by Divina Paredes

CIO50 2019 #18: Angie Judge, Dexibit

Mar 28, 2019
Business ContinuityCloud ComputingInnovation

“The biggest thing I’ve learned in my career is just how important it is to focus technology development on solving a problem and making sure that it’s an important, painful problem that’s mission critical to your customer,” says Dexibit CEO Angie Judge.

“It sounds simple, but it’s easy and very tempting not to and instead get carried away by the technology, a mistake so many people make,” says Judge, whose company has been signing up an elite customer base across North America, United Kingdom, Europe and Asia Pacific.

Dexibit’s ability to predict and analyse visitor behaviour powered by artificial intelligence is a unique, patent pending innovation in the visitor attractions sector.

Over the past year alone, its awards included the PWC Hi-Tech Awards for Most Innovative Software and Best Technology Solution for the Creative Sector and Spark TIN rising star for early stage companies.

Judge was also named the 2018 Westpac Woman of Influence for Business Enterprise.

Judge says she always wanted to use her experience in location analytics to build a company in the museum space.

“I originally thought it would be fun to use that to disrupt the visitor experience – after all, so many people visit museums and the like with their heads buried in their mobile phones,” she states.

“But it turns out, that’s not really an issue. As the dozens of companies who have attempted to create technology solutions with apps, VR/AR and bots around this have found through uptake failure, it’s a hard space to win, because there’s not a strong problem to start with.

“And yet, in a world where startups are funded off passion and measure their successes in social media reach, it’s easy to avoid facing reality until it’s too late.”

She avoided that fate when she pivoted her focus on the critical problem museums face: These venue did not know much about their visitors, hence, it was hard for them to generate useful insight.

“As a result, they mostly just guessed decisions instead, which was often exceptionally painful, like when an exhibition flops because its popularity has been overestimated or mistimed,” says Judge.

“The moment crystallised for me when I saw the security guards at the Smithsonian manually clicker counting every one of their 30 million visitors through the door.

“I could imagine the process of getting that data to a clipboard, into a spreadsheet, some poor analyst pouring over the numbers trying to take meaning, the manually distributed reports, a guess at what the future held – the sheer number of opportunities for human error, the lack of insight into what those visitors were doing once they were through that door. The ramifications of getting it wrong. That was the moment I realised I’d found a problem worth solving. “

It’s a lesson I’ve tried to remember constantly as Dexibit has grown, she says.

“Before we leap into trying anything new, we must be disciplined enough to come back to that question ‘are we solving a problem?’ and ‘is this a problem worth solving?’.

“Where we’ve gone down avenues in our product development that we’ve ultimately pivoted from, it’s usually because we haven’t answered this challenge well enough,” she admits.

“Asking this question, having the guts to turn away when the answer is no – and then if yes, learning as much as we can about those problems, becoming experts in them, helping our clients discover value in solving them – this is what ultimately brings customer centricity to our innovation.”

Innovation mindset

Continuous innovation is another lesson Judge can share based on her experience growing Dexibit from startup to a global software company.

Dexibit introduced artificial intelligence, in the form of machine learning and natural language, to aid with predicting and analysing visitor behaviour as part of its software as a service offering of big data analytics for visitor attractions such as the Smithsonian.

The use of artificial intelligence has powered much of the company’s international growth, establishing it as a market leader in the broader attractions category – from its initial base in cultural institutions such as museums, galleries and libraries through to a more recent expansion to the commercial attraction segment such as theme parks, stadiums and zoos.

Dexibit launched a new predictive analytics module with a number of forecasting models which use a variety of Dexibit’s machine learning algorithms to predict elements such as visitation and revenue – over a year ahead and down to the hour, as well as simulate new market offerings such as exhibitions to aid strategic and operational decisions which grow visitation, revenue and efficiency.

The company has also developed an insights module with several advanced analytics models that offer location analytics, sentiment analysis and digital correlation to help understand the visitor’s journey, happiness and market influences.

As Dexibit’s traditionally descriptive analytics took hold in the industry, the tech company naturally progressed into more predictive and then prescriptive analytics – inspired by its customers to disrupt their traditionally manual approach to strategic and operational planning, and visitor evaluation research, turning laborious planning and limited survey samples into data informed decisions supported by in the moment insight.

Dexibit worked with Callaghan Innovation and the University of Auckland’s data science programme for a research and development programme that will speed up its time to market.

The programme has three pillars:

The first is applying customer centric design principles. This means pairing a lead client for each research project as the focal point for initial analysis, model development and user experience prototyping, followed by beta testing across Dexibit’s client portfolio.

Second is the establishment of a machine learning as a service infrastructure to ensure the data team can implement and scale new models quickly, reusing a shared component foundation.

The third is a commitment to continuous improvement. “We are constantly tweaking enhancements to applicability, accuracy and insight,” she states.

These organisations were used to forecasting models that take months of involved analysis.

Dexibit’s technology has also paved the way for many of these traditional organisations to switch to dynamic business models, especially around ticket pricing and capacity management.

The resulting benefits to Dexibit’s customers include increased visitation, revenue and efficiency, alongside the achievement of social outcomes.

To enable their work on AI, Judge says they have product and data teams in their core engineering function. They then built a dual data science capability across product research, focusing on exploring customer facing models and data engineering, and building the underlying data infrastructure.

Operationally, the introduction of forecasting required a high degree of data integrity and business rule configuration across a variety of diverse venues. To handle this complexity, Judge says the team developed a series of configurable, interdependent feature sets to provide a strong and reusable data foundation across the various models.

Culturally, the company was faced with introducing an entire industry to the topic of artificial intelligence, she says.

Dexibit worked with industry executives, and provided a masterclass series to assist the latter’s business teams with understanding forecasting components such as contextual enrichment, featured factors, residual attribution.

“We are helping the industry to embrace this new technology through seeking to understand it and manage organisational change considerations such as ethics, rather than treating machine learning as a ‘black box’ solution,” says Judge.

As Dexibit has scaled globally and particularly set against the complexities of delivering for government clients, Judge has leaned on partners such as Callaghan Innovation, through its Build for Speed program which audits against industry best practices such as agile development and continuous delivery.

She says they also got assistance from Amazon Web Services in architecting their growing solution with new technologies.

Dexibit’s high performance team has been built on a foundation of integrity, excellence and diversity which starts with understanding its people as individuals, says Judge.

She says the company boasts an exceptionally diverse team across gender, ethnicity, orientation and belief and remains committed to the importance of this in its success.

“Unusually for a technology company and particularly in the data science field, Dexibit’s team is comprised of an equal gender balance, including at board and leadership levels,” she says.

Judge advocates for personal growth and mentoring of individual staff. As well as internal training and brown bag collaboration, the company provides opportunities for individuals to attend conferences and training aligned with each individuals’ development plans to support internal promotion or skill diversification.

Judge is also big on celebrating success, implementing a ‘plus one’ policy. This means the company’s leadership team always takes along another team member to attend them in external events.

More recently, Judge says Dexibit has also taken careful consideration of mental health.

“We have important rituals such as a weekly ‘peak and pit’ to encourage the team to support each other in alleviating the stress inevitable with a fast paced, high growth environment.”