Upstream: How to solve problems before they happen, by Dan Heath

Buy the book here!

DIGEST THIS! THE ESSENTIAL IDEA:

Heroism is revered worldwide, a phenomenon deeply rooted in mythology through to the present day. The need for a “Save The Day” hero, however, is a sign of something gone wrong, a sign of failure not success. It’s high time to evolve our approach – to give greater applause to a “quieter breed of hero” – those invisible heroes who annihilate problems before they even occur.

Given the current impact of Coronavirus, the sentiments of this book are needed now more than ever.

TOP TAKEAWAYS

• The need for a hero is a sign of system failure.
• Focus on changing the system.
• Problematise what has become normal, what has become acceptable.
• Detect problems before they arise by addressing early warning signs.
• Use key leverage points for greatest impact.
• Systems are complex –expect unexpected reactions.
• Implement ongoing feedback for success.
• Upstream work is limitless – you can always go further.

WHAT IS UPSTREAM WORK?

• Upstream work detects problems before they occur.
• It is preventative. It is proactive.
• It focuses on early warning signs, flagging the onset of a larger problem if unheeded.
• It involves systems thinking – looking at the systems behind problems and fixing these.
• Upstream solutions are broader and slower but when they work, they achieve long lasting results.
• Examples of upstream solutions include swimming lessons to prevent drowning, vaccinations to prevent disease, and visible police presence to prevent crime.
• Upstream is a direction … you can always go further. Swimming lessons are further upstream than life saving buoys.  Police presence is further upstream than burglar alarms.

WHY IS DOWNSTREAM WORK MORE POPULAR?

• Downstream work is reacting to problems. It is putting out a fire after it has occurred, it is treating for diabetes after it has developed.
• It is favoured because solutions are tangible, easier to measure, and short term.
• In solving a problem reactively but successfully – e.g. putting out a fire that saves lives – a saviour hero is created. And heroism is addictive. Everyone wants to be the hero that saves the day. However, the need for heroism is a sign of system failure. True heroes stop the fire occurring in the first place.
• Preventing a problem before it occurs is hidden work. Upstream work involves an inherent paradox: how can you measure success for something that did not happen? This defines the Prophet’s Dilemma – a prediction that prevents what it predicts because the prophecy galvanises forces to put in the work to avoid its occurrence. It leads people to erroneously believe that there never really was a problem in the first place.
• Upstream work involves invisible heroes saving invisible victims.
• There will always be a need for downstream work – we cannot prevent all fires from occurring, we can’t stop hurricanes. However, it is time for a new dawn to emerge – it is time to tip the balance in favour of Upstream work. “The world needs … a quieter breed of hero, one actively fighting for a world in which rescues are no longer required”.

BARRIERS TO UPSTREAM THINKING

1. NORMALISATION

• How can you solve an issue where no one sees it as a problem because its seen as normal? Where people are blind to the problem in the first place? This describes problem blindness.
• Solution: There’s a need to “problematize the normal”–to give a problem a name.
• EXAMPLE – SEXUAL HARRASSMENT: In 1975, journalist Lin Foley coined the term “sexual harassment”, giving a name to an issue that had been normalised in the workplace, making it something abnormal and unacceptable, and empowering women by defining a collective experience. This helped to enable a world in which women would no longer need to tolerate or accept such behaviour in order to keep their jobs.
• The second aspect of normalisation is the belief that outcomes are out of our control. “This is just the way it is.” That negative outcomes are natural or inevitable.
• It highlights our passivity and sense of helplessness. Instead of renouncing our power to make a change, we must believe in our agency to make a difference – to stop minimising problems as normal because we mask our power.
• INJURIES EXAMPLE: Sports trainer and doctor, Marcus Elliott, brought a different mindset to the New England Patriots NFL team – one averse to the traditional acceptance of injuries as an inevitable part of sport. Instead, his belief that injuries simply resulted from bad training (something that could be changed), inspired a new individualised approach focused on movement observations, assessments of muscle imbalances and targeted training for those most at risk of injury, leading to a 76% reduction in hamstring injuries following his intervention. His programme is an example of an upstream intervention – by focusing on early risk and warning signs of injuries (problems that had not yet occurred but could), and making interventions (training) he changed the course of direction to a more favourable outcome.

2.LACK OF OWNERSHIP

• Problems aren’t solved without someone taking ownership for solving them.
• “That’s not mine to fix” – a common approach by parties who are actually capable of fixing a problem. Its often those who suffer most from a problem that are left to find solutions. However, they may not be best placed to fix them.
• Taking ownership is particularly important in upstream work, which focuses on preventing a problem that hasn’t yet occurred, where there is little to no attention nor demand for a solution. In these situations, taking ownership is about becoming a visionary, a pioneer, about stepping up to become a leader.
• The question should not be “Who suffers most from the problem?” but “Who’s best positioned to fix it, and will they step up?”
• Taking responsibility for problems often means stepping outside of your immediate sphere of influence.
• CAR SEAT SAFETY EXAMPLE – In the 1970s, Dr Bob Sanders – a paediatrician in Tennessee – stood up and took ownership by answering the call from an article for paediatricians to advocate for child car seat safety. The article called for the widening of the realm of paediatricians beyond diagnosing and treating illness to advocacy and lobbying – undertaking leadership of a problem that wasn’t within their traditional remit. More young children were injured and killed within vehicles than outside, and the authors saw paediatricians – possessing the authority to positively impact outcomes by alerting parents to the dangers of children riding “loose” in vehicles, and advocate for change more widely  – as best placed to fix the problem, seeing no difference between restraints and immunisation as weapons in the preventative medicine toolkit. Following intense lobbying efforts, in 1978, Tennessee became the first US state to require car seats for children under four (with a subsequent repeal of a loophole in 1981) and by 1985 all 50 states had passed child restraint laws. Estimates indicate that 11,000+ children’s lives were saved by car seats between 1975 and 2016.
• Sometimes we deny our own ownership of a problem, requiring someone else to fix it, something external to us, and thus giving away our power.
• MOVE MY CHAIR EXAMPLE We’ve all been in situations when someone sitting in front of us blocks our view, we shift a little, and the person in front mirrors our action, creating an ongoing dance causing huge irritation. But we forget that we hold power in our hands – we can simply get up,  move our chair and end our frustration. What other irritating situations can you apply the “move my chair” mindset to, and take control of a situation for a more positive outcome? What if you were to tell a story of a frustrating situation as if you were the only one responsible for the outcome and move from a victim mindset to (co)-owner of a solution? (This does not apply to abusive situations which would lead to victim-blaming”).
• Sometimes, “That’s not mine to fix” is an issue of legitimacy. People are motivated to step up to help fix a problem, but do not feel it’s their legitimate place to do so as they haven’t suffered directly e.g. a young man is concerned about the high levels of date rape on campus but feels it may be inappropriate for him to join protests led by women.
• Solution: where appropriate, provide a sense of legitimacy by making it clear all groups can play a part in the solution e.g. simply by changing the title of a protest to include references to both men and women.

3. TUNNELLING

• Juggling multiple problems can lead to tunnel vision – there isn’t enough bandwidth to solve them all.
• It results in short term, narrow and reactive thinking – the opposite of systems thinking and preventative solutions which form the basis of upstream work.
• When resources are scarce, every problem becomes a source of stress – “life becomes a tightrope walk” -leading to reacting to fire after fire after fire, without the breadth of space – the bandwidth – required to prevent them in the first place. This explains the spiral of poverty. Time can also have this effect.
• Tunnelling leads to more tunnelling –if you can’t solve problems by treating the system, you are kept in an endless cycle of reaction.
• TUNNELLING NURSES EXAMPLE A study showed that nurses solved unexpected problems every 90 minutes on average – they were professional problem solvers. However, their creativity and efficiency, signs of a “good nurse”, masked the tracking of these problems, creating “a system that never learns”. Problems included having to borrow towels from other departments to cover a shortfall, and repeatedly having to find security tags required for new-borns to be discharged. The nurses were tunnelling, focusing on short term reactions to problems without the formal opportunity to feedback on these issues or space to consider potential preventative solutions.
• How to escape the tunnel? Build in guaranteed time and resources for problem solving – a space for upstream work e.g. some hospitals hold early morning forums where staff can flag regular issues and near misses, and discuss complexities for the day ahead. This could have been the ideal forum for the nurses above to raise issues e.g. security tags falling off babies, allowing for preventative fixes at the systems level.

7 QUESTIONS for UPSTREAM LEADERS

1.HOW WILL YOU UNITE THE RIGHT PEOPLE?

• Upstream work often involves volunteered efforts, “chosen, not obligated”. People volunteer to take ownership of a problem in order to prevent it, so it’s important to motivate people to undertake this work.
• A. Surround the problem with the right people – those close to the problem with the experience, expertise and authority to make a difference. To surround the problem, make sure all of the agencies playing a part in the multi-faceted system are involved.  (Similarly, Rebel Ideas calls for diversity of thought to surround a problem.)
• B. Align people’s efforts towards a shared vision focused on preventing specific instances of the problem e.g. stopping women from being killed rather than discussing domestic violence policy issues or what’s broken. Make the problem personal and real – focus on cases by people and names.
• C. Focus people on the use of the latest data – data for the purpose of learning rather than data for the purpose of inspection. The latter is often target focused, based on penalising those who fall short (which can lead to gaming), rather than learning why and how to improve going forward. When designing a system, ensure data will be useful for those on the frontline, to allow them to learn and adapt, and know in real time whether they are succeeding or failing e.g. such data could be used by teachers to focus more time on areas students are struggling with if they have data that reveals this in real-time. “You can’t solve a dynamic problem with static data.“
• DOMESTIC ABUSE EXAMPLE In 2005, Kelly Dunne, a leader in the anti-domestic violence field, saw that the only way to prevent murder in domestic abuse cases was to unite the fragmented groups having a role in them – police officers, parole and probation officers, victim advocates, hospital staff, a representative from the District Attorney’s office. She surrounded the problem by organising the Domestic Violence High Risk Team, bringing together representatives who had previously been working in institutional isolation and focused their efforts to prevent the deaths of women at greatest risk. They used name lists, reviewing cases one by one (specific instances of the larger problem) – making the problem personal and real rather than abstract and distant, using the latest data to inform their work. “Where was Nicole’s abuser?” “What has he been doing?” How can we help her this week?” Where would she go if she needed to escape? Who would pay for a hotel or taxi?” Collectively the team uncovered and addressed gaps within the system that could be exploited by abusers. Not a single woman was killed due to domestic violence related homicide in the 14 years from the formation of the team. “Not one”.

2.HOW WILL YOU CHANGE THE SYSTEM?

• The ultimate aim of upstream work is systems change for the better – a well-designed system is the best way to solve problems before they occur.
• INVISIBLE SYSTEM EXAMPLE For decades, fluoride has been added to water supplies to prevent tooth decay, a preventative invisible systematic solution that has been named one of the ten best public health achievements of the twentieth century.
• DAMAGED GOODS EXAMPLE – to solve the problem of bikes being damaged on delivery, VanMoof added images of flatscreen TVs to their boxes, leading to more careful handling by couriers and a 70%+ reduction in damages.
• Upstream work involves fighting against people and organisations who have become used to the system, who tacitly accept its flaws and the status quo, perhaps because in some way they benefit from the system or it’s simply much easier to become resigned to it.
• Courage sparks the start of system change by uniting people around a common cause, but there should never be an ongoing reliance on heroism – the objective is to eliminate the need for courage because change has been instilled within the system.
• Be in it for the long haul – realise that systems change takes time – it took decades or even centuries for these systems to form in the first place.
• Solutions should be systemic, not personal i.e. not reliant or dependent on the judgement of persons within the system e.g. to increase workplace diversity, organisations can systematically remove names from applications, and require recruitment from a wider pool of places.
• An important part of systems change is to give actors involved a sense of their own power to create change, a chance to express their agency, which could initially be in small ways leading onto successes in larger campaigns.
• LIFE EXPECTANCY EXAMPLE – Anthony Iton on moving to Baltimore in 1985 was shocked at Americans who shrugged their shoulders at urban poverty, who had come to accept it as inevitable. As director of the Alameda County Public Health Department, he and his team used data to analyse life expectancy by neighbourhood, something that hadn’t been done by the department before, revealing a stark gap of 16-23 years in neighbourhoods only miles apart. He found that there wasn’t just one or two or three causes of the lower life expectancy, but that it was literally everything – a multiplicity of systemic forces that created communities that were “incubators of chronic stress” due to a lack of control over their lives – over housing, finding good education, avoiding crime, finding jobs, healthy food etc. Lower income communities were being short-changed in terms of their life spans due to the system they functioned within. People with low incomes are not born being physiologically different from those with higher incomes…they are made that way – they are made by the system. The system had been perfectly designed to produce those results. Chronic stress led to ill health and shorter lives. Of course there are exceptions, examples of people rising above their circumstances and the systems they were born into, to succeed. However, badly designed systems lower the probabilities of this happening – in higher income neighbourhoods, where systems were well designed and life expectancy longer, the probabilities were overwhelmingly in their favour. Why should we be indirectly congratulating a system where individuals have to become heroes and overcome the odds just to succeed? Iton went on to focus on giving citizens a sense of their power through political campaigning, to reshape their environments and the systemic causes of their hardships piece by piece, gradually shifting the odds back in their favour. “Greater power leads to policy victories which leads to a better environment”. Successes in a particularly community – Fresno – included a new skateboard park, and opening up of 16 school playgrounds for public use outside of school hours. Between 2010-2018, 321 policy wins and 451 system changes were achieved across 14 communities. “Power works”.

3.WHERE CAN YOU FIND A POINT OF LEVERAGE?

• Systems change is complex and can take decades of effort – so it’s important as early as possible to look for a points of leverage, which can be used to gain and deliver maximum impact towards the outcomes you seek.
• Examples of leverage points include: targeting high impact groups – small groups of people having most impact on a problem, risk factors e.g. smoking, and protective factors that could lower the incidence of the problem e.g. youth clubs.
• Finding leverage points requires immersing yourself in the problem, to understand the issues that contribute most.
• SEPSIS EXAMPLE In aiming to reduce patient deaths, detailed case studies of the last 50 patients that had died at each hospital within the Permanente Medical Group in Northern California, revealed that a third had died due to sepsis. This had been an infection the hospitals had been relatively ignorant to, which consequently led to a focus on patients with sepsis and a 60% reduction in deaths due to this cause.
• EDUCATION EXAMPLE Chicago Public Schools (CPS) “was a system designed to fail half its kids”. In 1998, only 52.4% of its students graduated (the symptom). The mindset of those within CPS was an acceptance of the high dropout rate, and belief that the failure of students was either due to their own behaviour or lack of effort, or root causes beyond their control that were impossible for them to impact–poor families, student trauma, lack of nutrition, inadequate prior education. Outcomes were turned around through the use of insightful research, and problem ownership, which shifted focus onto the attainment of high school freshmen – something they could have an impact on, and a change they could make within the educational system at a key leverage point. The research had shown that achievement of students in their first year of high school was critical to their overall success. By 2018, the graduation rate shot up to 78%, an increase of 25%.

4. HOW WILL YOU GET EARLY WARNING OF THE PROBLEM?

• The aim is to design a “smoke detector” alarm system forewarning you of a problem to come, so you can take upstream action to prevent the problem occurring.
• In some situations, you do not want too many false positives, leading to alarm fatigue where people end up ignoring the alarms. But where the impact of missing a problem would be devasting you may be willing to accept a high rate of false positives.
• 911 EMERGENCIES EXAMPLE In New York City, Northwell Health used historical data to create their warning system, to strategically locate ambulances in areas of highest need. Data revealed that there was a spike in 911 calls on Fridays & Saturdays, during flu season, on July 4^th and New Year’s Eve, and at mealtimes in nursing homes (when caregivers are guaranteed to check on residents and discover something wrong). By pre-deploying ambulances within close reach of key locations at designated times they achieved a response rate of 6.5 minutes in comparison to the national average of 8 minutes.
• SCHOOL SHOOTINGS EXAMPLE Following the Sandy Hook School Shooting, the Sandy Hook Promise organisation researched other school shootings to develop an early warning system focusing on the mental health of potential shooters. Extreme feelings of social isolation, a strong fascination with firearms, bragging about access to guns, and acting aggressively for seemingly minor reasons – had all been warning signs that had been overlooked in past shootings. They launched a training program and anonymous reporting system to encourage students to raise any concerns they had with fellow students – effectively employing students as human sensors, as human warning signals. When this system was adopted by public schools in Pennsylvania 615 tips were received in the first week, and there were 46 suicide interventions, 3 major drug busts and warning of a school shooting threat.

5.HOW WILL YOU KNOW YOU’RE SUCCEEDING?

• Success is more tangible for downstream interventions. The overall aim is restore the situation to the previous state e.g. put out the fire – so success is easily measurable and evident.
• Upstream interventions involve prevention of a problem before it even occurs – so how do you measure their success?
• This difficulty can lead to ghost victories – superficial success which cloak failure.
• 3 types of ghost victories:
  • 1) Assuming success that is not attributable to your work although measures show you are succeeding e.g.  your team is hitting more home runs, but it’s not through your interventions but an external factor – a decline in pitching talent. Every other team in the league is also achieving greater success as a result.
  • 2) Short term measure success but not aligned with the longer-term mission. BOSTON SIDEWALKS EXAMPLE – in Boston, measures used to assess success in sidewalk maintenance actually worked against the longer-term ambition. Research revealed that prior “success” in repairing roads, serviced as a result of phone call requests, masked the fact that almost half of these roads were already deemed to be in good condition, and lower income neighbourhoods with roads in the poorest conditions were being neglected as a result. This was because roads were only being repaired on calls received – which in the main came from richer areas – and not on the basis of need or condition. Success was partially measured by the number of closed calls – and in doing so, it appeared the team had unwittingly been providing an inequitable service. The short measures indicated a ghost victory, going against the overall mission of ensuring walkability for all Bostonians, particularly those communities most in need.
  • 3) short term measure becomes the mission undermining the overall aim (leading to cheating or “gaming” measures). HOSPITAL WAITING TIMES EXAMPLE – an investigation revealed that patients had purposely been left in ambulances in order to achieve a maximum 4-hour waiting time target, measured from the point at which they entered the hospital. CRIME DATA EXAMPLE – some police officers found indirect ways to under-report crimes or downgrade them in order to achieve more favourable crime statistics. This included purposely trying to find holes in a victim’s story with the sole intention of downgrading a crime as serious as rape. When career performance and progress is judged on hitting certain targets, people will find ways to tilt the numbers in their favour.
• Ways to avoid ghost victories:
  • Paired measures – pair quantitative and qualitative measures to ensure real success e.g. for cleaning, measure “success” based on the size of the area cleaned during a period of time PLUS quality checks – spot checks for errors, customer satisfaction.
  • Pre-gaming – devote time to consider how short-term measures might be misused or achieved in such a way that would be deemed misleading.

6.HOW WILL YOU AVOID DOING HARM?

• In our quest to do good, to make the world better, how can we ensure we don’t unwittingly do harm?
• Remember that upstream interventions involving tinkering with complex systems – you should expect reactions and consequences beyond the immediate scope of your particular area of work, and not all of them may be favourable.
• Systems are too complex to be controlled, but they can be designed and redesigned – it’s about learning to dance with them.
• 1. Look at the system as a whole, not just the particular part you are interested in solving.
• 2. Ask – are you intervening at the right level of the system? And what are the secondary effects of your interventions? What will fill the void (of removing something e.g. banning plastic bags)? What will receive less attention due to focusing on the intervention? How easy is it to reverse the intervention if we end up unwittingly creating harm?
• 3. Test small, gain prompt and ongoing feedback, and implement quickly and iteratively – remember that your thoughts and planned intervention is only a theory – experiment and ask others to challenge your assumptions, create feedback mechanisms and measurement systems – and implement the feedback in order to improve.  Success does not come by foreseeing the future accurately, it’s impossible to foresee everything – we succeed by ensuring we have the feedback we need to navigate our way through.
• 4. Check whether there have been other similar interventions and learn from them.
• 5. On the basis of 1-4, make a decision as to whether to stage a full upstream intervention (similar to moving from pilot to launch) If the answer is no or negative to any of the above, or you haven’t created any feedback mechanisms, think carefully before proceeding.
• FALLING BRANCHES EXAMPLE – In New York City, cutting the pruning budget (pruning being an upstream activity) led to a surprising number of settlements due to injuries caused by falling branches. Greater harm was caused by only looking at benefits to part of the system – savings on maintenance – which in fact ended up being paid out in lawsuits.
• COBRA EXAMPLE – during the UK’s colonial rule of India, a British administrator decided to use incentives to decrease the numbers of cobras in Delhi. However he unwittingly made the problem worse – cash rewards for dead cobras created a cobra farming industry, increasing rather than decreasing numbers. Furthermore, on abolishment of the scheme, the cobra farmers released them as they were no longer of value.
• PLASTIC BAGS EXAMPLE– the banning of plastic bags has led to some unexpected consequences e.g. in San Diego, a deadly outbreak of hepatitis A in 2017 was attributed by some to the lack of plastic bags – people who were homeless had been using them to dispose of their waste, and the void led to use of less sanitary alternatives.
• CHARITIES EXAMPLE – Is the solution you are providing really benefiting the people it is meant to serve or those employed by it? A foundation aiming to increase the financial security of those on low incomes through financial coaching actually benefited its workers – everyone in the ecosystem got paid except those  beneficiaries – “they got coached”. The problem was not that the poor lacked financial know-how – it was because they lacked money due to a lack of adequate opportunities – a system that did not favour them.

7.WHO WILL PAY FOR WHAT DOES NOT HAPPEN?

• The cost of downstream work often far outweighs the cost of upstream work, but there is often resistance to pay for the latter despite the popular saying “Prevention is better than cure” e.g. estimates indicate that  for every $1 dollar spent adding fluoride to water supplies, society saves $20 in avoided dental costs.
• Preventative efforts succeed when the problem is prevented i.e. nothing happens. Who will pay for what does not happen? How do we get people to pay to prevent problems rather than paying for reactionary fixes, often at a much higher cost because the problem is far worse by that point?
• Part of the challenge is that many people and organisations do not want to invest in a programme that will provide rewards years in the future – that will pay them back eventually.
• Solution 1: Government funding -Seek private/alternative sources of funding initially, implement a study of the intervention to assess its impacts based on measures agreed in advance, if successful the government agrees to fund the intervention permanently. The government benefits from not having to undertake a big financial risk at the outset, and future funding is underpinned by evidence of the intervention’s success.
• Solution 2:  Create incentives for organisations to adopt preventative measures.
• ACO EXAMPLE  The Accountable Care Organisation (ACO) model allows primary care doctors to group together and share savings achieved by managing patient’s healthcare better, often achieved through proactive solutions – spending more time with patients, monitoring early warning metrics such as weight and blood pressure, ensuring these head in the right direction – and thus reducing the need for more costly downstream measures such as hospital visits.
• CAPITATION EXAMPLE – healthcare providers get paid a flat fee per patient to take care of all of their healthcare needs regardless of the number of interventions required. These payments are risk adjusted e.g. higher payments for an elderly person compared to a 25-year old. The capitation model incentivises upstream approaches such as providing free healthy food for diabetics, thus avoiding more costly downstream interventions. Pairing quantitative and qualitative measures e.g. patient health metrics and patient satisfaction surveys ensures providers do not game the system by providing less services – they receive less money where patients report being unsatisfied or they allow their health to deteriorate.

GOING SOLO, UPSTREAM

• How can you personally move upstream? Upstream thinking is not just for organisations, it’s for individuals too.
• There are lots of things you could invest in – how do you choose?
  • What do you care so much about that you are willing to commit to it year after year, through obstacles and defeats?
  • What problem are you willing to really learn about up close? Macro starts with micro – to help a million people you first need to understand how to help one; if you want to help solve big problems in the world, seek out groups with ambitious visions that have proximate experience of the problem.
  • Could you change the organisation you currently work for and improve the system from within?
• How can you engage in upstream thinking in your personal life, in relationships? Adopt the mindset of upstream thinking, take personal responsibility for issues and believe in your power to solve them – think of the Move my Chair example above.

HOW TO SUCCEED AT UPSTREAM WORK

• OVERCOME THE BARRIERS TO UPSTREAM WORK
  • PROBLEMATISE NORMAL – give a problematic experience a name.
  • BELIEVE IN YOUR POWER TO POSITIVELY IMPACT OUTCOMES How many problems in our lives and in society are we tolerating simply because we’ve gradually given away our power, accepting apathy and the inevitability of negative outcomes as substitutions? Because we have forgotten that we can fix them?
  • TAKE OWNERSHIP – you may not be the one to create a problem, but you can be the one to fix it. Become a leader. Move from “Can’t someone (else) fix this problem?” to “How can I/we solve this problem?”
• BUILD IN TIME FOR PREVENTATIVE THINKING without guaranteed time for thinking about the wider system and preventative solutions you can get stuck in tunnelling – in a cycle of reactionary band aid fixes. 
• BE IN IT FOR THE LONG GAME – it takes time and commitment – years and decades rather than days and months to see the fruit of your labour when undertaking upstream work, due to the broadness of the problem landscape and the complexity of various moving parts and actors within the system. Be impatient for action (change is not based on lofty ideals but is delivered through action) but patient for outcomes.
• HAVE HUMILITY – the humility to learn, to be wrong, to listen and take on feedback, to take on the hard and complex work and not be discouraged.
• CREATE A SENSE OF URGENCY – to start preventing a problem that will result far off in the future unless action is taken now, you need to create a sense of urgency to attract attention and demand for the problem to be to fixed. People and organisations are constantly dealing with urgent short-term problems – planning for speculative future ones by definition is not urgent. To compete with people’s daily concerns, to overcome indifference and the difficulty in convincing people to collaborate when hardship hasn’t forced them to, create an image and use language that captures people’s imaginations, that enables them to understand the world they could avoid through their efforts now. E.g. The term “ozone hole” created visual imagery that helped the public understand the need to take action to prevent further damage to the ozone layer – alike the urgent need to fix a hole in a roof or a boat.
• FIX THE SYSTEM, NOT THE SYMPTOMS Upstream work is concerned with creating change at the systematic level – with changing the processes and rules that govern us and the culture that influences us – for better outcomes.  Downstream work reacts to problems, treating the symptoms, allowing the root cause to remain – a perpetual cycle. At the heart of the upstream approach, if there is a problem or failure, its root cause is within the system, as “every system is perfectly designed to get the results it gets”.
• UNITE THE RIGHT PEOPLE – Surround the problem with those with the knowledge, experience and authority to make a positive impact.
• HUNT FOR LEVERAGE POINTS IN THE SYSTEM to deliver maximum impact towards the outcomes you want to achieve.
• GET UP CLOSE TO A PROBLEM to understand the key issues.
• SPOT PROBLEMS EARLY ON – create “smoke alarms” to spot early warning signs of impending problems and intervene.
• AVOID HARM –make sure you ask the right questions before pursuing an intervention.
• GET STARTED & GAIN FEEDBACK – Don’t obsess about formulating the perfect solution before getting started – ongoing feedback is key. Take ownership of the underlying problem and start slogging forward. You could spend time designing the perfect (untested) intervention and hope for the best, or start with a pretty good solution with numerous built-in feedback loops, that can’t help to get better over time.
• DETERMINE HOW TO MEASURE SUCCESS ACCURATELY to avoid ghost victories and unintended consequences.
• MAKE USE OF THE LATEST DATA – to solve dynamic and complex problems before they occur.
• CONDUCT TESTS AHEAD OF FULL IMPLEMENTATION. The benefits of pre-planning and simulations in emergencies and humanitarian disaster planning include:
  • 1) boosting readiness before the real situation occurs by revealing problems and developing improvements and tweaks to the system (e.g. implementation of learning from logjams in prior hurricane evacuations improved the contraflow process during Hurricane Katrina, saving lives);
  • 2) get stakeholders to know each other before having to work together in a real emergency and to understand the linkages in the system, as  “you don’t want to be exchanging business cards in the middle of an emergency”.

This is a book where a variety of examples are given that are key to a nuanced and deep understanding of the nature of upstream work – buy the book here.