|
Howdy, A few friends and I were kicking around the idea of doing Statistical Consulting as a side venture, and I was wondering about your experiences in the field! I'm interested in hearing it all, especially how you got your clients, what markets you focused on, the troubles you encountered, how much you tend to charge per client(if you're willing), what credentials you have and just anything you want to share about the business. Heck even if you think it's a decent idea for someone to try to jump into the market! Thanks!
|
# ? Dec 22, 2014 18:15 |
|
|
# ? May 5, 2024 17:18 |
|
Texibus posted:Howdy, The few people I've met that have succesfully started a consulting firm has had a network of clients available before startup.
|
# ? Dec 22, 2014 19:00 |
|
Absolutely you got to have a client source! I'm interested in the pitch though. I'd have to imagine everyone starts off with seeing a need and then having to say," hey, I can help you with that problem and you don't need to have me in your office 365 days out of the year to solve it." Definitely a good thing to consider though, thanks man!
|
# ? Dec 22, 2014 19:13 |
|
Texibus posted:Absolutely you got to have a client source! I'm interested in the pitch though. I'd have to imagine everyone starts off with seeing a need and then having to say," hey, I can help you with that problem and you don't need to have me in your office 365 days out of the year to solve it." My wife does a mix of stats and scientific writing consulting, and in her case it was actually the opposite - somebody asked her advisor if he knew anyone who could do some consulting and he steered him to her. When someone later asked me the same thing and I steered them her way too and those two first clients liked her work and started recommending her in their networks, it was enough of a base for her consulting to take off. Most people I know with consulting careers grew them that way instead of the other way - it was something they were already being requested to do for money before they established an independent career. My suspicion is that if you're trying to go more top-down, you're going to have to advertise same as any other business - maybe try Craigslist and some flyers up at universities?
|
# ? Jan 2, 2015 23:57 |
|
Is your wife a PHD? Seems like she already works in academia. Do you know how much she generally charges for the work she does and the type of clients she has (doing it for university research papers or helping business figure out problems). If that's all too personal I understand either way valuable stuff, thanks man.
|
# ? Jan 13, 2015 16:14 |
|
I work as one both professionally and freelance. Odesk and elance work ok. Better is getting in touch with uni departments you have a link to and asking around if anyone needs assistance on papers. This will not pay very much money. To get connections with business make it a sales pitch less focused on your technique and more focused on the business upshot. Any more specific questions I'm happy to answer.
|
# ? Jan 13, 2015 21:09 |
|
semicolonsrock posted:Better is getting in touch with uni departments you have a link to and asking around if anyone needs assistance on papers. This will not pay very much money. This is potentially more interesting work though. What are we talking here, $50/hr? Do you bill hourly or by project?
|
# ? Jan 14, 2015 00:45 |
|
semicolonsrock posted:I work as one both professionally and freelance. Odesk and elance work ok. Better is getting in touch with uni departments you have a link to and asking around if anyone needs assistance on papers. This will not pay very much money. Awesome! Maybe you could start with what credentials you have because presently that's what my buddy is concerned with most, the fact that he only has a bachelors but lots of industry experience working with medical insurance companies.
|
# ? Jan 14, 2015 15:43 |
|
Texibus posted:Awesome! Maybe you could start with what credentials you have because presently that's what my buddy is concerned with most, the fact that he only has a bachelors but lots of industry experience working with medical insurance companies. He's right to be concerned. All the ASA reports I've looked at over the years for compensation research have bachelors-level statisticians as an extremely small percentage of responders (<5%) . I did some consulting for small academic projects and other grad students while I was in school, but only after I'd gotten my masters by passing prelims. I ended up ABD rather than finishing the PhD, and that locks me out of some jobs.
|
# ? Jan 14, 2015 15:56 |
|
Understandable, did you try and work very much with businesses?
|
# ? Jan 14, 2015 16:01 |
|
Caveat, my info is 10 + years old and university based. If you are going to consult at a University level, it is going to be hard with just a bachelors. At larger universities, any research is grant supported, and when you are going for grants, they will account for the statistics support. Any larger department will have that one person who specializes in stats and experimental design and is 10 % on everyone else's grants and gets publications that way too. At smaller colleges, the math/applied math department will serve that function somewhat. You may be able to help grad students, but their advisors should be connecting them to resources within the school, plus there usually are 'stats for grad students' type classes. It is sort of a rite of passage to muddle through the stats for your thesis. The best you could do would be a sort of stats tutor. Having said that, I think the more important think is to have a business proposition. Not, I will help you do stats, but, I will help you do the statistics needed to show that your non-profit is not discriminating based on gender, race, age, etc when it provides services, or something related to the medical insurance field that your friend has experience in.
|
# ? Jan 14, 2015 16:22 |
|
Texibus posted:Understandable, did you try and work very much with businesses? Not at all. My field (biostats) doesn't have a ton of overlap with business stats once you get past the fundamentals. Most of my consultation work was for grad students or the school of dentistry. Now I'm in biotech/pharma and completely satisfied with my day job, though I may switch to part-time contract work years down the road for quasi-retirement.
|
# ? Jan 14, 2015 16:43 |
|
Yeah, I was hoping to spin a business type of consulting where education would be hopefully less of an issue. In academia I can see how it would be extremely important to have a PHD or ABD verifying results.SubjectVerbObject posted:
You nailed it, what you mentioned is exactly the type of work were shooting for! Nothing that needs to go into a paper but more, "hey we looked over your data and these are the trends that shook out, maybe it can help you make some money back up or keep you from getting sued." As far as the medical field, our fear is the huge amount of liability that can be assumed when consulting those companies. And to be clear, no real interest with working in the University circles. Silver have you ever done stats in that capacity for clients? I'd be interested in that experience the most. Texibus fucked around with this message at 17:19 on Jan 14, 2015 |
# ? Jan 14, 2015 17:04 |
|
I hate to even say it, but it may work better to do process improvement, Six Sigma, Quality, whatever is the new buzzword. The stats would be there, but the focus would be on improving the business. The downside is if you become a Six Sigma black belt, I will hate you.
|
# ? Jan 14, 2015 17:22 |
|
Texibus posted:Yeah, I was hoping to spin a business type of consulting where education would be hopefully less of an issue. In academia I can see how it would be extremely important to have a PHD or ABD verifying results. Your first comment here concerns me a bit. If someone needs a body to verify results, then that's the kind of work that could be given to a green-behind-the-ears kid straight out of school. Consulting work is inherently more challenging - your job is to tailor an analysis to the client's needs, while respecting the limitations they have. Often you're not brought on to a project until after data is collected, so you're stuck in retrospective analysis and have a need to assess data quality and robustness of your conclusions. Throwing everything at a wall and seeing what sticks is not a value-add proposition unless you have some way to control the false discovery rate, if data mining parallels statistical genetics.
|
# ? Jan 14, 2015 17:22 |
|
To be fair, you're talking with the business end of the operation and not the brains. I started the conversation so I could get my arms around what this type work would entail and gain insight on how to grow a client base. My buddy would handle the type of concerns you're worried about and sorry for concerning you with my poor way of explaining the business model. Basically, I should have just said we're looking to work strictly with businesses in some statistical consulting capacity (we don't know what that is yet), not do the type of work university students or departments would need help with which we couldn't offer because of the education limitations. I guess my question would be, couldn't a guy with a bachelors and 8-10 years of experience accomplish what you're saying or is his experience not going to offset that? Texibus fucked around with this message at 17:43 on Jan 14, 2015 |
# ? Jan 14, 2015 17:30 |
|
Engineer Lenk posted:He's right to be concerned. All the ASA reports I've looked at over the years for compensation research have bachelors-level statisticians as an extremely small percentage of responders (<5%) . I did some consulting for small academic projects and other grad students while I was in school, but only after I'd gotten my masters by passing prelims. I ended up ABD rather than finishing the PhD, and that locks me out of some jobs. Up until about 5-10 years ago, undergraduate majors in statistics were not common. There's been a lot of growth in those programs since then, so I would expect this to change eventually.
|
# ? Jan 14, 2015 19:49 |
|
Texibus posted:To be fair, you're talking with the business end of the operation and not the brains. I started the conversation so I could get my arms around what this type work would entail and gain insight on how to grow a client base. My buddy would handle the type of concerns you're worried about and sorry for concerning you with my poor way of explaining the business model. Basically, I should have just said we're looking to work strictly with businesses in some statistical consulting capacity (we don't know what that is yet), not do the type of work university students or departments would need help with which we couldn't offer because of the education limitations. It really depends on the experience. If it's directly related to this sort of business analysis, there's a strong case for it. If it's only tangentially related, it'll be an uphill battle. If I was approaching this, I would put together a number of different hypothetical case studies with high-level results and at least one sample report, then focus on businesses that would likely have similar issues. Cold-calling and selling a business on your services with a value proposition may end up with better results than being one of many options for a bid.
|
# ? Jan 14, 2015 20:41 |
|
Chiming in to say I'm extremely interested in this topic and wish I had more to contribute. I'm M.S. statistics guy, but I work as an engineer for a major defense contractor. Eventually, I'd like to get an independent consultant role, but I have no role models to follow nor idea how to do it. Everyone that I could emulate has gone up the engineer or business track, but I'm searching another option. Your replies are greatly appreciated! Pointed question: anyone have experience with statistical consulting of classified projects? I think that may be a niche I could fill, but I'm unsure how to start.
|
# ? Jan 15, 2015 08:20 |
|
Texibus posted:Is your wife a PHD? Seems like she already works in academia. Do you know how much she generally charges for the work she does and the type of clients she has (doing it for university research papers or helping business figure out problems). If that's all too personal I understand either way valuable stuff, thanks man. I'm a terrible husband and I don't actually know how much she charges per-hour, but I know it's a sliding scale - she charges students less than professors. The last time I talked about it I think she was charging in the mid-$20s on average, but based on her workload I think she could charge more. It's a sideline business for her, so she's not very aggressive about pushing it. She has a PhD in psychology with a strong stats background but not a PhD in stats specifically. My take on the issue of whether a PhD is necessary or not comes down to the question of who your audience is and how you can convince them to pay you $$ when they don't already know and trust you. In my experience the vast, vast majority of the work that real-world/nonacademic businesses do doesn't require PhD-level statistics, but having a PhD goes a long way towards establishing your credibility as a person delivering the data. I work primarily in the nonprofit sector doing research and program assessment stuff. Occasionally I come across something where my own PhD is useful for some wonky methods or analysis, but most of the time it's very basic stuff that anybody with a BS in stats or a MA in a stats-heavy field could do very easily. The problem comes when I have to explain to people that this or that program isn't working or needs to be changed based on my stats. They want to argue with the results, and the extra legitimacy that comes from a PhD was helpful earlier in my career when I didn't have a history in the field. If your work experience isn't enough by itself to be generating requests for consulting gigs, my suspicion is that you're going to struggle to get the kinds of big-dollar contracts you'll need to get to make a living as a consultant if you don't also have at least a master's-level statistician working with you.
|
# ? Jan 16, 2015 18:47 |
|
I have been consulting for about 5 years now. I have a background in behavior science and a masters in I/O psychology, so not stats, but on the scientific side of things. I am currently working on finalizing a contract with an agreed upon rate of $130/hr - if the contract is less than 3 months - and $95 if it's longer. This will be my best contract yet because the more I consult the more I feel am I worth. I started by charging $25/hr. I also have daily rates sometimes and that varies. When I give a workshop it is around $1200/day + time costs for customization, materials, etc...Texibus posted:Yeah, I was hoping to spin a business type of consulting where education would be hopefully less of an issue. In academia I can see how it would be extremely important to have a PHD or ABD verifying results. What you are describing here is not a Statistical Consultant...even if you want to do stats. You're being too high brow with that kind of posturing and that's why academia was assumed. Businesses want to hire people that help with metrics and poo poo. You want to do pretty simple stats work, right. Like you said, you're not doing work to meet peer review, you're doing work to help managers make decisions. I got into business consulting to avoid academia, which basically means I do applied work. I think of applied statistics as business analysts or Data Specialists or HR Metrics or whatever... you can literally call yourself anything you want. If you have a good background in stats then coming up with experimental designs to tie metrics into business results is rather straighforward. For HR, you look at how onboarding improves efficiency more quickly or how their new policy is effecting time to hire or whatever. For IT you relate things back to downtime or make an algorithm to predict when hardware will need to be replaced. Etc... That's kind of the easy part for you since I'm assuming you're competent. The hard part will really be marketing your services in a way that business relate to and then, once you have their interest, knowing how to adapt your services to their needs and close a deal. Being a consultant means selling your services and, for me, that was the hardest part. The first product I sold was an assessment I made to assist with hiring and succession planning. I offered consulting on how to use the results of the assessment as effectively as possible. I started off being all about the science because to me the product I put together was just so scientifically sound, which is hard to find in assessments. I thought it was a great differentiatior since I saw so many other assessments for the bullshit they are. I talked about Rasch analysis and logarithmic scales and r-values and poo poo. No one loving cared. I met with CEO's, VP's, mangers, and other educated people that could have firmly grasped the science, but they didn't care one bit. It was foreign to them and really not important for them to know it. What they did care about boiled down to two things: What actions will the results help us to take and what will be the impact of those actions (on our bottom line is applied). Marketing is the key here and that is the message you want to make crystal clear. You want to have models with catchy names that you can make a brochure for. You want to have a clear outline of what your process usually entails. Data is big these days and business are really starting to care about using all the data they collect in a meaninful way. You can definitely find work in the field if you have any social media skills or even just an average network. Can you sell a business on your ability to do that stuff? If so, you're good. I got my start because I was carrying out research and giving talks at professional conferences on my work. One day someone asked me to tie my work into their business and offered to pay me. From there I spent some time setting up a website, making sales material, doing legal poo poo etc... I had a lot to learn and I had already taken a consulting course during my I/O masters, so don't expect consulting to just like magically happen. Businesses expect a certain level of profesionalism from consultants and your first impression is important. If you don't seem competent in their business setting right away, they will think you are green and it's uphill from there. That means knowing the right questions to ask and unfortunately, a lot of those questions come from experience. My MS in I/O psych really helps me because it taught me about organizational structures and management theory and other business stuff. My point is that consulting for organizations is as easy as selling yourself. They sort of just assume you can do what you tell them so they're not going to look into your stats skill. If you can sell yourself, you'll get an opportunity. From there it's being able to put together a statement of work, negotiate rates, put together the contract(s), and deliver. Edit: Adding some stuff about rates Jewce fucked around with this message at 19:43 on Jan 16, 2015 |
# ? Jan 16, 2015 19:36 |
|
Velochis posted:
Given the complexities of federal contracting, I can't imagine that would be easy. Your best bet would be to get to know some small businesses that are involved in federal contracting as subs and see about working with them on an independent basis.
|
# ? Jan 19, 2015 03:04 |
|
|
# ? May 5, 2024 17:18 |
|
Thanks dude, exactly what I was looking for. All this might not be for us, but it something we wanted to explore and you have given some excellent insight into exactly what we're looking to do. I really appreciate your post, seriously you're awesome. Sorry if I mislead y'all from the true intent of this thread. Texibus fucked around with this message at 06:30 on Jan 20, 2015 |
# ? Jan 20, 2015 06:12 |