Following the success of the ITT webinar, which considered the role of data in exacerbating bias in the travel industry, host & partner TTG Media quizzed Beyond Analysis' Jordan Browne-Moore, Consultant & Bias Expert, on his top tips for considering bias in business.
Some snippets from Jordan's article with TTG are
found below. To access and read the full article visit TTG.com
"Competition, and the focus on the ethics and morals of companies & industries has never been so high, it is now evermore important to address bias and its effect on what we do, how we do it and to whom." Jordan said.
What can go wrong?
Limiting diversity in your customer base, leaves you vulnerable to the economic risks associated with narrow target demographic groups. Bias dampens our perception of our markets needs and thus our ability to innovate. And because everything starts with people, the tools and technology that we use to do business at scale, inherit our biases. Do we know who our customers could be? Have we unwittingly excluded wider groups? The missed opportunity could be substantial, if you shape your business propositions around a subset of what your market could be.
The effect on the brand
In today's landscape, biased businesses inevitably result in a weakness in the brand image, and a negative effect on potential and existing clients trust, leading to a halo effect on the entire industry. Are we as aware as we should be of the bias in our customer service environments? Negative experiences undermine an company’s ability to foster a diverse customer base.
Change starts from within
The focus on diversification may be improving, but it is not yet solved. Just as ensuring that your business proactively targets diverse markets, you must power the change from within and ensure that your internal teams are reflective of the diverse nature of the customers you want, and the world we live in. Diversification in teams not only promotes creative problem solving and adaptation to new problems but positively impacts customer trust.
Bias limits growth
Many types of bias effect our cognitive reasoning and decision making, however this is further magnified when utilising machine learning (ML) tools. Wherever ML ensues, the ramifications of less than prudent design and execution can appear. The most common shortfalls are of data being used incorrectly training models with bias, substandard data to begin with, or models being poorly optimised. Though the ramifications are serious, breaking the models into broad strokes helps data scientists to raise the important questions that you need to ask of your data modelling and how it is trained.
What steps do we take to avoid it?
"As data scientists, we have a responsibility to constantly check model design.
Have you intentionally/unintentionally ignored any group? Is the model optimised and methodology appropriate? Were features set in the training environment similar to those experienced in the current environment?... Are just some of the various questions we consider." He commented.
What should you know, when you start looking for bias in your business?
1) Realise that bias can happen to any business, big or small. (It's happened to global titans Google, Amazon, Apple too...)
2) Bias isn’t always easily seen so be prepared to dig to find it.
3) Be objective - don’t audit with bias. If you don’t have the resources to do it yourself make sure that you partner with a third party who you can trust to be objective.
4) Keep an eye on the shape of your people, those you recruit, reward and retain.
5) Schedule regular data reviews; audit your customer data, your communications, your target audiences, imagery, and messaging. Regularity is key - processes and models are journeys not destinations.
Jordan Browne-Moore is a Consultant & Bias Expert at Beyond Analysis. He has extensive academic grounding in Finance & Economics paired with unparalleled knowledge in computer programming, machine learning and AI. Jordan supports clients to understand their data and drives efficiency gains and business improvements, with a strategic focus.
Click here to watch the full webinar featuring Jordan or take a moment to read more about bias in machine learning and AI.
Got a question? Get in touch.