• Commissioners accessed and used information pragmatically to build a cohesive, compelling case to plot a course of action, convince others and justify decisions. The ‘art of commissioning’ consisted of juggling competing agendas, priorities, relationships, demands and personal inclinations to make the ‘best’ decision circumstances allowed. Commissioners balanced multiple pressures.
  
  
• Academic research was occasionally explicitly sought, but usually came pre-digested via NICE guidance, software tools and the clinical knowledge of local GPs. Negative research evidence did not trigger discussions of disinvestment opportunities.
  
  
• Local data, such as service evaluations, often trumped national or research based information in persuading commissioners on a course of action. Conversations and stories were important, these were fast, flexible and suited the changing world of commissioning.
  
  
• As soon as knowledge and information was obtained, it was modified through repeated processes of local contextualisation and engagement to refine the information and ensure the ‘right people’ were on board. Commissioners regularly accessed information through five main conduits:
  - Interpersonal relationships
  - People placement (embedded staff)
  - Governance (e.g. Department of Health directives)
  - Copy, adapt and paste (e.g. best practice elsewhere)
  - Product deployment (e.g. software tools)
  
  
• External providers were contracted for their skills and expertise in project management, forecast modelling, event management, pathway development and software tools. Contracts with external providers with the most value occurred when there were effective strategies in place to actively engage commissioners in learning, embedding and applying new skills, rather than just a wholesale outsourcing of work to an external provider.
  
  
• Commissioners have to be able to interpret and apply the information supplied. For example, it’s not enough to deploy a software tool and provide training on that tool. Contracts need to be explicit in requiring external providers to provide interpreters who can help commissioners to understand the data outputs and apply them to local strategies, problems and issues.
  
  
• Amongst the commercial and not-for-profit agencies in this study, one was not very successful, as the healthcare clients thought the external provider added little of extra value. With another, the benefits were largely notional and with a third views were mainly positive, with some concerns about expense. Because we studied external providers offering software tools, analysts often benefited more than those making commissioning decisions.

The implications of our findings are that:

• External providers, including academic researchers, who want to make an impact on commissioning should maximise their use of the different conduits that commissioners access, ideally employing commissioners’ preferred mode of communication through conversations and stories.
  
  
• To capitalise on the expertise of external providers, wherever possible, contracts should include explicit skills development and knowledge transfer components so that healthcare commissioners benefit longer term.