I am scheduled to be part of a panel discussion titled “Successful Statistical Consulting: The Practicalities.”
Here’s the abstract for that session (reformatted slightly for the web).
The Statistical Consulting (CNSL) Section has recently engaged in a lengthy series of exchanges centered around the practical issues of providing consulting services - - typically to individuals or agencies outside one’s current employer. Issues that should be addressed in these situations include:
- Nondisclosure Agreements (NDAs, CDAs etc.)
- Need for legal agreements
- Professional liability insurance, Professional Liability, Errors and Omissions, Officers and Directors, etc.;
- Terms of Rights and Ownership
- Need for and format of service contracts
- Organizational structure to use: Subchapter S, LLC, etc. (in US)
- Issues of data ‘sharing’, confidentiality, security
- Regulatory issues and liability
- Ethical issues
- Issues of data ownership
- Authorship privileges
This session is intended to allow participants to engage in direct discussion regarding these issues. The session will be structured to provide specific time limits for each of the topics to be addressed. Participants will be invited to present arguments on both sides of the aisle (as it were. The intent is to foster discussion and share experiences, not necessarily to formulate rigid guidelines
Here’s a brief outline of what I will talk about.
I have worked as a statistical consultant in several academic locations: the University of Iowa (as a graduate student), Bowling Green State University (as a faculty member), and the University of Missouri-Kansas City (also as a faculty member). I have worked for nine years as a consultant at a government agency and twelve years at a hospital.
I also have an independent consulting business, P.Mean Consulting, which I started in July 2008, though I had previously done a limited amount of private consulting. Here are some of the characteristics of my consulting practice. I don’t want to hold up my practice as a model that everyone should follow. I’m sharing this information to help offer some contrasts with how others set up their consulting business:
I have set up my practice as a sole proprietorship. I’m sure others will talk about this issue, but a sole proprietorship is the simplest way to set up a consulting practice. There are other options, such as a Limited Liability Corporation or Subchapter S, that offer protection against debts and clarify the roles of business partners and investors. But until I see the need to take out loans to support my consulting business (or until an angel investor wants to give me a million dollars), I don’t see any need for these options.
I also do not use a standard business contract. If a client wants to back out of a consulting agreement, I see no reason not to let them. So I’m pretty casual about this. If there are important issues (such as confidentiality) that need to be delineated, I would clarify them using a one page letter. If the client wants to draft a contract for me to sign, I’ll sign (after reading it, of course).
I prefer to retain ownership of anything I write. I don’t want to resell it, or hide it from my client. I want to retain ownership so I can re-publish the material on my website. I’ll try my best to write in very general terms if there is a privacy concern. That doesn’t always work, and some of my clients don’t want me to put anything related to their work up on my website. I’m disappointed when they request this, but I honor their request.
I have no liability insurance. The types of consulting I do are (in my possibly foolish opinion) not likely to lead to litigation. If I were to serve on a Data Monitoring Board, an Institutional Review Board, or some other position that can impact directly on decisions about patient care, then I would make sure I got liability insurance.
I don’t demand coauthorship as a pre-requisite on any projects. I get my fair share of coauthorship without resorting to any cajoling, wheedling, or threatening action. If I spent more than a trivial amount of time helping someone to write a manuscript, my only effort to suggest the need for a coauthorship is to point out the recent scandals associated with ghost authorship.
Because I am currently working halftime in the academic world and halftime as an independent consultant, I want to talk about the differences I have seen between the two types of positions.
As an independent consultant, you aren’t the gatekeeper. At my government job, I had to sign off on any research conducted in my division, and I was expected to withhold that my signature from “bad” research. I suspect that the gatekeeper role is common in most government agencies and in many industrial settings. It is part of being a good steward of the taxpayer’s or stockholder’s money.
You’re typically not a gatekeeper in an academic consulting environment. For graduate students that job is reserved for the thesis/dissertation committees. Faculty members who come to your consulting center live a pretty much ungated life, as long as they can get their research proposals through the human subjects review board or the animal care and use committee. In an academic setting, you are not expected to tell people that they can’t do their research.
It’s even more extreme as an independent consultant. Some clients will listen to you if you tell them that their research is fatally flawed, but if they want to be stubborn about it, you have no leverage other than the strength of your persuasive skills. It is ridiculously easy for your client to fire you and find a different independent consultant if they like.
That doesn’t mean that you should not try to act as a gatekeeper in these settings. By all means, be blunt and let them know that they are collecting the wrong data, or that their approach is unlikely to pass muster during the peer-review process, or that they are being self-delusional. Just do this with the knowledge that your only trump card is refusing to work with that client if they persist in their flawed research.
As an independent consultant, you have to bill regularly. This is one of the aspects of independent consulting that I hate, but it is important for two reasons. The first and most obvious reason is that if you don’t bill for your services, you will starve. But regular billing is also critical for giving your client feedback on how much you are costing them. If you let the invoices slide and then sock them with a big invoice three months later, your clients will be understandably upset with you.
Some academic consulting centers bill their clients and some don’t. But the billing process in an academic setting is typically handled by someone else.
You need to think long and hard about your billing model. Two common models are: bill by the hour or bill by the project. If you bill by the hour, you keep tabs on the amount of time spent on the project and charge a fixed amount per hour of work. Most clients will want you to place an upper limit on the number of hours that you work. If you bill by the project, you need to be explicit about what will be produced. You should structure the project with several discrete completion points where you can ask for money for the portion of the work already done. You should be cautious about clients who ask for work beyond the agreed upon scope. These require additional fees, and need to be negotiated carefully.
As an independent consultant, no one will subsidize your professional development or your independent research program. Attending conferences like the Joint Statistical Meetings is not something that you can bill to one client or another. They won’t buy new books for you or support your professional development. It’s all on your own dime, though the expenses are tax-deductible.
I’m currently writing a book, and my boss at my academic position is quite happy to provide me with some of the time needed for this. She is also supportive of an effort of mine to get my own NIH grant.
As an independent consultant, you do not have a support network. Unless you are very fortunate, you will not have enough billable hours to justify hiring an administrative assistant. I had a full-time AA at my hospital job and it is the one thing that I miss the most from that position. No AA means that you have to schedule your own appointments, prepare your own travel, document travel expenses, and a host of other things.
As an independent consultant, you don’t have access to a company lawyer, or to benefits specialists.
As an independent consultant, you have to provide your own insurance and health care benefits (though in my case, I rely on my spouse for insurance and health care benefits).
As an independent consultant, you can pick and choose your customers. It may seem like heresy to turn down a paying customer, but being an independent consultant means that you don’t have to work with someone you don’t like and you don’t have to work on a project you don’t like. Furthermore, you can work on soliciting business from desirable customers. I haven’t walked away from too much business, with one exception. I love to travel, but with family commitments, I have tried to limit any work that involves travel. I charge a higher rate for consulting that involves travel. When I told a client I would only do my work by teleconference, I lost the job, but I wasn’t too disappointed.
If you’re struggling to find billable hours, then this may be a moot point, of course.
As a consultant for an academic consulting center, you pretty much have to help anyone who comes in. In extreme circumstances, you might be able to refuse to work with someone, but it is not easy. It helps if there is more than one of you at the consulting center, of course, because then you can hand off difficult clients to a colleague.
As an independent consultant, it’s easy to keep your boss happy. I’ve had some very nice bosses over the years. They have typically allowed me a fair amount of latitude on how I schedule and how I conduct my work. But even the most flexible of bosses can’t be as good as being your own boss. There’s something quite liberating about not having to ask anyone for permission.
You can find an earlier version of this page on my original website.