Tips for Starting a Data Consulting Business

As a side-gig, I'm interested in working as a data analytics consultant to small businesses (i.e., use R to free their data from spreadsheets and turn into dashboards, reports, shiny apps, etc.).

I'd love to get advice from folks about their thoughts and experiences on providing this type of service. For example, how did you start getting clients? What do you find that people want (e.g., descriptives, forecasts, dashboards, reports)? Do you find that people care that an analysis was produced with programming (as opposed to GUI), or they simply care about the final product?

I'm open to any feedback you might have on this topic. Thanks.


I've been on the consulting side (though not R) and I've been on the procuring side and here are my thoughts:

Look to solve client pain points. Whilst we (the R Studio Community) might care about the latest R version, the latest developments in R and new packages, I think clients will (likely) only care so far as how it can benefit them, which for a business could be small scale such as productionising reports or performing advanced analysis which can help them achieve their business objective (cutting costs, gaining customer insight, growing revenue). In summary, really focus on the client and their needs.

I'd say it totally depends on the client, but as I said above, focus on their pain points. I'd strongly recommend to set expectations. Don't promise the world before you know what data you have access to, the cleanliness of the data and what resources available etc.,

I don't think it should matter, in fact, I would prefer programming! However, my feeling is that (especially in the UK) there is a trend for the "democratisation of data" and allowing a non-technical audience to reproduce analysis which is why Tableau, Qlikview, Power BI, Alteryx etc., are so in demand at the moment. I emphasise that this is just my intuition and feeling of the market - I haven't looked at market or sales stats to back it up.

1 Like

I went through this myself, and I learned two very important lessons:

  • especially at the first meetings: listen very carefully to the customer. Be patient and do not insult him by jumping to conclusions (this came very hard to me).
    Remember your relationship will be about his needs, not your wants - so if you want to do some fancy math and he needs a couple of simple reports then simple reporting it needs to be.
  • focus on potential clients with money. A perfect client has a need of your services and money to afford them. Alas, customers such as these are very, very rare.
    It is easier to convince someone with money that he has a need of your services than to convince someone with a dire need of them that he has the money.
    For this reason it is often easier to work for big companies, who are used to paying real money for real services, than a honest small business owner who has little idea what he is talking about, but has a nephew that is handy with excel and "could do this over an afternoon for some ice cream money".

The best way to get clients is through personal contacts. This is where networking comes in handy.

Does a previous employer of yours require your services? Or a former colleague at a different company?

@Mark6 Thanks for the advice. I prefer programming too! I would like part of my pitch to be that although drag-and-drop software (e.g., Power BI) may have a place, there is tremendous value in having a human program an analysis. I'm just not sure the extent to which potential clients care.

@jlacko Great point about listening carefully. As to your second point, I agree, but my thinking was that many (if not most) large companies already have full-time analyst(s), but maybe that assumption is wrong. I do think many companies meet both criteria (i.e., large enough to pay a consultant but not big enough to have a full-time person).

martin.R I agree about personal contacts, who I've started reconnecting with. Because of issues with conflicts of interest, there are limitations to tapping into my professional network, who mostly work in the same industry (higher education).

1 Like

Interesting thread. I work at a bank but feel I might be more effective consulting for banks. There seems to be a great need for the kind of skills that come with people who are well-versed in performing in-depth analysis in R, as well as understanding and responding to client needs. The combination of the two, I think, is what makes for a great consultant. From my experience, we're much more eager to pay an external consultant to automate work, provide guidance, and help us move away from the traditional way of doing things at banks than we are to listen to our own employees on the same subject.

One thing I will caution is that when you start consulting you're essentially starting your own business. That's a quick way to end up in a dangerous place. There are some great books on that, the E-Myth Revisited might be worth reading. Running a business is very different than earning a salary, and just because you enjoy doing something doesn't mean you'll enjoy running a business that sells as a service that thing you used to enjoy doing.

1 Like

@pedram Thank you. I've had the same, strange, experience at my work, where leadership is more keen on paying consultants or buying the latest and greatest software than listening to current employees who have the ability to do the same thing with existing tools.

The subtext of my question is my naivete about running a business, so I appreciate the advice about E-Myth.

1 Like