By interview @SimonCocking. Great interview with Finn MacLeod from ZenLikeFocus

How did you get into Data Visualisation and Analytics?

I started with a PhD with Professor John Lewis, in an area called Large Deviation Theory – a specialisation of probability which deals with the likelihood of rare events. John had recently been involved with founding a company called, which had started applying this kind of mathematics to network monitoring issues.

Then I finished in UCC, and did a research fellowship with Professor James Gleeson in the University of Limerick. he had just started doing research into building models of virality on the Twitter network. So I guess the seeds were sown early.

Why didn’t you continue in Academia?

At the time I was utterly fed up with the analytics side of things – and literally ran away to comedy and the circus. I got a once off gig as a Circus Clown which was paying twice the hourly rate of a research mathematician, and I thought “fuck that”!

I had a yearly gig to go to Venice and teach a mix of circus and English to kids, just around the time of the film festival and the Venice Bienalle. It was great fun.

The worst pitch I ever did…I won’t say which startup accelerator…but they opened with the question “what did you do before Beautiful Data?” and I said “Circus Clown” — and it basically spiralled downhill from there. They had no sense of humour.


So how did you get your first clients?

We won an Enterprise Ireland CSF award to visualise mathematics in new ways. The process was really effortless. But it didn’t work out commercially, and we moved into dashboards. Which was the start of the data visualisation business, Beautiful Data.

We then landed a gig for Thomson Reuters to analyse key influencers at the Dublin Web Summit, and this took us further into the world of social data.

Why do you call data visualisation “CEO porn”?

Because its universally desired by CEO’s. In theory it should be able to condense all the information about their company into nice easy to use dashboards. In practice what they want is not necessarily good for them. There can be all kinds of problems, underlying infrastructure issues, and the difficulty of representing the CEO’s thought model of the business.

What’s worst thing about startup life?

Running out of money. It sucks. What they don’t tell you if you start a startup is that it is very, very hard to get social welfare if you run out.

There was a moment that I had 30 euros in my account, and a ticket to fly back home to Scotland. But it was Christmas, and the last bus from Cork to Dublin was full. So I got onto the bus and pitched everyone there, and no one budged. After about five minutes this really nice Polish guy came off and offered me his seat. Fair play to him.

That’s why one should build sales and marketing mechanisms before your product. If you have them, then you can always sell other peoples stuff until you get yours rolling. The ZenLikeFocus project came out of that philosophy.

What is

ZenLikeFocus is a method of growing your following on Twitter in a really targeted way. There are a number of different strategies for growth in social media and many of them are monotonous and haphazard. We use smart analytics to keep track of all the details and make the best choices. A company still needs to say interesting things online, but we optimise the strategy. This makes a difference to growth and sales.

In some ways we do “SEO for Twitter”. We all know that sites on the frontpage of Google get lots of hits, some if they’re on the second page, and almost nothing after. Twitter is exactly the same. When you get a big enough Twitter influence, you start surfacing in search results a lot more. And Twitter is a huge search engine, bigger than both Yahoo and Bing.

The offering as a service is really new. We soft launched just before Christmas.

Can’t you just buy followers on a site for a fiver?

Yes but they’re not real. The whole point is to find customers that are willing to buy your product, or followers who are willing to engage with you. We’re decreasing spam and fake accounts by targeting better. Marketing managers don’t have time or knowledge to perform the number of checks that we do, so they’re more likely to randomly follow people.

For example, when we select an account to follow, we might check the network structure of an account’s followers, or verify against data sets from outside of Twitter. These checks are closer to what is used in fraud prevention, rather than just looking at an account and guessing whether you think it’s relevant. Your numbers won’t jump overnight, but they will grow consistently with the right kind of people.

What is it that makes Twitter information valuable?

Social network data can be used in many interesting ways – Techcrunch was probably one of the earliest to the table to monitor niche activity groups on Twitter for spikes in activity. If they saw a spike, they’d get a journalist on the phone and get the latest scoop before anyone else. Some of these ideas spun out into Marshall Kirkpatrick’s company Littlebird, which is actually quite similar to what we’re doing. If you remember the time the White House Twitter account got hacked – and a fake tweet claiming that Barack Obama was injured knocked millions off the Dow Jones – you know that this data has real influence.

Twitter has been struggling recently – does that concern you?

Not at all, that seems more market driven than anything else. Twitter is live, kicking and important for two reasons. Firstly it’s where the journalists hang out, so if you want press, it’s a great way to find them, especially niche bloggers. If these guys like you on Twitter, they spread you through their other channels.

Secondly, Twitter is great technically, and it makes data available in a really generous manner to third parties, unlike LinkedIn say. This means you can do good data science. When the *** goes down, like in Egypt, Tunisia or the London riots people still turn to Twitter. A while back, we did some proposals for a company called Newsflare, who are a competitor to Storyful, based in the UK. They find videos that are on the edge of going viral, reach out to the owners and buy a licence, then sell that content to larger distributors, like the BBC. They wanted to automate the process of finding videos, just before they went viral which means you have to poll the YouTube likes or comments to see if they are increasing quickly. However, at the time we were thinking about the problem, YouTube wasn’t updated frequently enough and twitter was a more reliable way of getting “previral” prediction data.

What’s Next

We’ll look at Instagram next, as well as developing better visuals and analytics to show people the types of followers they have and the ecosystem that surrounds their keyword niche.

If you would like to have your company featured in the Irish Tech News Business Showcase, get in contact with us at [email protected] or on Twitter: @SimonCocking

You May Also Like:

Pin It on Pinterest

Share This