How much data analysts (and other data roles) get paid in UK nonprofits
For six months now I’ve been running a small side project. I collate a weekly list of data vacancies in nonprofits. What else is there to do on a Sunday morning after all? You can see previous newsletters and sign up to have each one delivered here.
Since it’s been six months and I work for a data social enterprise, I thought I would have a look at the data.
The dataset
It’s not a very rich dataset. For each vacancy I capture:
the job title as advertised. There’s a range of these ranging from the general Data Analyst to the quite specific Head of Data & Insights (Marketing Ops).
the salary as advertised
the name of the organisation
the URL where I saw the vacancy. This is not persistent as these URLs tend to vanish when the vacancy is filled or the closing date passes.
A single salary figure
First I focused in on the salary. What is the going rate for, for example, a data analyst? I wondered. I had to do a little cleaning of the data (of course). Salaries are typically given as a range, they may include London weighting (which I find particularly curious in these days of remote or hybrid working), they may be for full or part time roles and they may advertise “plus additional benefits”.
I chose to take the upper end of any salary range including any London weighting that was applied. I also chose to take the full-time equivalent salary. I also discounted any additional benefits. That gives us a single salary figure for each post. This might be higher than someone would get paid if they took that job (because they might start on the bottom of the scale, be outside London and only be working part time). I think that’s worth bearing in mind when interpreting the data. I’m open to the idea that I should use the low point of the range or the mid-point, and I’d love to hear any arguments people have in favour or against.
Standardising roles
I then standardised the role type. I didn’t have much information with which to do this because I only have the job title. I’ve ended up pulling out what seemed to be the key role type from each role name.
For example:
Data Analyst is coded as Analyst
Data & Information Manager is coded as Manager
Data Selections & Reporting Officer is coded as Officer
Over six months I have listed 273 unique vacancies (there is the occasional duplication in the newsletter and I’ve removed these in the dataset). These have been classified into 42 job types. Some job types occur much more frequently than others.
68 Analyst
41 Officer
40 Manager
19 Head of…
17 Lead…
12 Engineer
7 Co-ordinator
7 Director
6 Data Scientist
6 Senior Analyst
5 Architect
Median Salary for each job type
We can calculate the median salaries advertised for these job types.
Median advertised salary for different role types
Showing only job types with at least five vacancies in the dataset
Co-ordinator is the job type with the lowest salary (£25,000).
Median salaries then increase gradually in turn for each of Officer, Analyst, Senior Analyst, Manager, Architect, Data Scientist, Lead, Engineer and Head of.
The median salary for the Head of job type is £55,645. There is then a big jump to the Director job type which at a median salary of £87,000 is the highest paid job type in this chart.
Analyst is the most frequently appearing job type. Analysts have a median salary of £37,000.
Salary ranges for different job roles
Median salaries are interesting but the range of advertised salaries gives us additional information. There are many different ways of showing the range of salaries. A simple approach is to calculate the quartiles. The lowest quartile is the salary where 25% of jobs would be paid less and 75% of people would be paid more. The highest quartile is the salary where 25% of jobs would be paid more and 75% would be paid less.
Looking at the range from the lowest quartile to the highest quartile shows you the middle range of salaries, 25% of jobs are paid more, and 25% are paid less but 50% of jobs are within that range.
Range of salaries for different job types
From lowest quartile to highest quartile
Showing only job types with at least five vacancies in the dataset
Director is the job type showing the largest range, from £66,276 to £105,000.
The Architect job type shows no range at all. All the Architect job types for which we have salaries were advertised at £43,788.
Data Scientists show the second highest range of salaries from £ 38,310 to £54,655.
There is considerable overlap in the salary ranges of Analyst, Manager, Senior Analyst, Data Scientist, Lead, Engineer and Architect job types. Co-ordinator and Officer job types overlap in salary and are the lowest paid job types in our dataset.
The data used in both of these charts is available as a csv.
Benchmark limitation
I hope that these figures are useful but I think it is worth emphasising some limitations.
The figures are based on advertised vacancies. We can’t tell from this dataset whether organisations were able to recruit to these posts. Over a longer period of time we might be able to see readvertisements appear in the data but so far we haven’t detected this. We also can’t tell if candidates were able to negotiate higher salaries than advertised or where in an advertised scale a successful candidate was initially paid.
Posts are often advertised with a salary range. This is not the same as the range of salaries shown above. The range of salaries shown in this post is the range of maximum salaries advertised.
I don’t record whether the post is office-based, hybrid or remote. These differences might have real value for some candidates (for example people might accept a lower salary to be fully remote). Many posts advertise a salary plus benefits and I don’t capture these benefits in any way. It is possible that some candidates would accept a lower salary for some other sort of benefit.
The original data is based on searches that I perform in different advertising services. I include all jobs I see that I regard as “data jobs”. I might be missing a proportion of jobs that are advertised in places that I’m not looking. I also exclude jobs advertised by recruitment agencies unless they name the nonprofit they are recruiting for.
Future analysis
I hope to update this benchmark in six months time. At that point I would also like to add more analysis.
There are other questions it would be interesting to ask of this dataset. I would like to know if larger organisations pay more for the same role (or perhaps smaller organisations feel they have to offer more). I will be able to answer some of these questions in the future because I capture details of the recruiting organisation. The dataset is not quite large enough yet but, assuming I continue with the project, in a further six months I think I will be able to provide some analysis about size of organisation and, perhaps, even types of organisation (as in housing, environmental, social care etc).
I think it might be interesting to see if there is any relationship between this analysis and our annual State of the Sector analysis.
Get in touch
Please get in touch if you have comments or suggestions about this approach to benchmarking.
We’d love to hear from you if your organisation is looking at developing data skills and might benefit from our assistance.
And if you are advertising a data vacancy you would like to be included in my newsletter, or just come across a vacancy that should be included, please let me know.