When should service providers (core or external) expect authorship?

Question

There are a number of areas of biology research where specialist knowledge requirements or expensive equipment mean that it's more efficient to outsource work to core or external facilities. This seems particularly common for next-generation sequencing (here and here), transcriptomics and proteomics, but Science Exchange is based around this concept and I'm aware of proposals for similar models in other areas, e.g. for work with pathogens that require high containment. The Diamond Light Source offers something similar, too.

On the whole this seems like a good thing; it has the potential to reduce duplication of investment in infrastructure, increase use of nationally-funded resources and facilitate multidisciplinary and interdisciplinary research. However, I've recently heard a few people assume that a commercial service provider is ineligible for authorship, even in situations where I strongly suspect they would normally include someone as an author if they were doing the work for free. I'm also aware of situations where collaborations have faltered due to differences of opinion over this.

Personally, if someone has made a substantial and unique contribution to designing the experiment, running the experiment and analysing the results, I think I'd be uncomfortable not adding them as an author. Not to mention that such a paper could easily end up including experiments that none of the authors knew how to design or conduct. I'm pretty sure there are a lot of papers out there now reporting NGS which none of the authors could actually have generated without help.

How should commercial service providers be acknowledged in papers reporting work they assisted with, and does the fact that a service is paid for change the criteria for authorship? Is this affected by whether the PI pays for the access, or if the access is paid for from a central budget?

Good-subjective answers only, please. I'm particularly keen to hear personal experiences or evidence from the biological sciences (hence the biology tag), but would welcome answers that compare and contrast this to other fields.

Answer 1

In materials science, there are many companies that will provide a service for payment. Examples include places such as the Evans Analytical group (once Charles Evans) for materials analysis of many types. You send them a sample, they send you results from generally standard analytical techniques. It is a purely transactional relationship - they get paid, you get the data. You generally would indicate in a paper that the analysis was performed by them, but they do not get authorship. The same goes for various other things, like ion implant houses - you send a sample, they implant with the ion species and dose. You are getting a standard service performed, not intellectual effort.

There are other situations where you need to interact with the external entity to develop a new technique, or new twist on the technique. This is no longer purely transactional, and may result in authorship. Clearly the reverse can also happen - if you are a beta site for a new instrument and are working closely with the company to test it and expand the capabilities, you would get on any papers written by the company as well.

My analysis: if you and anybody else can walk up and pay for the service, they are not co-authors. If you have a joint project, with dollars and intellectual effort from both sides, they are. And then there is a broad grey middle ground...