Design for Lasting Impact that Goes to Scale

Design for lasting impact that goes to scale

Rainer Arnhold Fellows design iteration flow format (DIF)
August 2011

Here is the primary tool the Rainer Arnhold Fellows Program uses to help people with a good idea design for maximum – meaning scalable - impact.

This is a design tool to give you the best shot at real impact that will go to scale. The basic approach is this: identify exactly what impact you want, lay out what behavior must happen to get it, formulate the interventions that will drive the behavior, and organize the whole thing into a systematic impact model that can scale up. The result of the process– your model in the DIF – provides a vehicle for easier ongoing iteration and evolution of your idea and your model.

There are three parts to the DIF:

Part one: sketch the basic model (steps 1-7)
Part two: flesh out the model (steps 8-10)
Part two: flesh out the model (steps 11-13)

There is a separate worksheet template to make it easier to use this format.

The best way to use this is to:

  • Read through the whole thing quickly
  • Do part one – don’t move on until you’ve got a completed version that you like. Run it by someone familiar with the DIF if you can.
  • Flesh out your model in part two to the degree that is useful to you and that allows enough flexibility to easily make changes – too much detail and you’ll be over-committed, too little and you won’t capture what really is distinctive and innovative.
  • Look over part three and use what’s helpful. Don’t worry about completing it unless it’s useful to you and/or it provides material for a helpful discussion.

Part One: the basic model

1. Mission

Decide on and say exactly what you’re trying to accomplish with your big idea. This is what everything you do will be designed toward. Capture it in eight words or less, and include a verb, a specific target population or setting, and a big outcome that implies something to measure:

  • “Get African one-acre farmers out of poverty”
  • “Prevent HIV infection in Brazil”
  • “Save coral reefs in the South Pacific”

2. Big Idea

This is about your central, distinctive idea about how to create lasting impact at scale - the idea at the core of all you do, that you use to create change for the better (some people might call this your “theory of change”). To formulate the big idea you need to know two things: how you’ll change behavior, and how that can go to scale. Get your idea down to a sentence that captures your special sauce. Keep working it until you really like it. Be aware that it may change after you’ve gone through all the way through part 1.

  • "Design, market and sell money-making products that farmers can afford and will use”
  • “Eradicate devastating invasive species from islands so that endemic species and ecosystems can recover”
  • “Use existing community groups to provide poor farmers with the integrated delivery of farm education, credit and access to cash buyers they need to make a decent living”

3. Impact

Identify the single best indicator – an outcome, not a behavior - that you would measure to know if you’re fulfilling the mission. This is critical: it defines what you’re designing for:

  • “Increase in farm income”
  • “Decrease in HIV infection rates”
  • “Biological indicators of coral health”

4. Behavior mapping

Impact comes from action, from someone doing something differently. What do you want people to do differently to create that impact? What are the end-user behaviors that will directly lead to the impact you’re looking for? Design for impact is focused on behavior – how to drive it and maintain the behavior(s) that create impact.

To start, put down the most critical behavior change that must happen for impact.

  • “Farmers adopt new set of farming practices”
  • “Teenagers practice safe sex”
  • “Island communities guard reefs and maintain sustainable fishing levels”

Behaviors that directly create impact don’t happen in isolation, though -getting to impact requires a connected sequence of behaviors. To make sure that impact really happens, you’ve got to “connect the dots” and run through all the behaviors necessary for impact. List them in sequence as a simple flow diagram of who must do what all the way to impact. Start with your key behavior and use it as the starting point for a list. It may be that not everything fits into a chain of sequential behavior, so just list any additional behaviors off to the side – that’s why we call it a map.

5. Interventions

Look at your list of behaviors. Figure out which ones would happen without you and take them off the list. Of those that remain, put down – in brief - what you’re going to do to make them happen. These are your interventions. This is the list of the things you have to do – the activities that you must craft into a scalable model. Bear in mind that changing behavior is hard, and that ideas that require you to change more than 3- 5 behaviors are going to be pretty complicated – i.e not very replicable – to implement.

This is the step that often sends people scurrying back to do more research as it becomes clear that they didn’t think about some aspect of behavior change or don’t yet know the state of the art.

One useful tool: try looking at each necessary behavior in terms of conditions and incentives: are the conditions in place such that the behavior can happen, and are the incentives there so that it will happen? This is a systematic way of checking that the intervention is sufficient and it often suggest way of adding or modifying interventions.

6. Route to scale

A big idea is only a big idea in the context of how it will eventually scale up. If you don’t design your model with the route to scale in mind, it probably won’t happen. If you think about it, there are five ways to take impact to scale:

  1. By growing a really big organization
  2. Via the market
  3. Via governments
  4. By co-opting other NGO’s
  5. Viral spread of behavior

Pick the one that best captures how your idea will lead to impact at scale in the long run. It might take you an intermediate step or two, but identify what route will get you to a million – that is a really useful frame of reference. See Part 3.A for more detail.

7. Impact Model

This is a way to get you thinking systematically and a format that allows you to continually mess with your model. Don’t panic about the form; just try it. You can use a boxes-and- arrow format, some other kind of flow diagram, or simply a list. Use a pencil and blank paper - or post-it notes - if that works better than doing it on a computer.

The point is to arrange your activities in some sort of sequential order in a way that describes a systematic process that goes all the way to impact. Create new boxes to fill any gaps in the logical flow; you don’t have to limit yourself to the interventions from the previous step. Make sure there isn’t anything important that you need to do that wouldn’t fit into or be represented by a box. There is no proscribed, “right” way to map out the boxes and arrows; just keep fiddling with it until you have something that makes sense to you. The result is a flow diagram of your impact model.

Part Two: flesh out the model

8. Detail

The devil is of course in the details. Once you’ve sketched your impact model to capture a replicable process, provide the details that bring each step to life. Write a brief narrative that captures what is distinctive and necessary for each step. Imagine that you’re trying to describe what you do to someone interested in doing it for herself.

9. Additional detail

It may that you’ll want additional detail in the form of sub points to capture what’s essential. Don’t add more than are needed to ensure that everything essential to a successful process finds a place in your model.

10. Scalability Audit

If you did all the stuff above, you now have a reasonably fleshed-out model. The Scalability Audit is a systematic reassessment of that model per the five Mulago scalability criteria, and it’s one the most important elements of this design process. These are the criteria:

  1. Real Impact?
    • Behavior chain: does your model ensure that the dots are connected all the way to impact?
    • How will you measure impact1 - can you sketch out a good-enough approach that includes the right indicator, quality numbers, and attribution?
  2. Cost effective?
    • Can you see a way to measure/calculate the cost per impact?
    • Do have the numbers to make any kind of intelligent projection?
    • Looking at the most expensive parts of your model, is there anything you can strip away?
  3. Lasting behavior?
    • Look at the conditions and incentives in play for the key behaviors you address in your model – can you make the case that they will last?
  4. Replicable?
    • Is the model systematic and simple enough that someone else could do it?
    • Could it be adapted to a wide array of settings and still reach your target population
  5. The right route to scale?
    • Does your model fit your chosen way to scale – or more to the point, is it specifically designed for it?

Use each of these questions as tools to go back and poke at your model. The most useful way to do this is to write out your answer to each – imagine that you’re trying to persuade a skeptic that you’ve got each covered, then be that skeptic. Fiddle with the parts as you need to; let the process reverberate all the way up to your big idea. Remember, the scalable model is the nucleus of all your work. It’s worth as many trips back to the drawing board as it takes.

Part three: going deeper

11. A deeper route to scale

With your model in mind, revisit how it will scale up. Remember, this is about how your model will create impact at scale, and not necessarily how your organization will scale. Here again are the five possibilities we talked about under “Big Idea:”

a. By growing a really big organization

  • Upside: high degree of control over implementation and quality
  • Downside: huge – and ultimately limiting - fundraising and management commitment
  • Best where quality of execution complex models is paramount

b. Via the market

  • Upside: essentially limitless potential infrastructure and capital
  • Downside: relatively few interventions that benefit the very poor or the common good have sufficient potential for profit
  • Best whenever someone can make money off key parts of the model without getting you off mission

c. Via governments

  • Upside: leverages policy and infrastructure
  • Downside: corruption, instability, and inability to implement complex interventions
  • Best for the delivery of “pulic goods” via simple, bombproof interventions at big scale

d. By co-opting other NGO’s

  • Upside: lots of them, leverages existing subsidies
  • Downside: NGO’s don’t often adopt or do a good job of implementing interventions developed by others

c. Viral spread of behavior

  • Upside: zero ongoing expenditure
  • Downside: there are very few models that exemplify this and there is no way to control quality
  • Best whenever you can get it; rarely happens

When we first talked about the way to take impact to scale, it was about identifying the ultimate vehicle for scale – what will do the heavy lifting at the scale of millions. What we mean by scale depends on context – the size of the problem and The Path to Scale is a brief description of how you’ll get there from where you are now. Sometimes that path requires sequential or even simultaneous combination of the various ways to scale. Put together a sentence or two that describes your current thinking:

  • “Via a really big organization: we’re going to build an organization that can deliver services to an ever-larger population, while building a fundraising juggernaut”
  • “Via the market: we will use donor money to prove a business model, grow it through self-generated expansion capital, then create infrastructure to assist others to use and adopt our business model.”
  • “Co-opting other NGO’s: “We will build an organization big and visible enough to allow us to form effective partnerships and get others to implement our model.”

12. Stage of Organization

The stage describes where you are on your path to scale, and it implies a lot about the kind of organization you need to deliver your model at this stage and move on to the next. Here are the usual stages things need to go through – pick the one you think you’re in:

  • Idea: constructing a starting-point model, looking at failures and best practices to date. Not much on the ground yet
  • R&D pilot: work on the ground at a scale that allows you to sort out just what your model is and how it really works
  • Proof-of-concept pilot: you’ve got a replicable model; now you see if it creates the behavior and impact you thought it would
  • Limited expansion: you expand operations to a size that allows you to work out the kinks prior to scaling up
  • Scale-up: dramatic expansion of impact, via your chosen path to regional, national, international, and, eventually, galactic scale

13. Financial model

This is about the money – how you’re going to finance the organization and its work. The financial model’s design is driven by these four fundamental questions: 1) what will maximize impact for the target population over time, 2) whether your impact model includes a revenue stream and 3) what is your intended path to scale, and 4) given 1, 2, and 3, what is the best source of capital (or, conversely, given the sources of capital, what structure makes the most sense for you)? Here are your choices:

  • Pure non-profit: no internal revenue stream, fully subsidized by philanthropy
  • Non-profit hybrid: some kind of revenue stream within the impact model, but still subsidized. Types of subsidy include:
    • Start-uponly
    • Start-up+expansiononly
    • Start-up+expansion+ongoingoperations
  • For profit: Two kinds –
    • Marketratesofreturn,usingconventional,mainstreamcapital
    • Sub-marketratesofreturn,using“socialcapital,”“patientcapital”orsome other source of financing (debt or equity that for whatever reason does not seek to maximize financial return).

Sometimes it makes sense to create a functional hybrid with two organizations, one for- profit and the other non-profit (we take a dim view of these, in most case – too complicated and too many pitfalls).

Another way to help sort out the for-profit vs. not-for-profit question is to look at the various kinds of financing out there and imagine with designation will give you access to the most of the right kind of capital. There are essentially three big buckets of money:

  1. Philanthropy and grants
  2. Earned revenue
  3. Loans: above and below market rate
  4. Equity: high and low expectation

Not-for-profits can use 1, 2, or 3; for-profits can use 2, 3 or 4. it’s worth keeping in mind that most high-social-impact businesses that target the poor are relatively low margin and unlikely to attract serious equity investment. In any case, what should guide your choice of financing and hence structure is this:

“What kind of captial will lead to maximum impact for my target population in the most timely fashion?


There, you’re done. If you’ve looped back around and fiddled with the pieces to the point where what you’ve got feels right, you’re ready to take it for a test drive in the real world. The DIF is structured so that tweaks are easy and iteration painless. Come back, revisit, and reiterate in a few months.
OK, have at it - the world is waiting.