**4. Conclusions and Future Work**

In this paper, we have described the design of the Active2Gether coaching system in detail. The coaching system aims to encourage physical activity among young adults by combining evidence-based behavior change techniques with elements from modern (mobile) technology, such as location monitoring and model-based reasoning.

The effectiveness of the system is currently being evaluated in a three-month trial with more than 100 participants between 18 and 30 years old. To determine the added value of the tailored messaging, a three-armed design has been chosen. One group uses the full version of the Active2Gether system, a second group uses the Active2Gether system but does not receive tailored messages, and the third group uses the standard website and app that belong to the Fitbit tracker. The participants start with an intake questionnaire that contains questions about their personal situation, their current exercise behavior and perceptions about physical activity. After three months, a similar questionnaire is sent out. The participants are also asked to wear a validated activity monitor in the first week of the intervention and after three months. This allows us to conclude whether using the system leads to a significant increase in physical activity. However, because several behavior change techniques have been employed in the system, it is difficult to identify *which* technique actually influences behavior. In the case a positive overall effect is found, we will further analyze which messages were sent to users to identify the contribution of specific techniques (focusing on specific determinants) on the behavior change. In addition, future research is needed with variants of the system in which only specific components (i.e., social comparison, self-monitoring and goal setting) are functional.

In the current paper, we have discussed the architecture and the implementation of the Active2Gether system. In addition, we have shared lessons learnt during the design, implementation and evaluation of the system, as well as recommendations for further development and improvement. We believe that these insights and the detailed description of the technological choices will prove helpful to designers and developers of healthy lifestyle interventions to produce effective and appealing coaching systems.

**Acknowledgments:** This research is supported by Philips and Technology Foundation STW, Nationaal Initiatief Hersenen en Cognitie NIHC under the partnership program Healthy Lifestyle Solutions (grant no. 12014). The authors would like to thank Anouk Middelweerd, Saskia J. te Velde and Aart van Halteren for their contribution to the conceptual design of the Active2Gether system.

**Author Contributions:** M.C.A.K., A.M. and J.S.M. conceived, designed and implemented the described system. M.C.A.K., A.M. and J.S.M. wrote the paper. All three authors contributed equally to this article, and are therefore in alphabetical order.

**Conflicts of Interest:** The authors declare no conflict of interest.
