The proliferation of sustainability reporting has been accompanied with stakeholders’ concerns over the credibility of sustainability reports. This worry has resulted in the assurance of sustainability reports and accordingly, professionals have sought to create a new auditable market and academics to understand this phenomenon. Though previous research has shown that assurers and clients jointly decide the terms of the engagement, assurers’ perceptions of sustainability assurance and assurer–client relationships, in which the engagement takes place, have been overlooked. We try to close this gap by conducting an interview-based study which firstly, examines how assurers view and experience sustainability assurance engagements and secondly, investigates how assurers characterize their relationship with clients. We lead 12 interviews with assurers working for either Mazars, BDO, Grant Thornton, or RSM, covering 11 countries worldwide.   The empirical findings indicate that assurers work with sustainability reports that are unbalanced and inaccurate. Particularly, they view the inaccuracy of nonfinancial data as a challenge to the issuance of sustainability assurance statements. We find that inaccuracy is perceived difficult because assurers lack technical expertise and are unpracticed in auditing qualitative information. Further, when assurers cannot verify the disclosed information, they rather terminate the engagement than issue a presumably misleading assurance statement. The termination of an engagement is, however, the ultimate response and until then, assurers collaborate with the client to be able to issue an assurance statement. Hence, collaboration is considered key to the completion of a sustainability assurance engagement. This research contributes to a better understanding of practitioners’ perspectives on sustainability assurance engagements both practice and relationship-wise. We also develop practical implications and future research avenues.

Cette thèse se focalise sur l’amélioration de l’évaluation de performance des processus métiers collaboratifs. Il s’agit de poursuivre l’évolution de la collaboration entre l’entreprise et ses partenaires. Trois niveaux d’abstraction ont été d’abord identifiés : Métier, fonctionnel et applicatif. Ensuite, nous avons développé une approche descendante allant du niveau métier au niveau applicatif. Dans le niveau métier, des différents d’indicateurs clés de performance ont été proposés à travers la méthodeECOGRAI. Dans le niveau applicatif, nous avons proposé un référentiel d’analyse contenant des indicateurs techniques fonctionnels tels que la durée, l’input, l’output, et non-fonctionnels notamment la maturité, le risque, l’interopérabilité à partir des traces d’exécution. Nous avons proposé ainsi un modèle ontologique en vue de capitaliser et enrichir la sémantique de la performance de ces processus. Nous avons proposé un modèle ascendant pour l’agrégation des indicateurs technique au niveaumétier. Le principal objectif de cette agrégation est la corrélation entre le comportement de l’application métier agrégé à partir de l’exécution et l’évolution des indicateurs métiers. Un autre modèle de gestion des événements métiers a été également proposé pour consolider le processus d’apprentissage de notre approche. Par ailleurs, pour assurer la convergence de la performance, nous avons combiné entre la gestion des traces et la gestion des évènements métiers. Cette combinaison permet d’accompagner l’évolution des processus métiers collaboratifs pendant leur exécution. L’accompagnement évoqué avant favorise l’obtention d’un diagnostic sur la performance pour servir à la prise de décision. Cette dernière est liée étroitement à la détection des alertes et particulièrement à l’anticipation des déviations de la performance le plus rapidement possible. Pour valider lacontribution scientifique de cette thèse, une étude de cas a été réalisée sur un processus de création de devis dans le cadre du projet européen FITMAN. / This thesis focuses on improving the performance evaluation of collaborative business processes. It is about pursuing the evolution of the collaboration between the company and its partners. In the beginning, three abstraction levels were identified: Business, functional and application. Then, we developed a top-down approach from the business level to the application level. In the business level, different key performance indicators have been proposed through the ECOGRAI method. In the application level, we proposed an analytical repository containing functional technical indicators such as duration, input, output, and non-functional, including maturity, risk, and interoperability based on execution traces. We have thus proposed an ontological model in order to capitalize and enrich the semantics of the performance of these processes. We proposed a bottom-up model for the aggregation of technical indicators at the business level. The main objective of this aggregation is the correlation between the behavior of the aggregated business application from the execution and the evolution of the business indicators. Another business event management model was also proposed to consolidate the learning process of our approach. Moreover, to ensure the convergence of performance, we have combined traces management and business event management. This combination allows to accompany the evolution of the collaborative business processes during their execution. The aforementionedaccompaniment favors the obtaining of a diagnosis on performance to be used for decision-making. The latter is closely linked to the detection of alerts and particularly to the anticipation of deviations in performance as quickly as possible. To validate the scientific contribution of this thesis, a case study was carried out on a process of creation of quote within the framework of the European project FITMAN.