From Small to All: Designing Scalable Pilots

Re-posted from CAWST (original blog post: http://www.cawst.org/blog/bydate/2016/06/from-small-to-all-designing-scalable-pilots/)

Photo

Ampthill, UK and Calgary, Canada – Although many pilots and trials seem to succeed at first, they often fail to go to scale, remaining as islands of success in a sea of mediocrity. Dissecting our WASH sector’s track record of delivering solutions at scale is a valuable opportunity to learn and improve. This blog explores the issue of scale, stimulated by a recent discussion triggered by our respected colleague, Jan Willem Rosenboom of the Bill & Melinda Gates Foundation. We conclude by proposing three ways to build hardy pilots that transcend common pitfalls on their way to scalability.

From innovation to scale
When a new WASH approach or technology[1] is devised (usually a variation on an existing theme, for there is little that is truly new under the sun), those who have devised it and those who make decisions about implementing new things want to try it out at limited scale before scaling up. This is sensible, especially as rolling out at scale prematurely could put people and resources at risk. “Let’s try it out first, to see if it really works” is a highly rational strategy prior to scaling up. The questions are, how should that trial, demonstration or pilot be designed, funded, implemented and monitored to provide a solid basis for scaling up? And how should the trial or pilot be extended into the application at scale?

The answers to these questions depend on where the new approach or technology is positioned in the economy. Broadly speaking there are three possible business models.

Market-based solutions
In this instance, the innovation sits within the market, so scaling up is mostly a matter of consumer uptake. Successful examples of this are irrigation treadle pumps, the now-ubiquitous yellow plastic jerry can, and household water treatment technologies, all of which are marketed directly to households.

In this private sector / market model, new technologies (for it is mainly technologies rather than approaches which fit here) go through an arduous process of conceptual design, prototyping, proof-of-concept, iterative testing and re-design, final production and launching to market. Many fail. The fact that many fall by the wayside is normal. The few that succeed are robust and fit for survival, at least while the environment in which they evolved or were developed persists. The path to scale is strewn with failures, and the high number of failures is a reflection in part of the inadequacies of communication between designers and potential users. This communication gap is increasingly being addressed through user-centred design[2] processes.

Solutions requiring significant policy change
The second possibility is that the approach or technology requires significant public policy change. An excellent example is the adoption of Community-Led Total Sanitation (CLTS) into national policy.

A new approach or technology which can only go to scale with public policy backing or approval faces a different set of challenges. It still needs to start small, but its path to scale will require convincing not only the user market, but also key decision-makers in Government. The clearest example of such an approach is CLTS. It started in the work of one nongovernmental organization (NGO) in Bangladesh, but through a combination of factors it rapidly and organically grew. We suggest that the reasons that it went to scale included (a) the desperate need for a sanitation approach which worked (nothing else was working at the time), (b) the visible but informal evidence that this approach did indeed work, and (c) the push provided by a charismatic promoter of the approach and his professional colleagues[3]. Incidentally, it is noteworthy that the approach was taken up into national policies over a period when the formal, rigorous evidence that CLTS worked did not yet exist. Policy is not made entirely on the basis of complete, rigorous, research-derived evidence.

Subsidised roll-out
In the third possibility, public or nongovernmental organisations (with their donor partners) might adopt a new technology or approach and convey it to beneficiaries in a highly subsidised or free manner. The rolling out of immunisations by Governments, or distribution of mosquito nets by Government and NGO programmes are examples. In the WASH sector, most capital investment in water supply is highly or wholly subsidized; in contrast, in sanitation and hygiene the current wisdom is that subsidies should generally be avoided.

This third way of progressing to scale – through highly subsidised roll-out – is only possible through very large Government / donor programmes. Free bed nets and immunisations, for example, are delivered at scale through national programmes funded by Governments and international donors. Part of the reason we have not yet reached scale with water supply improvements in many low-income countries is precisely because of (a) the need for a high level of subsidy of capital costs (a need which will only increase as we pursue the more demanding Sustainable Development Goals (SDGs), and (b) the gross insufficiency of national budgets (even with donor support) to provide these subsidised investments.

Optimal scale
It is common to think of scale as a single programme or intervention that reaches an entire country or region, but scale may mean different things in different contexts. In many cases, there is an optimal scale for working within existing systems. Scaling up may then mean replicating at that natural scale, resulting in multiple projects or programmes adapted to each context, as opposed to expanding into one huge programme that needs to work for everyone.

In market-based solutions, getting to scale means progressively saturating the potential market. Once every household has purchased a latrine slab or biosand filter, there is no further market save for replacements, upgrades and the needs of new households. With interventions delivered by multiple actors to a common standard approved by public policy, scale implies that all actors are using a common approach and focusing their investments on it. In interventions delivered by Governments, we look for the replication of an approach at the natural unit of public administration, typically (but not always) the district local Government.

In conclusion, we posit three practical and tangible suggestions about how to design pilots so they are valuable towards achieving optimal scalability.

  1. Think of the pilot as merely phase one of the at-scale programme. There is a good reason why most innovations start out small. Implementers need to build the capacity of their team, figure out how to engage the community to create and sustain demand, test what financing model is realistic, create practical systems to monitor for improvement, and begin to understand how the approach will need to be adapted for different contexts. From the pilot and beyond, focusing on the outcomes and impacts we seek is critical – inclusion of the most vulnerable, sustainability of results, lives changed for the better. If our eyes are not constantly on these end-goals, we are likely to miss them.
  2. Plan for scale from the outset. There is widespread agreement that reaching scale requires thinking about scale from the outset. We would suggest that planning for scale means (a) identifying which of the three models above (or what combination of more than one of them, for the categories are not mutually exclusive) will take the innovation to scale, (b) working out the financial and affordability constraints of getting to scale (maximum price points for market-based technologies, affordability of technologies to be rolled out through public funding), (c) figuring out what capacities (human resources, skills, institutional arrangements) will be needed for delivery and sustainability at scale, and (d) in all cases identifying the aspects of the operating environment which may need to change in order to reach scale.
  3. Remember that there is no scale without user acceptance. Above all, whatever new approaches and technologies are proposed, user acceptance is a make-or-break condition. You can lead a horse to water, but if it isn’t thirsty, you can’t make it drink. Designing for scale involves designing with the potential users.

Successful, scaled-up programmes and interventions exist, and they usually started out small. If we are to achieve large scale impact, we must learn from what worked and what didn’t, and design better pilots.

Richard C. Carter, PhD, FGS, CGeol, FCIWEM
Millie Adam, BSc

Richard Carter, Advisor at CAWST, has worked in the natural and social science and engineering of water resources, water supply and sanitation in low-income countries for the last 40 years.  After periods in consultancy, academia and the INGO world he now directs his own consultancy firm (www.richard-carter.org).

Millie Adam, Strategic Initiatives at CAWST, is an engineer who has worked in water treatment, corporate social responsibility and international development for the last 15 years. She works directly with our CEO to build CAWST’s international profile and collaborations.

_________
[1] In this blog, ‘approach’ is taken to mean a way of going about promoting, managing or financing some aspect of WASH. Examples include CLTS, sanitation marketing and health clubs. ‘Technologies’ refers to components of the physical hardware needed for water supply, sanitation or hygiene.

[2] http://www.userfocus.co.uk/consultancy/ucd.html

[3] Kamal Kar and Robert Chambers leading many others.

 

More information on scaling up:

http://washtechnologies.net/en/

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s