AI for good: how technology can help society’s most vulnerable

In his second inaugural address, Franklin D. Roosevelt said: “The test of our progress is not whether we add more to the abundance of those who have much, it is whether we provide enough for those who have little.”

With this in mind, in this blog I’m going to tell you the story of FareShare. Here’s how the charity, together with Peak, is making sure that AI and Decision Intelligence do not only benefit those who have, but also those who need.

My hope is that, by the time you have finished reading this, a seed of imagination will have been sown, a spark of inspiration lit – and that more people will join the journey towards building Decision Intelligence solutions for good.

In November 2021, Peak hosted its first Decision Intelligence summit, AltitudeX, in Manchester. I had the honor of sharing the stage with the amazing Miranda Kaunang (Head of Development) of FareShare Greater Manchester (FSGM). If you don’t know who FareShare are, you may have heard about their work via the impactful campaign by Marcus Rashford to provide food for families in need during the summer holidays and the 2020 and 2021 lockdowns in the UK. In that case, you will have certainly seen pictures of the Manchester United forward in action at the FareShare warehouse in Manchester, the main benefactors of his campaign.

In her talk, Miranda told us their story. FareShare set out on a mission to reduce waste by redistributing good quality food that would otherwise have risked being thrown away. They’re making this dream a reality by taking on the role as the middle party between food donors and charities. In FareShare, companies that want to provide food have a go-to organization to donate to, while those charities in need of food have a dedicated partner on hand to support them.

Working with more than 350 frontline charities and community groups, and with the support from donors and volunteers, FareShare Greater Manchester distributed millions of meals to school breakfast clubs, older people’s lunch clubs, homeless shelters and community cafes in 2019 alone. These figures reached even greater heights in 2020, which we will cover shortly, due to “the Rashford effect.”

Making such an impact on your local community does not come without its own set of challenges, however. Surplus food is donated from the food industry, and not only is the frequency and magnitude of donations irregular, but the very items received are unknown until they show up at the warehouses. Coordinating deliveries under these conditions is a largely manual process, and allocating food in such a way that as much of it as possible gets distributed on time is an immensely difficult task (the stuff of nightmares for most supply chain managers.)

On top of it all, the COVID-19 pandemic came. Not only did it hit our most vulnerable communities the hardest, but it also decimated the very workforce trying to help them – 47% of FareShare volunteers were forced to self-isolate during lockdown. Through the summer of 2020, the number of people accessing food via FareShare Greater Manchester rose from 26,000 a week to 50,000 a week. 

The heroes at FSGM were working harder than ever before, to meet a higher demand than ever before, with fewer volunteers than ever before. In four months, with a volunteering force slashed in half, they dispatched 971 tonnes of food; the equivalent of 2.3 million meals. As impressive as these figures were, there was a real need to change how they operate, in order to ensure that as much food as possible was distributed in the face of such challenging conditions. Enter Marcus Rashford.

The England footballer’s involvement kick-started a massive nationwide campaign that captured the hearts of the country, from Greater Manchester Mayor Andy Burnham to the royal family, to your local supermarket. The support was overwhelming. At Peak, we started calling this effect “the Rashford effect,” and its impact can be observed in the plot below.

The Rashford campaign resulted in an enormous increase in donations, with the vertical line marking the start of the campaign. After a summer of crazy volatility, the campaign subsided somewhat, as all campaigns naturally do, but not without effect. Once the donations were relatively stable, they rested at a level almost twice as high as before – most likely thanks to the raised awareness of the fantastic work that FSGM does. The difference? 52,000 meals. Every. Single. Week.

Raising awareness like this can sometimes yield unexpected results, or put you in touch with unexpected people. In this case it was Peak and FareShare. Geographically, we were just 30 minutes down the road, but in another sense, we were worlds apart. Our journey together started out with a Peak team charity walk, when we collectively hiked the mammoth 813-mile distance from Land’s End to John o’Groats to raise money and awareness for FareShare’s cause. From here, our story evolved to become what it is today: a mission to leverage Decision Intelligence for good.

With a burning desire to fulfill the great promise of Decision Intelligence, the data science team at Peak was presented with FareShare’s challenges. Thankfully, ideas came flowing with the intensity of a river in the Mancunian autumn rain. A key skill for any data scientist is to take a seemingly complicated problem, understand it, and then break it down into digestible components to solve.

With FareShare, the same can be done simply by realizing that it operates like any other business – just one with a different set of KPIs (key performance indicators). If a regular company depends on revenue, FareShare depends on food donations. Where a regular company would reduce their inventory if they frequently generate a lot of waste, FareShare would rather distribute this food for free to those people in need. It’s the same processes, the same levers to pull, only seeking a different outcome. The conclusion? If Decision Intelligence can help profit-driven companies, then it can help nonprofits too.

Decision Intelligence for good

It’s a bold promise, and those sort of promises are nothing without action. Just like how the partnership started with a hike, we once again find ourselves ready to walk the walk, rather than just talk the talk. So, what are we actually doing?

With ideas abound and plenty of people itching to add their contribution to the project, we decided that building a Decision Intelligence toolkit was the best approach. The idea is to build any and all solutions that would enhance decision making at FareShare. In the first phase of the project, we’re focusing on donations. Two solutions are currently being developed in this area:

1. Personalized impact reports, generated at scale

Suppliers will sometimes receive impact reports outlining how their donations have been used, where they’ve ended up and the impact they’ve made. Producing these reports and conducting the analysis, of course, takes time; time that an already-stretched FSGM team simply doesn’t have. 

Our first solution is a portal, put together by our insights team, that allows the team at FareShare to grab key stats and generate a report packed to the brim with detailed analysis – ready to send to any supplier with a single click of a button. The whole process takes less than 30 seconds, down from days.

2. Donations forecasts

Coordinating deliveries from an unknown quantity of unknown product donations is challenging. The first step to aid in this gargantuan task is to produce a forecasting model, predicting the amount of donations for all the various types of food.

We hope that the benefits are two-fold. With a good estimate of future donations, allocating food fairly will be an easier task. We also intend to use it to spot opportunities to take on more community food members (charities), increasing the reach of FareShare Greater Manchester.

Peak’s data science hackathon

What would a set of tools be without a good toolbox to keep them in, to keep things organized? When our first solutions are developed, we’ll be looking to deploy them on the Peak platform in a structured dashboard – complemented with even more insight and decision making tools.

This stage of the project is being executed in less than two weeks using a format that risks making some of Peak’s other customers green with envy. We call it hyper-delivery. A Peak-wide hackathon, engaging the whole data science team in an intense week of development, will be the culmination of the first phase of the project.

So, that’s our plan for the next three months – with much, much more to follow. I want to finish this post off by referring to two of Peak’s five values: open and responsible.

Open

It’s easy for anyone to write a blog post about doing good without backing it up. For this reason we are asking you to hold us accountable. Check in with us in a couple of months and we will be following up on our progress on the journey we have set out on. 

The FareShare x Peak partnership was made public at our inaugural AltitudeX summit in Manchester, so if you missed the talk, you can rewatch it here.

Responsible

FareShare Greater Manchester is providing a jaw-dropping amount of food to families each year in a herculean effort to help. Despite this, families are still going hungry in the fifth largest economy in the world. How can you help? Helping is easy; it can be as simple as buying a couple of extra groceries to put in the food boxes at supermarkets, or it could be offering up a couple hours of your time to volunteer at your local branch. 

I hope by now that this read has given you a couple of ideas of how the potential of Decision Intelligence is not limited to the revenue streams of large corporations, but can also be used as a force for good. If you want to learn more about the work we are doing just get in touch with me or anyone else at Peak or FareShare and we will be more happy to carry on the conversation.