Designing dissemination strategies to maximise stakeholder engagement

Abstract Policy‐focused research projects must engage with an array of stakeholders to ensure that policy recommendations are relevant to a large proportion of the population and align with the needs of organisations and policy‐makers responsible for their implementation. Strategies of stakeholder engagement and dissemination are key to any research programme, particularly in e‐inclusion where stakeholders are fragmented. This article uses examples from the Bridging Research in Ageing and Information and Communication Technology (ICT) Development (BRAID) project, funded under the Seventh Framework Programme of the European Commission, to examine the effectiveness of different stakeholder engagement strategies and make recommendations for future policy‐focused research projects. The BRAID project findings indicate that certain engagement strategies are more successful that others at integrating particular stakeholders. Based on this, we recommend that projects evaluate the types of stakeholders they are successfully reaching, design their dissemination strategies accordingly, and use a variety of engagement activities to ensure that all stakeholders are represented.


Introduction
Strategies of stakeholder engagement and dissemination are key to any research programme seeking to influence organisational, social or political policy. This is particularly true for e-Inclusion, where stakeholders are inherently fragmented. The purpose of the European Commission-funded project, the Bridging Research in Ageing and Information and Communication Technology (ICT) Development (BRAID) project (2010-2012), was to create a roadmap for ICT development that responds to demographic, economic and health challenges associated with ageing and the ICT sector in Europe and internationally, including the fragmentation of stakeholders and research efforts.
As opposed to other projects focused on particular aspects of ageing, which also experienced fragmentation issues, the BRAID consortium recognised a need to more broadly address the larger goal of ageing well through three main objectives: create a dynamic ICT and Ageing roadmap to address and analyse older people's needs, develop a research strategy capable of evolution over time, and expand the stakeholder network of contacts [1,2].
Some have cautioned that engagement strategies must address a key assumption that simply bringing together diverse stakeholders does not necessarily result in improved policies and practices [3] and often consideration is not given to how to best to integrate different types of stakeholders and what methods are most appropriate. BRAID engaged a diverse array of stakeholders by combining in-person workshops with

Identifying the stakeholders
Stakeholder identification and categorisation are critical to ensure successful engagement with representatives of all relevant groups. As the BRAID project evolved, the original four categories [6] were regrouped into six categories, including policymakers, academics, civil society organisations, industry, end users and the media, following Finn and Wright's [7] exploration of successful coordination mechanisms for different stakeholder categories. The stakeholder categories, their relative roles, and the means of engagement for each particular group is described in Table 1. Unsuccessful -There are not any recorded instances of media professionals attending workshops N/A Workshops do not often engender press attention as being part of a "good news story".

Challenges
All categories Good -The website had thousands of hits and required extra bandwidth in April 2011 and the newsletter had more than 3500 subscribers.
Top hits from Third countries included hits originating from Australia, USA, Ukraine and China The website and contact list were not designed to differentiate between different categories of stakeholder.

Forums and discussion boards -The use of forums and discussion boards required a low level of involvement and passive engagement by participants
All categories Unsuccessful -Although contact list members and newsletter subscribers were made aware of the BRAID virtual forum, it did not receive any posts.
N/A Difficult to incentivize stakeholders to engage with the virtual forum between workshops.
Wiki -The use of the wiki required a high level of involvement and a variable level of engagement ranging from passive (accessing content) through to active (contribution of original content) All categories Unknown -The BRAID wiki has been receiving contributions from project partners but it is too early to evaluate its use by other stakeholders.
N/A Difficult to incentivize stakeholders to contribute to the Wiki.

Policy makers Unknown
The BRAID conference, as with many conferences, is geared more towards professionals and experts, not end users. The final conference registration information is not yet available for examination.

Academics
High -A high proportion of abstracts submitted were from academics

Press releases -Press releases require a low level of involvement and passive engagement by participants
All categories Excellent for generating interest in / downloads of deliverables. Press releases were carried by a number of other associations.
How this interest was differentiated by stakeholder type is unknown. The project brochure was not prepared until the preliminary results of the BRAID project were ready to be circulated. As such, it was only able to be distributed relatively late in the project. Second, the nature of the distribution meant that the consortium was not able to keep track of which types of stakeholders were presented with the brochure, since it was given freely to anyone interested in the project.

Academics
Good -Academics attended events at which the brochure was circulated. Industry Good -Industry representatives attended events at which the brochure was circulated. CSOs Good -CSO representatives attended events at which the brochure was circulated.

End users
Fair -Some end users attended events at which the brochure was circulated. Media Unknown 1 Throughout this table, we use "end users" as shorthand for primary stakeholders including end users, informal carers and family members as described in section X.X. 2 Workshop attendance figures are courtesy of BRAID 5.2.

Stakeholder engagement and dissemination strategies
Planning for the dissemination of knowledge not only involves understanding where and when but also what should be communicated and how it should be presented to different audiences [8]. Therefore, a stakeholder engagement and dissemination strategy should be based on a sound understanding of the stakeholders to be targeted and the coordination mechanism(s) most appropriate to optimally reach particular stakeholder audiences.
Stakeholders are often varied and heterogeneous, with different levels of interest or power. As such, Wright and Cairns [9] present a framework for this, positioning stakeholders in a grid along an x-axis indicating interest and a y-axis indicating relative power. This framework is useful for understanding where stakeholders might be located in terms of interest and power, and it provides a way to understand how to impact particular groups of stakeholders, or even specific individuals.
The BRAID plan was developed to determine the methods by which their issues -embedded in broader societal issues including policy, e-inclusion efforts, employment, community care, health care, standards, insurance and business models -were to be communicated between the project team and to the various, aforementioned stakeholder groups. The overall aim of the stakeholder engagement and dissemination activities was to ensure stakeholder involvement was an integral part of the process, and at different phases, this involvement took the form of sharing information, consulting, engaging in dialogue or deliberating on outputs.
Engagement and dissemination strategies can range from "passive" to "active". In the former, stakeholders are recipients of information without active engagement, whereas active strategies require commitments of time to information products, partners or other stakeholders. Likewise, stakeholder involvement can range from "low" to "high". A low level of involvement may simply take the form of receiving and reading final reports. High-level involvement may involve regular community activity or attendance at workshops.
The BRAID strategy used passive activities with low involvement, including newsletters or website access, but established pathways for increased involvement through a project Wiki, surveys, commenting on deliverables and participation in workshops and interviews. This range made dissemination scalable, meaning hundreds of users could remain passive recipients of information but could easily increase their active involvement via the project's community portal.
The engagement and dissemination activities were designed to permit the formulation of a baseline and monitoring of trends, develop BRAID's roadmap and impact, link all of the individual work, and address the levels of stakeholder engagement. The stakeholder involvement techniques provided feedback and engagement pathways and formed a method for disseminating the results.

Stages of dissemination
There are many communication tools for particular audiences and these may change over time. The BRAID dissemination strategy used multiple tools that are listed in Table 1. Using multiple methods and producing different forms of information where appropriate increased the spread of the research and knowledge. Additionally, appropriate methods may be stakeholder-and stage-dependent. The BRAID strategy was divided into four different stages as described in Table 1.

Understanding the impacts
The success of the BRAID engagement and dissemination strategy demonstrates the relative efficacy of utilizing different engagement strategies accordingly for different types of stakeholders and illuminates gaps in this analysis. Table 1 examines, where possible, how well each strategy engaged different types of stakeholders.
First, although some engagement strategies were successful at involving stakeholders, there are knowledge gaps about how well engagement occurred. For example, Web-based resources and press releases were good at generating stakeholder interest. However, they did not adequately permit examination of which stakeholders were visiting the website, receiving press releases and downloading these documents.
Second, some strategies, particularly virtual activities requiring high involvement were potentially ineffective due to a lack of incentives for stakeholders. The ineffectiveness of the virtual forum was surprising, because the forum was initiated at the request of stakeholders [10]. Both the forum and Wiki required a high level of engagement with little external incentive to participate, and this likely contributed to the difficulty in generating interest in these dissemination activities [11].
Third, some strategies work well for particular types of stakeholders, but not others. While project workshops and conferences worked well for industry representatives, academics and CSO representatives, with some success for policy makers, they did not engage end users. Also, the integration of disparate stakeholders often resulted in a significant amount of time spent establishing common ground and agreeing upon terms of references rather than engagement activities [12]. Finally, despite the success of the press releases in generating interest in project deliverables, they did not translate into specialist or mainstream press attention, and this should be a key objective for such projects since press attention can increase access to other stakeholder groups.

Recommendations based on BRAID
First, in order to assess the efficacy of different stakeholder engagement and dissemination strategies, projects and other initiatives must find creative ways of monitoring and evaluating who is accessing their materials, which could be accomplished through a web-based survey given when visiting the website or downloading materials that collects stakeholder category, country, reason for visiting, and contact details. This could have the dual impact of identifying who is visiting the project website and increasing the size of the project's contact list.
Second, strong and clear incentives for stakeholder engagement must be developed. The benefits of social media or networking for stakeholders need to be understood by project partners and highlighted for stakeholders, by identifying the significant benefits for stakeholders. Future projects should consider the paraphrased questions from Roper et al.'s [13], for stakeholder analysis (listed in Table 1), which can assist in understanding and maintaining stakeholders' motivations for involvement. Walther [4] suggested that anticipation of future inter-action assists in generating commitment within computer-mediated communication. Better analysis and understanding of stakeholders' motivations could have assisted BRAID in improving use of and commitment to the Wiki and online forums.
Third, project partners should also be cognisant that events involving different stakeholders will need additional time to establish common ground, terminologies and understanding. Partners should consider stakeholder-specific events when multiple groups are not needed. Notably, end users have different needs, interests and motivations, and events geared towards other groups may not be valuable for them and could actually discourage further participation.
Fourth, projects and initiatives such as BRAID must prioritise integration of the media when devising engagement and dissemination strategies. This is a key area where further research would be useful. To engage end users, appropriate language and use of the popular media, such as newspapers, magazines, radio and television, are recommended to provide broader reach and understanding of projects.

Conclusions
This commentary has presented several dissemination tools for successful stakeholder engagement and proposed recommendations that can be used by other policy-oriented projects to improve engagement strategies.