A Human-Centred Approach For Designing Data Solutions
In 2004, Bank of America hired the design firm IDEO to help them attract more customers. When IDEO started the engagement, they did something unexpected, rather than going into Bank of America’s offices, they went to the field and observed how people handled money and recorded their spending habits. They found that people were rounding up their expenses because 1) it is easier to calculate, 2) they felt safe as it added a buffer in savings. However, people were still struggling to save money.
This insight inspired IDEO to design the “Keep the Change” program. Every time a customer made a transaction using Bank of America’s debit card, the purchases were rounded up and the change went to a savings account. The bank also contributed a fraction in the savings account each time change was deposited. More than 2 billion dollars were saved through this program. At some point, the stats revealed that 12.3 million-plus customers enrolled in the program since its launch in September 2005. So, what did IDEO do?
Instead of trying to change the strategy within the bank, or devising different marketing tactics to lure customers based on assumptions, IDEO looked at the problem from the customer’s point of view. They observed and engaged with the customers to understand their habits, feel their pain, and used that knowledge to design an unconventional solution. Such an approach to designing solutions is called Human-centered design.
The human-centred design approach places people at the centre, enabling designers to get closer to the people they are designing for, understand the actual problem behind the conditions they describe, generate ideas and build quick prototypes to test, learn and redefine.
It encourages designers to practice EMPATHY
The Observation (a.k.a Inspiration) phase encourages designers to drop all preconceived notions and empathise with the users — feel their experiences and know the motivation behind what they do. When designers see a problem from a user’s perspective without any bias towards a predetermined solution, they can find the origin of the problems described by the users. That helps in targeting the root cause rather than treating a symptom.
It connects PEOPLE at every phase
All the phases in this process involve the users. Designers work closely with the users in the Observation phase, they take feedback from users on sketched solution concepts in the Ideation phase and involve users in testing and experiencing the real solution in the Implementation phase. By including the people in the Observation-Ideation-Implementation phases, this approach gives ownership to them and let them have a say on the transformation that affects them.
It develops TRUST and CONFIDENCE
Involving users at every phase, taking their feedback and mitigating all concerns develops the organisations’ trust in the designers. The designers feel more confident about the solutions as those are built and tested iteratively based on constant feedback. It ensures that solutions are accepted in the organisation and the people would use them.
It fits well with modern TECHNOLOGY and PRACTICES
Modern technology platforms like the cloud enable rapid prototyping by allowing the business to build resources dynamically eliminating the overhead of managing and maintaining them. The cost is based on usage and so experiments can be very cheap which encourages iteration. Agile methodology can be applied to develop products, it is built to support changes and offer more flexibility than traditional approaches. This allows designers to release changes quickly and constantly.
A survey conducted in 2019 reveals that adoption of Big Data and AI solutions remains a major challenge in 77.1% of the companies, with executives indicating that 95.0% of those challenges are stemming from cultural changes (people and process), and only 5.0% relate to technology.
It is a well-known fact — the misalignment between business strategy and data strategy cause several project failures, and the lack of communication between businesses, data and IT create major gaps in what users want, and what they get. One of the prime factors that contribute to such situations is that the solutions do not solve the problems the users face.
Data doesn’t mean anything to the business until it is put into a context. Designers must understand that context by working closely with the business and its people. A lot of design decisions are made based on past experiences or applying templates from previous projects, and that reduces the chances of solutions to work in a different environment. Moreover, approaching a problem with a predetermined solution in mind can distract designers from the actual issue and move their focus on just delivering a product or a service.
Data solutions will only succeed when designers look beyond the printed requirements and find what has not been stated. Designers should form a connection between the people, processes and technology — this requires them to look at the bigger picture and think more than just categorising, integrating, and storing data. Designers must empower each user (and collectively the organisation) with the knowledge and insights that help them make the right decisions for the business.
In practice, data solutions projects come in different shapes and sizes, they have their own dynamics, and run with a defined set of principles often aligned to the specific organisation (or department). So, applying this approach to every project can be tricky. However, there are some principles that designers can follow in any environment that will help shape solutions in a human-centred approach.
- Forget about previous projects when approaching a problem. Look at the problem with a fresh lens.
- Empathise with the users — know whether they are describing a cause or a symptom.
- Don’t suggest solutions before knowing the root cause. Talking about solutions before understanding the problem can create boundaries in the design thinking process or even distract users from telling their story.
- Generate meaningful and actionable problem statements using Point of View (POV) method.
[User…] needs [need…] because [insight…]
— Data Scientist/s (User) need a sandbox because they want to run Machine Learning experiments.
— Data Analyst/s (User) need the latest data in the Data warehouse every morning because they send daily reports to external customers.
- Ask “How might we” questions to generate ideas. If you find out users cannot trust the data, say, “How might we develop users’ trust in data”, to generate as many ideas rather than suggesting solutions like “Implement Data Quality tools”, or lining up products.
- Decompose complex problems into smaller, manageable parts. Analyse each part, ideate and synthesise different ideas into possible solutions.
- Involve the users. Take their feedback on different concepts before building solutions. This could save time.
- Build, Measure, Learn
Eric Ries came up with the Build-Measure-Learn concept in the Lean Startup Methodology. It states “The fundamental activity of a startup is to turn ideas into products, measure how customers respond, and then learn whether to pivot or persevere.” The same concept can be applied here. Build small, ask users for feedback, learn and redefine, if required.
Data solution projects are often a result of enterprise-level data strategy. So, while designing solutions at the lower-level, designers should also remember the bigger picture. It is important to apply governance methodologies for reviewing alignment with the business strategy and ensuring the smaller solutions fit well to solve the enterprise-scale data management puzzle.
If the intent is to “Design for the People”, the worrying statistics on the adoption of Data and AI solutions will change in favour of successful implementations, generating value for the business and shaping a better-connected organisation.
Connect on LinkedIn.