Why timing matters when it comes to behaviour change
Changing behaviour is one of the toughest challenges in marketing. People are more likely to change their behaviour when a message reaches them at a moment when they can actually act on it. AI and contextual data can help you identify that moment more accurately. In this blog, we use an energy campaign to show how this works, which AI applications can help, and how you can start using AI to drive behaviour change yourself.
Focus on relevant moments
Reaching people is not the same as changing their behaviour. Reach and visibility are important campaign objectives, but they do not automatically lead to action. The message, context and timing all play a role too. That is why behaviour change starts with different questions: what behaviour do you want to encourage? When are people able to act on it? And which data can help you identify that moment?
Behaviour change in practice: using electricity more intelligently
A campaign focused on the energy transition shows how this works in practice. As we move towards more sustainable energy use, we are consuming far more electricity than we used to. This is partly due to the growing use of technologies such as heat pumps, electric vehicles and induction hobs. As a result, pressure on the Dutch electricity grid is increasing. Grid congestion occurs when more electricity is being consumed or fed back into the grid at the same time than the network can handle. Expanding the grid can help, but smarter energy use can also reduce these peaks.
And it can start with small changes. When the sun is shining brightly, there is plenty of solar energy available. By, for example, running the washing machine or charging an electric car at that time, people can shift their electricity use to a more favourable moment. But how do you change a deeply ingrained habit? By delivering the message when taking action is both possible and logical.
Contextual data determines the moment to act
Network data showed which regions were experiencing the highest pressure on the grid. Whenever temperatures in those areas rose above 15°C under sunny or partly cloudy conditions, a targeted message was activated: ‘The sun is shining. Use electricity now.’ The message was not shown to everyone, all the time. It was only activated in the regions and at the moments when people could take action straight away. Combining grid, location and weather data made this targeted activation possible. It was one of several ways AI and data were used throughout the campaign.
How AI solves concrete campaign challenges
What sets this approach apart is not how much AI was used, but how it was used. At every stage of the campaign process, one question came first: does AI solve a specific problem here? If the answer was yes, it was used. If not, it wasn't. The result was a series of applications, each addressing a different pain point:
Producing two commercials in one day was too expensive and time-consuming.
An AI layer applied to the original footage eliminated the need for costly post-production, making it possible to produce both commercials in a single day.
Customer service was receiving the same questions every day.
An AI version of a customer service representative answered the most frequently asked questions in short videos. This reduced the number of phone calls about those specific questions by 50%.
Website visitors were searching for answers but couldn't find them straight away.
An AI-powered search function provided an immediate answer instead of a list of links, with a simplification button for users who preferred clearer, easier-to-understand language. It now serves more than 200 users a day.
A generic message about saving energy is ineffective if someone can't act on it at that moment.
By connecting weather data with grid data, the campaign only appeared when the sun was shining and grid pressure in the region was high.
Campaign content quickly becomes outdated as the news cycle moves on.
An AI agent scanned news sources every day for relevant developments and new search terms, keeping the campaign up to date without the need for manual monitoring.
Dozens of channels, each with their own definitions of reach and results.
By bringing all the data together in a single model and connecting it to an AI model, manual reporting was eliminated. This allowed specialists to focus on analysis and optimisation instead.
Results: more conscious electricity use when the sun is shining
An impact assessment showed a change in reported energy behaviour. 18% more people now consciously use electricity when the sun is shining. 21% shift their electricity use to sunny periods. The campaign reached 95% of adults in the service area. In total, 5.8 million people saw the campaign at least once.
Five principles for behaviour change through AI
The approach used in this energy campaign can also be applied to other behaviour-change challenges. Whether you work in energy, retail, healthcare, financial services or the public sector, these five principles can help you use AI in a targeted way to drive behaviour change.
1. Define the behaviour you want to encourage
Start by getting clear on the behaviour you want to change – and why and when you want it to happen. Be specific about the action you want people to take, such as charging their electric car when solar energy is readily available.
A clear behavioural objective gives your campaign direction and makes its impact easier to measure.
2. Find the right moment
A message is far more likely to prompt action when it reaches people at a moment when they can actually do something about it. Identify the contextual signals that can help you spot that moment, from behavioural and customer data to location and weather conditions.
The goal isn't to communicate more. It's to communicate when your message is most relevant.
3. Map your campaign journey from start to finish
Look at your campaign process end to end. Where are teams spending too much time on manual tasks? Where could additional context improve decision-making? And where is analysis slowing things down?
Once you've identified these pain points, you can assess where AI can make a real difference.
4. Bring the right expertise together
A data-driven behaviour-change campaign calls for close collaboration between marketing, data, creative and customer service. Bring these disciplines together from the outset to make sure insights, strategy and execution reinforce each other.
AI can support and accelerate the work, but people remain responsible for the strategy and the decisions that shape it.
5. Test, measure and scale
Start with one clearly defined use case and decide upfront how you will measure its impact on behaviour. Test your approach, learn from the results and refine it as you go. Once you've established what works, scale it up.That way, you're optimising for meaningful behaviour change – not simply for reach.
