It all started with a gamble.

After founding Saberr with a university friend in February 2013, Alistair Shepherd started visiting entrepreneurial competitions, not with the aim of winning but to predict who the winners would be.

saberrheadshot800
Alistair Shepherd © Saberr

“Our first test of the algorithm was at the University of Bristol,” Shepherd wrote in a blog post in July 2015. “They were hosting a business plan competition as part of their Spark course. We wanted to see if we could predict which team would win without any knowledge of their skills, their experience, their demographic nor the idea they were working on.

"We were hoping to be roughly right, roughly the teams with better relationship quality would rank higher than those with lower relationship quality. But instead we got the precise ranking of all eight teams spot on.

“The probability of getting this correct is 1 in 40320.”

How it works

“We started by looking for patters in online dating data,” Shepherd told Techworld, “if you think about a team in classical psychological terms you will look at personality profiling, is someone introverted or extroverted, for example.”

The problem for Shepherd and his team was that patterns only helped them identify compatibility based on shared characteristics or interests, which is unique to every pair. So they started to go deeper by looking at shared values.

“The thing we found interesting was that it’s not just your values that matter, but your tolerance of other people’s values. If you meet someone that is intolerant of other people’s ways of thinking or working, then they are hard to get on with.”

So Shepherd and his business partner started to develop their own algorithm and mapping it to teams of people. The result is a ‘resonance score’ which the company believes is the best possible predictive measure for whether a group of people will work effectively together. Saberr also claims to be able to predict who will be a good fit with your existing team when hiring.

Saberr’s survey concentrates on deep values rather than shared interests. Managers will then receive a breakdown of compatibility scores ranging from 0-100 (100 being the most compatible) for each co-worker.

The startup’s algorithm then determines a compatibility rating by predicting people’s alignment on certain key values. Put simply, if two people value the power of words over numbers they should gain a strong compatibility rating and therefore be teamed together. This can be scaled up across a whole team so that an average shared rating can be seen by managers who can tweak the team makeup accordingly.

Engineering success

Shepherd comes from an engineering background, so his work has always been grounded in absolutes. Can we absolutely predict the success of a team, or an entire organisation, based on their ability to work together?

He frames the idea in the concrete terms of engineering: “Before you build something you do a lot of computational modelling and the majority of your testing on a computer. So when you build it in the real world it behaves as closely to the test as can be expected. We wanted to do this with HR. People are fundamental to business but we are hugely complex in the way we think and the way we work.

"There are people we get on really well with, and when we do our performance and output are really high. The opposite means our output is a lot lower. So my question was: ‘What if there was a way to predict that and model it ahead of time.'

“The implication if you can do that are phenomenal in terms of teams that are designed for high performance. It does good things for the performance and for the team involved, who are going to be a lot happier and have less stresses if they are working with people that they are compatible with."

The name

So, where did the name come from?

“The name comes from two places,” Shepherd explains, “the Spanish word saber, meaning to know, and sabermetrics.” This is the concept of gaining a competitive edge through effective use of data analytics in baseball that Michael Lewis wrote about in Moneyball. “You can draw many parallels between what we’re doing and what sabermetrics did for baseball. It comes down to engineering a better outcome for a team,” Shepherd explains.

Funding

Shepherd thinks that many startups are too hung up on funding. “It can be so consuming for a lot of founders when it shouldn’t be. It’s not the goal, the goal is to have resources and the best way is to sell your product. Of course funding is an objective when you don’t have a product, and we were very lucky to get significant seed funding ($1.2m according to AngelList) on basically a powerpoint and a good idea and a couple of case studies. That gave us the capital to build.”

Read next: 11 ways to fund your startup

So Shepherd is in no hurry to acquire VC funding. “For us it doesn’t make sense to take VC capital,” he says. “There’s lots of interest but we are making the conscious decision not to take it.”

Shepherd isn’t looking to become ‘The Uber of HR’ but rather he wants to carve out a lucrative niche for Saberr. “Ours is not a market that is winner-takes-all, like Uber and Airbnb,” he says. “It makes sense for them to take funding and dominate the marketplace. HR is the biggest expenditure of almost every company on the planet, so it doesn’t require dominance for us to be extremely successful.”

Customers

Naturally Shepherd has been met by a few raised eyebrows along the way, mainly from the older, regulated industries: “It’s not necessarily a reflection of the people, but more the processes they have to work with. For example, the banking industry isn’t particularly willing to introduce new procedures. It’s an old industry and set in their ways, and they have to be careful and secure. The legal profession is similar.”

Shepherd has been surprised by one industry’s response to his product though. “We are seeing good take up in the enterprise space. There is a real desire to change and innovate their practices, and they have the resources to make that change, so selling into this works extremely well for us.” Capco, Deloitte and Ebury are all listed customers of Saberr services.

Naturally the tech sector is interested too: “Unsurprisingly, the tech world is very open to using data about people to make business decisions. Probably because they are pushing that narrative in other areas: that tech and data will change your life,” says Shepherd.

What next?

Saberr is still extremely young, but they are building a steady portfolio of clients and a good amount of referral custom.

One problem they may face is that although their algorithms are clearly very effective they have no data bank of their own, so insights can only be gleaned on a per-customer basis. As opposed to someone like Insidesales, who can apply their billions of data points to a predictive analytics platform for recruitment purposes.

Being the founder of a tech startup is an ongoing struggle for Shepherd though: “This is my first real company, and I’m still very young [27], so the biggest challenge is to learn things you would normally learn, but then also immediately apply that for a whole business that I am solely responsible for. Which means we make mistakes, daily, and that’s incredibly painful.

“Truth be told we have no idea what we’re doing. We know how we want to get there, but we are learning every day.”