“Scaling is the road, not the destination.”

How best to scale up innovations? The paper on this topic in our recent Strategy documents drew on external expertise as well as our own. Among those we consulted was Lennart Woltering, Scaling Advisor at CIMMYT. He takes issue with some of our views. We asked him about that – and for his advice on scaling more generally. You'll find the main interview below and Part 2 here.

Syngenta Foundation: You say that for scaling, «10% is the innovation and 90% the context». What do you mean by that? 
Lennart Woltering: As a global research institute, CIMMYT brings innovations into a wide range of contexts. The same innovation has a very different ‘scaling pathway’ in Zimbabwe to that in Bangladesh. Socio-economic, agro-ecological, cultural and many other contextual factors play a role. Fortunately, one can also innovate to change the context! In the pilot phase, you want to see if the innovation works; during scaling, most energy and resources should shift. Here you need to understand and improve the context for the innovation to perform beyond the pilot environment. This requires much greater multidisciplinary efforts and strategic collaborations. It also needs long-term investments, not only financial but also in the development of organizational and institutional capacities at the local level.

What would be an example of that at CIMMYT? 
The Cereals Systems Initiative for South Asia (CSISA) project in Bangladesh. CSISA aims to sustainably intensify agricultural systems by encouraging the use of surface water irrigation and efficient agricultural machinery. It’s not only about having the appropriate machinery. It’s also about improving the enabling environment to help sustain and enhance the scale of innovations’ benefits. 

How does that work?
By tackling other scaling ingredients such as attractive business models, collaborations, local knowledge and skills, and leadership. For example, CSISA focused on creating business partnerships to encourage private sector leadership and investment in machinery markets. It also addressed numerous other elements to improve access to market information as well as enterprise development*. 

What about scaling outside agriculture? 
To take a current example: Although the COVID-19 vaccine is often the same, vaccination rates vary a lot between countries. Each nation has needed to innovate somewhere else, be it in awareness, distribution channels, and/or other capacities.

On Twitter, you urge us all to «Go beyond the numbers». What does that mean in practice? 
Projects may reach an impressive number of beneficiaries, but this is no guarantee that they can and will continue to use an innovation after the project ends. With funding often tied tightly to the number of “direct” beneficiaries attributable to the project, project implementation teams tend to chase the numbers of end-users to please donors. According to Goodhart’s Law, “if a measure of performance becomes a stated goal, humans optimize for it, regardless of any consequences”. That makes the measure useless. 

But surely with such massive problems as hunger and poverty, lots needs to be done on a very large scale?  
Yes, but these are very ‘wicked’ problems. They don’t have quick fixes. The more we focus on reaching numbers fast, the less is invested in what matters: local capacities, listening to changing demands, and questioning whether the outputs really result in better outcomes. Scaling is a process: It is the road, not the destination. On that scaling road, the direction is often more important than speed. So: Recognize and address root causes and invest in local capacities and relations to improve the enabling environment. This will result in high numbers that can be sustained or grown after initiatives end. Don’t try to do it the other way round ("First reach high numbers and the rest will follow…").

Your Scaling Scan helps «determine the strengths and weaknesses of one’s scaling ambition». What are some key strengths and weaknesses you repeatedly see enabling or blocking scaling? 
The Scaling Scan builds on the 10% / 90% idea I’ve just described. After developing a scaling ambition, we look at ten ingredients to predict the bottlenecks and opportunities. One ingredient is the technology or practice; the other nine revolve around the context. That includes the policy environment, value chains, and accessibility and affordability of finance products, for example. Typically, the innovation itself and associated evidence and capacities are very conducive for scaling. Finance and business cases are almost always the bottlenecks. 

"Scaling is an art and a science"

You’re glad that scaling is «finally getting recognized» as a science and an art. Historically, mass-scaling like the spread of religions or languages has tended to include some coercion. Some people complain about getting coerced into Covid vaccination. How far can science and artfully replace coercion in scaling? 
Completely, I hope, in agricultural development. But why do I say scaling is also art? Because it requires a lot of soft skills such as trust-building, navigating power dynamics, and recognizing and acting during opportune time windows. That is not something that someone can prescribe with scientific formula, or fully capture with scientific methods. 

As well as the science and art of scaling, there is also the mixture of action and research. What is currently the most interesting scaling research to accompany the action, and who’s doing it? 
What we call scaling in the agriculture sector has been known much longer in the health sector as “implementation science”, hence researching the actions. The Brookings Institute works on real-time scaling labs in education is refreshing: It focuses on learning rather than steering projects. Canada's IDRC takes a hands-off approach to its implementing partners; it focuses on learning and improving. 

What about your own overall organization, the CGIAR?  
Here it’s great to see the interest and demand in systematically planning interventions using scaling tools and frameworks based on decades of learning about scaling. We have matured in that we now appreciate that scaling doesn’t happen automatically when one has a good innovation. It is an art and a science, and one that requires recognition of roles and responsibilities within organizations and among different actors, especially by linking different sectors.  

Who in your Community of Practice and/or elsewhere in agricultural development is doing scaling particularly well, and why? 
The education and health sectors have been ahead of agriculture. It is a great learning adventure to interact with them. Agriculture is catching up and there are many good examples in the Scaling Up Community of Practice (CoP). The IDRC has helped me see the difference between scaling for the public good and the dominant scaling approaches that focus on expanding operational scale, growing market share, and achieving commercial success. The latter is not wrong, but incomplete when applied to development efforts where the goal is social impact and the public good. IDRC focuses on finding not a maximum scale for a few, but on balancing the positive impacts for society and the environment beyond the target group and area.

Another organization from which I am learning a lot is Catholic Relief Services (CRS). They recently launched their new strategy 2030. CRS is setting its land restoration platform to scale water-smart agricultural practices, not for the sake of large numbers, but embedded in a larger strategy for the transformation of landscapes in Central America. It was very enriching to work with them.

«Systems change» is a buzzword used by many, including the Syngenta Foundation. You and others also use the term but point to the difficulty of knowing which system to change, without negative effects on other systems. What is your advice to agricultural organizations in this area? 
Inflationary use of the term is causing confusion. Our Scaling CoP has been focusing on the interfaces between scaling and systems change – there is also some good CoP website material about this. It is really important to specify what systems are changing. For example: scaling fertilizer use will probably make farmers richer ("systems change = poor to rich farmers"). However, it consolidates an agri-food system reliant on fossil fuels, which some people consider bad. The project is opposing its efforts to change away from this dependency. Scaling doesn’t always only lead to improvements. One needs to keep a healthily critical view of what systems the intervention is changing and be transparent about it.  

Another favorite bit of development jargon is «enabling environment». Using a real-life example, how would you explain what that means for an agricultural innovation? 
A 2018 ICRISAT study by Verkaart et al. concluded that the enabling environment was responsible for the remarkably fast spread of improved chickpea varieties in Ethiopia. There were almost no barriers to success – there was good market access, accessible extension services, and the chickpea was already an important crop and thus held value in the community. 

We need to be careful about saying “this innovation is ready to scale”. Under what conditions? ‘Stage-gating’ decisions for advancing or stopping programs need to consider this as well. Maybe the innovation does not meet the criteria in Kenya, but can it be in Ethiopia, or what in the context is holding it back? Stage-gating can be helpful if it prompts such questions – use it for “pause-reflect-act”, not “stop-go”. 

You’ve pushed back at our work on scaling with the question «Who decides what is good and what should scale?» How does CIMMYT tackle this dilemma? 
By engaging in the early research stages with farmers of both sexes and other decision-makers. This way, the stakeholders engage with the scaling process from the beginning. There are some good examples: in Southern Africa, a range of conservation agriculture practices are tried, tested, and evaluated by farmers and extension agents; CIMMYT plays only a supportive role. The farmer decides which combination of conservation agriculture practices works best for him/her. This results in slower but more sustainable adoption. The new Crops to End Hunger initiative puts a lot of attention on responsive breeding, considering the needs of farmers and others in the value chain. 

However, your question harbors a very tricky problem. See my comment above about fertilizers, for example. Who am I as a Westerner to decide if farmers in Africa “shouldn’t use fertilizer”? They currently apply less than 20 kg/ha, compared to a global average of 136!

You and others highlight the difference between the ability to scale and willingness to do so. Presumably more ag development organizations are willing than able? Or what is the issue here? 
Many development organizations overestimate their capacity to scale. They see it simply as an extrapolation of what was done to deliver the pilot. To use an analogy: if someone moves from using a bike to a car to a bus to a plane, not only the carrying capacity differs but also the skills required. But there is another important element to this: handovers and exit strategies. At the end of projects, the local private and public sectors are the only people who can sustain any changes. Lack of motivated and capable local organizations is a major reason why these projects too often lead to scattered and short-lived improvements. We need to get serious about organizational capacity development and really addressing local needs. Then local organizations will have the capacity to innovate, collaborate and scale, and be willing to invest their political and financial capital.

How will the ‘One CGIAR’ reform help CIMMYT and other CG centers to scale up better? 
I am hopeful that it will break silos between institutions. That would allow us to be more outcome-oriented, asking “What does it take to get this community out of poverty?”, rather than finding a problem for a solution we’ve developed. Collaboration across institutes will be more easy and fruitful because we can draw on a much broader base of colleagues and experiences. More emphasis on systems change and food systems transformation will also help fit our solutions into the larger picture. More space for cross-regional and cross-program exchange and learning are also critical benefits of One CGIAR.

On a more personal note: We assume that you sometimes scale down your thinking about scaling up. What do you do in your spare time? 
We have a two-year-old son and since the COVID-19 situation in Mexico has been rather tough, I am spending most of my spare time with him. That is a blessing, considering how much I was traveling before… 

Here’s some biographical information about Lennart Woltering. 

*Important factors for the success of CSISA include spatial and market targeting to facilitate demand creation, training, and awareness building of appropriate machinery, capacity development of mechanics to repair and source spare parts for appropriate farm equipment, assurance of after-sales services, and technical advice deployed through state extension partners, and support for emerging service providers by linking them to farmers and farmers' organizations, as well as developing business models that generate new income-generating opportunities.