Lack of engagement or failure to sustain engagement over time

This is a common challenge for all collective intelligence initiatives and may be caused by practical considerations such as contributions or tasks that are too difficult for volunteers, or inappropriate incentives for people to take part in fully.

• Standardize tasks and publish examples to make contributions easy.
• Ensure technology is appropriate for the location and participants.
• Gamify the experience so it is easy and fun to contribute.
• Tailor incentives to differing profiles and motivations.

Low motivation to contribute due to other, more pressing, issues and concerns or participation fatigue

Some issues associated with climate change may seem irrelevant or disconnected from the main sources of stress in the daily lives of target communities.

People become frustrated after being asked to contribute multiple times, without evidence of concrete benefits from their engagement in the past.

• Co-design initiatives and processes with the people taking part to make sure they respond to their interests.
• Communicate the interconnectedness of climate adaptation with health and economic sustainability.
• Share data with and communicate outcomes to people so they can benefit directly.
• Provide financial incentives to enable contributions from a wider range of people.

Hopelessness and perceived lack of efficacy of climate action

Only a few studies focus on the psychological impacts of climate change on people in the Global South, particularly where climate change is already causing disruption to homes or livelihoods. Early evidence suggests that the negative emotional impact of this experience may lead to mental health challenges that decrease motivation to engage.

• Draw on local, traditional and Indigenous environmental management solutions.
• Work with local NGOs and community leaders to broker relationships, helping to build trust, understanding and resilience over time.
• Communicate benefits and positive outcomes associated with collective action to participants.

Lower participation from certain groups e.g. women, older people, people with disabilities

In some parts of the Global South such as sub-Saharan Africa, uptake of technology such as mobile phones and use of the internet is higher among working-age men than women and older people. This means that collective intelligence initiatives may struggle to engage these groups. These disparities can be difficult to quantify because collective intelligence initiatives often don’t report participation data disaggregated by gender, age, disability, etc.

• Choose technology that already has the broadest reach within target communities and specific groups, e.g. SMS.
• Ensure technology is accessible, e.g. use pictograms to capture data and provide training on how to use devices.
• Use storytelling, drawing and other creative assets like 3D maps to make complex climate concepts more accessible to different groups.
• Design initiatives to fit around the responsibilities and lives of target groups e.g. women, older people.

Small datasets or “patchy” data

Some collective intelligence projects collect datasets that are limited to hundreds of datapoints making them unsuited for larger scale data analysis. And even when data is collected at scale it can be “patchy” meaning that there are gaps or inconsistencies in time or missing locations. This is particularly common for biodiversity data or monitoring of environmental variables like air pollution and water quality.

• Supplement small data with insights from qualitative methods or other data to understand climate adaptation issues in context.
• Apply emerging machine learning and statistical techniques for low-resource settings that can handle small-scale datasets.
• Be more strategic with small scale data collection to target known evidence gaps.
• Use statistical techniques that can model distributions based on incomplete data to fill in the blanks.

Messy and inconsistent data protocols

Collective intelligence initiatives often reinvent the wheel when designing their data collection process. This makes it difficult to integrate datasets between different projects working on the same topic, which limits their impact.

Collection of qualitative insights or data about preferences and attitudes towards climate issues is often more ad-hoc. Data tends to be collected as free text, which becomes increasingly difficult to make sense of when participation numbers are high.

• Embrace messiness when using data for storytelling and advocacy.
• Standardize data collection protocols between projects to build larger and more consistent datasets.
• Use language models to cluster qualitative inputs and identify patterns in unstructured datasets.

Concerns about data quality

Studies comparing citizen generated data with data collected by experts in citizen science have shown comparable quality in several citizen science environmental monitoring projects. But concerns persist among decision makers and institutions, particularly when there is no official guidance about how and when to use non-traditional data sources.

• Integrate machine learning algorithms that authenticate data submitted by participants into data collection tools.
• Use prizes to reward the highest quality contributions.
• Use data for specific purposes where it is “good enough.” For example, using citizen data to identify when thresholds are surpassed or new hotspots of unusual activity as a tool for targeting official data collection.

Decision makers (public and private) fail to act on citizen generated data.

A lack of technical skills amongst decision makers can make it difficult for them to interpret data and models – especially when issues are complex and intersecting, for example health policy makers that need to interpret how environmental variables affect the risk of infectious disease.

The data collected by citizens and outputs of models does not always match the specific interests of policy makers or comes at the wrong time in the policy cycle.

• Establish early connections between decision makers and communities to ensure data is relevant to policy gaps.
• Ensure collective intelligence designers understand the policy process and receive advocacy training to cater their comms to policymakers.
• Create and disseminate advocacy and communication materials as stories are often more convincing for decision makers.
• Develop technical skills training and simple explainers for decision makers. Make data visual and tailor it to relevant policy questions.
• Remain realistic about what is in the collective intelligence designer’s sphere of influence and ultimate control and design accordingly. Plan which decisions to target based on timing throughout the year, i.e action targeting budgetary decisions needs to be taken well in advance to fit the calendar of when such decisions are taken.

Data from collective intelligence isn’t shared directly with the people contributing.

• Involve people in setting the research questions and collect data on issues that directly impact their lives.
• Plan how feedback will be delivered and communicate results from the outset. Use communication channels that already have widespread use and work with trusted community champions.
• Give people access to the data that is relevant to the decisions they’re making. Integrate this into digital tools so individuals receive real-time information tailored to their needs.