Ann Mei Chang
Lean Impact was written by Ann Mei Chang. She graduated from Stanford in computer science and spent 20 years in Silicon Valley as a senior executive, most recently at Google. In 2013 she changed job paths into humanitarian work and became the Chief Innovation Officer for the Mercy Corps, and then USAID. Very quickly she discovered that it was much more difficult to “innovate for purpose, rather than for-profit”.
Outline
This post will follow the same three Principles of Lean Impact that Ann Mei identifies: think big, start small, and relentlessly seek impact. She then identifies engines for growth in lean impact, and ways that funders can promote innovation in the field.
1. Think Big
Startups in Silicon Valley are encouraged to think big. All companies are expected to show a ‘hockey stick graph’ that demonstrates anticipated massive exponential growth at some point in the company's future.
In order to make a dent in solving the many social challenges that face us, organizations must shift their mindset from a “linear growth path of service delivery” to the “exponential growth path of transformation”. This is a mental shift of ‘trying to do some good’, to “trying to actually reach a substantial number of people affected by a problem.”
Consider, for instance, the rapid spread of mobile phones. Today, more households in Africa have access to a mobile phone than sanitation. Why is it then that enterprises involved in social innovation progressive (at best) linearly? For instance, access to critical necessities such as water, electricity, and sanitation grow at under 1% a year. Ann Mei currently sees a mismatch between global problems and the strategies being deployed in the social impact sector.
Goal Setting
Ann Mai encourages the reader to set “a big audacious goal” based on the scope of the problem at hand. Too often the goals set by social impact organizations are too limited in their scale, and unable to achieve significant results. She identifies unclear and conservative goals as the characteristics that hold back many organizations.
Goals are too often unclear. They seek to “reduce hunger” or “improve health”. This would be the same goal as Google trying to “make more money”. Goals must be clear, measurable, and actional.
Goals are too often conservative. They often start with the available budget, or limited personnel available. Instead, the goal should start with a long-term (eg. 10 years) goal that clearly tries to find a solution that will reach a significant amount of the affected people. Ann Mei suggests that social impact organizations set a goal that raises the question: “are we trying to empty the ocean with a spoon?”
As part of the goal-setting process, Ann Mei encourages social impact groups to imagine what success looks like if they achieve their long term goal. It is essential to focus on this final success. Too often, organizations, define success based on a smaller intermediary goal around a project, rather than the end vision.
One adopts a different mindset when the goal is exponential, not linear. Ann Mei notes how Google’s Astro Teller observes a different mindset is needed to improve a car’s fuel efficiency by 10% than by 10x. This difficulty in achieving 10x fuel saving is different than 10%. Quite often in incremental improvements, all the linear options have been tried and already exhausted by other groups. However, with exponential ideas, there are unexplored opportunities and paths.
An example that Ann Mei uses that I like, is to compare starting a dry cleaning business versus starting Amazon. “The difference between these two ventures lies in both certainty and scope”. The dry cleaner has a predictable path forward, yet only can achieve local results. Amazon has an unpredictable path forward but has the potential for enormous scale.
"If both your problem and solution are well understood, you can simply execute on a defined plan. On the other hand, when the solution is unclear and the needs are vastly underserved, we must take greater risk, set higher ambitions, and test multiple alternatives."
Ann Mei says, in the social sector “we are creating a lot of dry cleaners” even though the problems we are tackling have a level of uncertainty and scale more like Amazon.
To summarize: the long term goal must be ambitious and include a plan to impact a sizable number of those affected by the social challenge.
It is important to highlight, that Ann Mei suggests a single organization can't achieve many of the most ambitious long term goals. Often success requires multiple groups working together, new policies, and new markets.
So, how does one identify the right solution?
2. Start Small
Although the goal should be long term and ambitious, the approach to solving it should start small and progressive. This is because the winning solution is unknown. It also makes it possible for small groups to start testing ambitious hypothesis, with minimal resources.
If the solution to solving a goal were obvious, it likely would have been already accomplished. However, the social sector often comes in because both the government and the private sector has failed to date. [I’d add this means, the private sector, to date, has failed. It will be required later on we’ll see likely for the success of the intervention.]
The job of the social impact enterprise is to test if it can deliver value on the three critical hypotheses that must be proven with evidence in every social impact organization.
1. Value Hypothesis: can you offer a product or service that people really want? Do people embrace the intervention and recommend it to all their friends? Or does it cause confusion and disinterest?
2. Growth Hypothesis: is the intervention able to achieve massive economies of scale. Is there a demonstrated route forward of how it will grow to meet the need.
These first two hypotheses come from Eric Ries’ original book, The Lean Startup. Most of his book revolved around explaining what these are and how to test these hypotheses. I recommend reading that book, or my blogpost summary, for the background on this.
The third critical hypothesis to test, and Ann Mei Chang’s new addition to Eric’s work, is adapted for organizations who do not measure success via profits but via social impact. That hypothesis is
3. Impact Hypothesis: the hypothesis in brief means: “does it work?”. Does the intervention achieve the desired social impact? But, in addition to this, does it do this in the most efficient and effective way possible
The job of the social impact organization is to solve these three hypotheses: value, growth, and impact. The process to test these hypothesis follows the following simple four-step process:
2.i. - Identify assumptions
Most ventures actually are built entirely upon assumptions. Recognizing this is the first step to success. An assumption likely involves question such as, “what must go right for your product or service to work? What could go wrong and cause this to fail?”
2.ii. - Generate a hypothesis and test each assumption
Each assumption needs to be broken down into one or several clear hypothesis. A hypothesis is a statement that can be tested with an experiment, and generate data that is able to validate or refute the hypothesis. The metrics for testing the hypothesis must be determined before the experiment.
Each hypothesis is tested via a build, measure, learn feedback loop. “The most critical indicator of successful innovation is the speed we can run through the cycle”. This approach is different than the typical program based grant, which takes 5 years to complete. In Ann Mei’s opinion, traditional grants focus more on “preventing fraud” rather than learning what is most effective.
“Your solution should evolve. Not from debates in a conference room but via data from the actions and behaviours of real customers.”
For instance, a group helping provide training for youth may measure the cost of training per job seeker, the cost of training per job matched, and the retention rate of the youth as employees. (clear measurable outcomes). They may have a hypothesis that a different training schedule makes it possible for candidates to stand longer in hospitality jobs, and in turn have higher retention in their job. This is a clear hypothesis tied to their core metrics. It is possible to test the hypothesis now and see if the new training schedule works.
2.iii. - Validated learning
The goal of testing assumptions via hypothesis and MVPs is to reduce the risk associated with each assumption, by trying to determine which are valid and which are false. This creates learning, which enables a social enterprise to invest further resources into those approaches that have shown to be most effective.
Eric Ries in Lean Startup, like to point out that activity without a hypothesis, is pointless. One cannot learn if one cannot fail. Without a hypothesis to test, it is impossible to know if the outcome from a set of activities was what is intended.
2.iv. -Pivot or persevere
The last step is to evaluate the progress that has made testing each hypothesis and assumption. Is the organization getting closer to discovering the correct path for its value, growth, and impact hypothesis? If so, it may continue to persevere. But, if the organization has achieved sub-par results, and is unable to improve with the current approach, the organizations may decide it is best to use the accumulated learning to ‘pivot’ and attempt a new path.
The goal isn’t merely “to do some good”, but to be able to achieve massively scalable solutions that have a plausible path for helping many with a need. This leads to the third sections: relentlessly seek impact.
3. Relentlessly Seek Impact
When an organization shifts its goal to achieve maximum social benefit, it has to put aside its “vanity metrics” and “tactical metrics” in place of innovation metrics.
Vanity metrics (or tactical metrics) speak about, how many people were reached, and the total number of dollars raised. But such metrics speak little about (1) if people want the thing, (2) if it is growing, and (3) if it works.
Innovation metrics demonstrate how your intervention is better than the approach of other organizations. It explains to the organization and its funders why a particular intervention can reach more people at a lower cost, or provide a better intervention. “If you can’t do it better than the other groups out there, you shouldn’t be competing for scarce funding and attention.”
The two metrics of social impact innovation include the
1. Unit Cost: cost per unit of intervention
2. Unit Yield: social impact return per unit of intervention
For an intervention to have massive scale, it must achieve lower unit costs or higher unit yields. When this happens, exponential growth is possible.
When Ann Mei moved from Silicon Valley into social impact organizations, she started asking questions that were routine in her prior work, such as “How well are these methods working? What improvements have been made to the grant since it was designed in an office oversees years before?” Could more good be done with the same funding? In your geographical area, are you delivering the best possible solution per dollar spent
To her dismay, most organizations did not consider their work through this rigourous lens.
4. Funding
To over-summarize a large section of the book, funding should be directed towards helping projects that have the potential for massive imapct, understand if the interventions they propose (1) are valuable, (2) are able to grow, and (3) have true impact.
The right funding model for this is focused on funding the anticipated need for a problem, not focused on funding a specific grant. It also means that flexible funds will be required for an organization to iterate on improving its ideas and hypothesis.
This new funding model changes the relationship between the donor and the recipient. In the new model, the goal for both the donor and the recipient is to figure out 'what works'. In the old model, the goal of the recipient is always to show the donor that their funding is "is working" and potentially downplay problems with what is being funded.
See further the separate post: Why philanthropy & grants hamper innovation
5. Engines of Growth in Social Impact
See Further: the my post Engines of growth in social Impact
Conclusion
Ann Mei encourages reads to
- Think Big when considering the social challenges we face and that organizations should propose solutions that if successful, would address a sizable number of people affected by the issue.
- Start Small: the pathway forward is unclear, and by starting small with testable hypothesis, an organization can test each hypothesis using the build, measure, learn feedback loop. It must discover if its intervention is (1) valuable to people, (2) has a strategy for growth, and (3) actually works and has the impact desired.
- Relentlessly Seek Impact: a social impact organization doesn't have profits to prove it is successful at finding the right product-market-fit. Instead a social impact project it must demonstrate through metrics how its activities and interventions are the most cost effective option, with the greatest impact.
To conclude, Ann Mei poses some questions to consider when applying the principles of Lean Impact to your organization include:
- “Does it work?
- What experiments have been run to further expand the social impact?
- Do the beneficiaries love and demand what is being offered?
- How will you reach the size of the population with this need?
- What is in the research pipelines?
- How has cost-effectiveness improved from last year?”
These are the types of questions an organization seeking to follow the principles of Lean Impact will ask itself, and will help it achieve 'success'.
Articles in this series:
- Lean Startup + Lean Impact: Combined Short Summary
- Why philanthropy hampers innovation
- Engines of growth in social impact
- Lean Impact: How to Innovate for Radically Greater Social Good (detailed summary) (this article)
Other articles you may like:
- The Lean Startup: Key themes from the book by Eric Ries
- Pair Together: Design Thinking, Lean Startup, & Agile
I highly recommend reading the full book Lean Impact: How to Innovate for Radically Greater Social Good, by Ann Mei Chang yourself.
Please note. The quotations used in this blogpost are jotted down while listening to the audio version of the book. They may be slightly inaccurate or accidentally missing. Please contact me with any revisions to the accuracy of these quotes and the original text.