we called Innovation Days, that were very focused on tech. Initially we used to battle to even get product managers involved.
A year later I reported to a CIO who absolutely understood the value of innovation as something you do in practice. Then a new CFO arrived and he is the least CFOish CFO that I have ever met. He drove the design thinking agenda for the whole company, even sending people on training courses for design thinking. We once had a chat and he said, “What do you need to make this work?” I said I need resources, and his response was, “Ok, don’t talk about it; you’ve got what you need, now just do it.” Because of the executive level support, it was really easy for me to recruit people into the program.
HI: You had budget for resources up front, support from the executive team, a mandate to innovate, training in design thinking and a workforce keen to participate. What did it all add up to?
MC: We took a lot of ideas, stretched the program across the entirety of Fairfax, developed fast and tested quickly so we could start commercialising ideas. We didn’t create the next BuzzFeed, but we created a lot of good things.
Because of the executive level support, every couple of months – at the senior management meetings – we would stand up and say, “This is what we did and this is why we did it.” We’d talk about our successes, but we’d also have a laugh at the things that we had failed with.
It was an agenda item that people, including the CEO of Fairfax and the CEO of Domain, would speak about: we want to be innovative, and we want to support innovation. How do we do more of it? How do we do it better?
HI: Speaking of how, how are you addressing the challenge of capital cost?
MC: Everything is always a challenge in terms of cost, so having support is key.
People in a business unit can say, “I don’t have to get my boss’ permission to participate. I’m just going to reply with my ideas, and then actually join in.” But, you know, collecting ideas is easy; that’s not innovation. Everyone has ideas. The innovation is what you do with them.
If you build something and stick it in front of people, that’s an experiment. Innovation is invention taken to market – and that last step is the challenge. You need support to drive that agenda.
HI: Can you share with us a favourite quote?
MC: Sure, my favourite quote on innovation is from Joichi Ito, the Director of the MIT Media Lab and Chairman of the Board of PureTech Health. He is also on the board of The Sony Corporation, The New York Times Company, The MIT Press, and a bunch of others.
He said, “If you want to increase innovation in a company, you have to lower the cost of failure.”
The reason that holds true is because there’s no such thing as a failure tolerant culture. We always say, “You have to create a failure tolerant culture,” but that’s nonsense.
The failure tolerance comes from the cost of the failure being low enough to dismiss it as an experiment, as opposed to a failure. The way you do that is you don’t buy infrastructure and you don’t buy software. You do MVPs (Minimum Viable Products); you build the smallest possible product you can.
HI: Can you give us an example of how you reduced the cost of failure at Domain?
MC: One example with is a Domain app I was using.
I just got a new phone. So now I’m in the ‘dev heaven’ of Android versus iPhone with a far less restrictive app approval process. The dev guys can release Android features four or five times a week with no issue, and they regularly use feature switching to turn features on or off for some people in the live environment.
Using this as a target sample for new features on the phone we’d add a feature and show a little splash page [in the app] effectively saying “do you want to use this feature?” If you click the button. you get another screen saying “Hey, this is coming soon; click here if you think it’s something you would use.” Click yes and the screen says, “Thank you for making Domain better,” and then it disappears.
It’s a really quick, low-cost way of market testing improvements. It takes, maybe three hours for a designer to put the pages and the little splash screens together. Maybe an hour for the tech guys to work in a feature-switch (like a toggle to turn it on) – and it’s in the next release. When they’ve got enough data, they turn the feature off so, even if you don’t update the app, it disappears.
As far as the cost of failure, if no one says, “Yes, I would use the feature…”
HI: Nothing’s broken – or, effectively, it’s like breaking a nail instead of a bone.
MC: Right! You’ve written off four hours as opposed to writing off a rack of infrastructure and six months of a team. So the cost of failure disappears. It’s now an experiment.
HI: Going back to the issue of financing innovation, we touched on a scenario unfamiliar to most, which does away with the process of writing a business case and going through layers of approval before starting a project. We’ve agreed that this kills innovation, so can you explain further how this works at Domain?
MC: The short answer is to imagine reverse pitching. Rather than pitching for funding approval upfront we build, test, evaluate and then pitch the outcome.
The finance guys came up with the process; the leads of each team – the dev, ops and product managers – run a one day presentation with finance, who get some documentation in advance. Each team has 30 mins to talk about the essence of their business case – what they’ve done and how much of it was Capex vs Opex. Their outcome determines how much leeway and resources they get to play with next time, in advance.
In other words, here’s a budget, go build, prove the value and then we will give you more. As long as you perform you are free to innovate.
HI: Is there an issue when recruiting skilled resources to keep this rolling?
MC: We approve positions and how much we’re paying at the beginning of the year, which means we have ‘X’ number of people in tech. Additionally, we have a particular target we’re expected to hit re Capex versus Opex. So, gone is the habit of having to do a project initiation form, which then goes to a governance committee that meets fortnightly, which then gives you an approval to do business.
That is how you kill innovation, by making it jump through all those hoops. Actually, if you get out of the way, then people will do stuff.
At a recent conference the CTO of Netflix was was talking about their talent. Basically he was showing off, when an audience member from one of the Fortune 100 stood up and said, “The stuff you’re doing is very impressive, but we don’t have the engineers that you have.”
The CTO’s off the cuff reply was, “We poached them from you. We just hired them and then got out of the way.”
That’s exactly like us. We have an amazing team with incredible talent and our CEO has skin in the game. He is personally invested, so he fights the fight for us. He is not technically inclined, but he understands Domain is a technology business and he understands the value of product and techonomics. Translation at a board level is not an issue.
We put forward out budgets and asked for 35% more people this year than we had last year in tech and product. We met with the leadership team, ran through things and, when the numbers went up, I held my breath. The CEO only had one question; he said, “Is this going to be enough? Because I don’t want to ask twice.” It was very, very different to the conversations I used to have.
HI: How has all this changed the way your team works?
MC: The language has changed. Now, every time a team does its planning the product owner in that development team asks, “In the next fortnight how are we going to deliver business value?” instead of, “Where do we want to go?” Everything is framed through the matrix of: does that add value to the business? Is there something on our list that would add more value?
Keeping the lights on adds a lot of value. Because we’ve got a technically savvy product team, we don’t need to do maintenance sprints. Our product guys look at it and say, “If we don’t do that, then my features aren’t going to work, and the sites is going to go down, and I’m going to get caned in the management reviews.” Because they understand this, things get done.
To me, having a cohesive team facilitates the ability to think in this way. The product guys are focused on business value – not only in dollars but also audience [capture] – so they get a suite of whatever creates business value.
HI: Undoubtable is that greater audience capture assists other parts of the business; is this hard to track?
MC: It’s actually not that hard.
For instance: if we know that you came to Domain via a Home Alert Email and we know that you created four inquiries this month, we know what your value is to us as a Home Alert subscriber. This means we can then say that the people receiving those [Home Alert] emails are worth ‘X’ and, if we can double that, they’re worth 2’X’. Therefore, if we want to build something that’s worth half ‘X’ we can determine what we need to do.
If you have enough data, you can create a way to put a dollar value on most things.
HI: What’s your plan to extract value from all the data you collect?
MC: We have a fascinating guy here who (I think) lectured at UTS. Yes, he’s a true data nerd; but when we have our team showcases every fortnight we typically make him go last because everyone wants to hear what he’s got to say. When people are getting tired, you put him in and they snap to attention.
We’re building a small data tech team with him who are aggregating data from everything we can get. They put the data into AWS [Amazon Web Services] and then use a visualisation tool over the top. We started off with 25 licenses and we’re always facing demand to increase that.
As gatekeepers, we create data sets that are very easy to interrogate, so you don’t have to figure out how to join things. You just connect to the inquiries set or the listings set and the data you want should be there.
Beyond that, just have fun.
HI: Is there anything we haven’t touched on that you think is critical to helping your peers to change the way their board views innovation as a risk?
MC: Thinking back to my old team, I’d say this: when you’re firmly stuck in the old world, with a high churn rate and you’re trying to drive a culture of change, one of the best levers you can pull is consciously targeting the right culture in your recruitment. If you don’t, then you hire people who are a cultural fit for where you are now, not where you want to be. In that scenario you end up trying to change yourself as well as the people you just hired.
So, step one is targeting people who’ve shown that they can perform in the culture and the environment that you’re trying to create – your aspirational culture versus your current culture. Once that happens they’ll cause conflict. There might be polite disagreements or a blowup; but, what you do with that conflict determines whether you will be successful.
To drive change you have to back the new person and say to everyone else, “You need to listen to them; if you don’t, then you might not be in the right place.” If you don’t support the new person, they’ll leave and you will be stuck where you were.