Why Human-Centered Design (HCD) Doesn’t Always Work for International Development
Human-Centered Design is a creative problem-solving approach that puts the customer or beneficiary at the center of the design process, using a replicable process driven by empathy, collaboration, and co-creation. In the last ten years or so, HCD has become increasingly popular for designing products, services and programs to fight poverty in international development, inspired by successes from the private sector. It’s most widely recognized by the number of sticky notes on the wall in meeting rooms around the world!
Instead of driving initiatives by asking questions like “how will we hit our targets?” or “what will our donors allow us to do?” we start by asking more meaningful questions like “what does the consumer need?” and “will this solution add value to their lives, and if so, how?” It forces us to get to know our target populations’ needs, wants and barriers, and then design for them - all by asking good questions, actively listening, and getting feedback from testing our ideas with them.
As a human-centered designer myself who has spent the last decade applying a variety of HCD methods to international development problems, I believe in the power of human-centered design. However, as I see lots of development dollars going towards HCD approaches, I have questions about when HCD works best and when it might not be the most appropriate approach. To this end I worked with the Savings Learning Lab at Itad on a publication that takes a critical look at the application of the HCD process in international development to understand what's working, what's not working and under what conditions the application of HCD could lead to better outcomes.
What has worked well in applying HCD to international development?
To design something people will use and benefit from, we need to understand their motivations and needs, why they say yes or why they say no, their day-to-day experiences, their problems and their goals.
I ran a number of design sprints with fintechs where we focused on specific challenges the businesses were facing, many of which had to do with improving the sales journey for the customer. One fintech wanted to improve sales and usage of a bank account immediately after sign-up. The design sprint included prototyping and testing a highly visual brochure, simplifying language and modifying the customer journey to encourage sales staff to show the customer how to cash-in immediately after signing up for the account. These changes led to a 50 percent increase in sales and increased account activity.
In Cambodia, the microfinance institution (MFI) AMK used an HCD mapping exercise to understand their clients' cash flow, revealing the critical insight that monthly repayment for a traditional loan would be impossible given the irregularity of their income. With this understanding, the MFI designed an innovative and highly successful loan product with a flexible credit line over a two-year period.
HCD is especially effective when addressing a narrow or focused problem with a clear path to change, such as designing marketing materials that communicate effectively with the target market, finding strategies to increase the number of sign-ups or usage of a service, or improving a digital user experience. For these types of problems, HCD can help us think outside the box and give us ways to test our ideas early and cost effectively.
What hasn’t worked in applying HCD to international development?
But there are areas where HCD hasn’t worked as well. Complex, large-scale systemic problems such as poverty, income inequality and restrictive social norms are difficult to address with HCD. Having emerged from the private sector, HCD focuses heavily on outcomes like uptake, usage, retention and conversion rates as early signals of demand, which is equated with success. But these indicators are more suited to measuring short-term performance. There is little evidence linking longer-term social impact to HCD, as the approach does not usually go far enough to show whether applying HCD methods to international development interventions leads to social impact.
In development, design engagements tend to be one-off and short-term (three to six months), with an emphasis on improving desirability of products and services but not enough follow-through to ensure a lasting impact. For example, UGAFODE, a Ugandan MFI, embarked on an HCD journey which revealed that rural customers needed a way to access their savings. Through the design process, they developed a "savings scratch card" that could be widely distributed through small retail shops or kiosks. However, the design company was no longer involved when the pilot was implemented, and UGAFODE had limited capacity to collect feedback and translate customer insights into improved features. The product was consequently unable to gain traction and never got off the ground.
As most programs hire international design firms to run HCD projects, HCD can also suffer from the phenomenon of "parachute research," in which outside consultants come in to the country only for the project and then leave. Without being immersed enough in the local context, they cannot provide the local insights which are often necessary to design a successful product for the target market.
Finally, not all organizations are set up to apply HCD methods effectively. HCD is fast-paced and sometimes requires quick and multiple changes. This can be difficult for large commercial banks, for example, where product development and process changes tend to take a very long time. Fintechs, on the other hand, tend to be more agile and iterative and better able to take on HCD processes.
So what do we take away from this exercise?
The HCD process helps us better understand value propositions for customers, which development organizations tend not to focus on as much as the private sector. However, we can’t only focus on desirability of products. Feasibility and viability must be considered from the beginning as well, in order to make sure a project and its impacts will be sustainable. And at the end of the day, regardless of the process we use, we need to keep challenging our assumptions and keep designing with the user at the center of the process.
There is a problem with this blog that stems from its source, a Savings Learning Lab document ‘Human Centred Design in International Development - A review of what works and what doesn’t’ by Rathi Mani-Kandt (the author of this blog) and James Robinson 2021.
The central issue is that the shortfalls that Rathi and James report aren’t problems with the HCD approach at all. Instead they are problems with the planning and execution of the projects they have chosen to report. So, the problem is with the practitioners of the projects and their leadership, not with the process of HCD. The blog isn’t a review of what works and what doesn’t – it’s a review of ‘what’s worked and what hasn’t’. Quite different. The blog and the underlying report are specific and anecdotal and not general and conclusive - and I think its important to recognise this.
So let me pull out two or three points from the blog:
(i) Rathi says HCD is a process driven by empathy (so does the report) but it’s not, it’s a process driven by evidence - and that’s a massively important point of difference. HCD is a live, changing, practitioner led discipline, but it hasn’t let go of its core commitment to evidence. HCD is not some magical, inductive, creative process.
HCD uses qualitative research methods that place the human at the centre of the analysis and uses observational and other techniques derived from disciplines (amongst others) such as anthropology, psychology and traditions of participatory design to gain insight into people’s authentic needs and create design solutions to meet them.
Empathy has a role, but only a role - the integrity of the process is not reliant on the researcher empathising with the people they are researching, it’s about using observation and discussion to gain insight as a way of understanding needs and experiences – this moves the emphasis considerably and if the culture is not yours, you’d better be sure to have locals on the team.
I’ve been reflecting on how a confusion between evidence and empathy arose, and I wonder if it doesn’t lie with the work of one or two of the big-brand US design agencies who have been active in the development space. They may not reflect best practice in HCD, or to be more generous, they may just reflect their own ‘in-house’ forms of HCD.
(ii) So back to the blog, Rathi says HCD is ‘especially effective when addressing a narrow or focused problem with a clear path to change’, well, that’s true in part. HCD can effectively tackle these issues, but that’s not the only or indeed major value of HCD. Broad, deep, complex, ’wicked problems’ (as they are known in the discipline) are very amenable to the broad contextual framing that HCD has of systems problems. HCD is nothing if not a discipline based on systems thinking. Service Design, another HCD term for systems-based problem framing, addresses just those things.
Whilst marketing comms, sign ups and service usage are, as Rathi says, all tractable by HCD, they are by no means representative of what HCD can and does achieve.
(iii) So ‘what hasn’t worked’. Well, Rathi tells us that poverty, income equality and restrictive social norms are not the ideal problem candidates for HCD. Frankly I’m not surprised. But I would be surprised if HCD as an investigative methodology couldn’t be an extremely useful tool amongst others in tackling these issues. I would be even more surprised If a funder had ever commissioned an HCD project team who claimed to be able to tackle poverty.
The blog asserts that the origin of HCD lies in the private sector (not true by the way, it lies in the military and in early computing) and that this means success is inevitably quite short lived and limited. This makes me a little hot under the collar because it is so palpably wrong. The very selective examples given to make the point (such as conversion and retention) just aren’t representative of the range of success metrics that can be achieved.
Let me take a big example. Done right, HCD is quite capable of identifying systemic problems in an organisation that point to a need for restructuring to deliver better long-term outcomes for *whomever*. That’s nothing whatsoever to do with conversion rates! In fact a team I worked with achieved precisely this in Central Asia a few years ago
Finally, the blog reports a series of well-known best practice guidelines that are used in the blog as exemplars of problems with the HCD approach.
So, looking at those:
• Yes, it is useful for HCD practitioners to stay engaged with a project – that’s why parachuting in external experts and not skilling up local practitioners is generally fruitless, and for the same reason that is why it is not a good idea not to try run an HCD project in Africa from a campus on the West Coast of the US.
• Yes, it is important to make sure that solutions are contextually relevant and achievable, it’s also important to make sure they are credible
• Yes, HCD can be fast paced - but it doesn’t have to be - and if the process is wrong for the organisation it is aiming to support, it will fail.
• Yes, for HCD outcomes to be sustainable the culture of an organisation needs to be supportive, and preferably customer centric.
• Yes, ‘Desirability’ is not the central point of HCD in development, or indeed almost anywhere.
• Yes, Viability and Feasibility are critical to account for (early) in solution development.
There are no surprises here other than that the author seems surprised. I however am surprised that a 101 understanding of HCD methods is described as ‘news’.
I’d love to open a discussion about the real and extensive capability of HCD to tackle issues in International Development, and perhaps also consider why some projects appear to have lacked the leadership and experience necessary to deliver on this, and what we can learn from and do about this.
Hi Mark,
Thanks for the comments on the blog and appreciate the enthusiasm for HCD. As one of the authors, I want to reiterate that I am someone who ardently believes in the value of HCD, and I am very happy to have you engage in a debate. The title may seem like a full critique of HCD in International Development, but the paper does delve into exactly what you say--- "what has worked and what hasn't" and yes, how practitioners plan and execute HCD in international development. So yes, exactly, it is not meant to dismiss or praise HCD as a design process in itself, rather, find trends in how it has been applied and what has worked and not worked at this stage - admittedly (as we mention in the paper quite candidly), we were not able to review every existing project or program, and we don't claim for this to be a review of everything.
In terms of evidence, metrics, and empathy. I would continue to assert that empathy is central to HCD, and that by saying empathy is critical, we are not excluding evidence as important. And I agree that HCD as an investigative methodology is extremely useful in tackling large-scale poverty and international development issues. We don't say that it can't solve wicked problems, rather that it has not yet to date been super successful at that for a few reasons, including - lack of ability to connect HCD "the process" to longer-term metrics of success, as well as the inability of the sector / organizations to fully invest in the process (these are not the only reasons). Again, we focus on practical ways it has been applied to date.
We are simply stating that there have been short-falls in how it has been applied and that we should investigate, talk about these, discuss them and change the way we plan international development + HCD projects, so that we can leverage the full potential of HCD processes. You say a lot "when done right, HCD has the potential to..." and I agree, it has high potential, much of what we state in the first section of the paper. And what we wanted to do, is reflect on how it has been practiced, and understand if there are better ways to practice it to get to the outcomes we want.
And perhaps the blog reports "nothing new to you" in terms of the "series of well-known best practice guidelines" you state, as you might have a more advanced understanding of HCD, but again, we are simply stating that we shouldn't "drink the juice of HCD" and keep putting tons of funding into HCD projects, without being aware of these shortfalls in the experience to date and looking for ways to overcome them.
Finally, we are not asserting that HCD "was invented" by the private sector, rather it was popularized by them in various ways, especially tech firms, which became mainstream enough to be picked up by development organizations as a way to think outside the box.
I would like to point out that this type of debate is healthy for refinement of our thinking around HCD. HCD is amazing, but it is not always the answer/silver bullet, and has to be implemented correctly. Would love to continue a healthy debate and discussion as we are able to. Hoping your collar is less hot! :)
Thanks so much and happy HCD'ing!
Cheers,
Rathi
Good article emphasizing the reality that you cannot only have a hammer in your toolbox, you do need other tools as well. Specific tools for specific tasks indeed. What needs to be mentioned is that if you drop the D from HCD, it becomes an approach and not a methodology. Being human-centered is a great guide in all situations.
Couldn't agree more, Gerhard! HCD as an approach is indeed a great place to start.
Leave a comment