Nature Survey Spring 2025: First Baseline Data Report

1. This document presents a summary of the preliminary results from the Nature Survey undertaken over the Spring 2025. Before this survey was developed, anecdotal evidence pointed to risks such as knowledge loss, skills gaps, and a lack of diversity in the current nature workforce. The Nature Workforce Survey was developed to address the lack of formal data, to collate baseline data to inform policy, strategic workforce planning. The survey is a foundational first step in understanding the nature workforce.
2. This summary provides a snapshot of the status of the nature workforce in Wales, the survey was open from 18 March to 18^th August  2025. The survey received 1,180 valid responses from both paid and unpaid individuals. Anyone who thought they ‘’contribute to nature recovery in Wales’’ could respond.
3. Caution should be taken in interpreting and re-presenting the results, they only reflect the views of those who self-selected to participate. Participation both to the survey itself and in answering each question was voluntary.  Not everyone answered every question and many, questions were intentionally filtered by Smartsurvey  e.g ‘If you are in paid work.. what is the source of funding? In that case only paid workers were asked to answer the question. Therefore, the ‘base number’ varies for each question. If the number of responses was less than 1180, the actual number of responses (base number)  is reported.  .  
4. Data should be interpreted with caution and are subject to response bias; finding should only be used as a general guide to the views and experiences of respondents, rather than as a comprehensive representation of the entire sector. Participation was voluntary and included paid staff and volunteers, any role, any sector, Wales-wide.  While the data provides insight into the nature workforce, it only reflects those who choose to respond. In all cases any re-presentation of these results should be prefaced with ‘of those who responded’.
5. The first baseline data results are at Annex A. Further detailed analysis will follow. 

Lessons learnt; first reflections on methodology and approach

6. The methodology was broadly successful, overcoming many of the challenges of a dispersed and complex workforce. It removed the need for organisations to collate data. Prior to the survey, it was not known how many people were in paid work in nature, or how many would respond, but thought the number may be less than a hundred. In terms of volunteers, these were even harder to predict because each person was only counted once, including those that volunteered and did paid work. Participation was therefore higher than anticipated, with over 1,000 usable responses.
7. However, there was an uneven response across Wales and 649 usable or incomplete / empty responses, which suggests areas for improvement. The question on highest  qualifications referred to NVQs levels with a look up table for equate specific qualifications. It would appear that (other than NVQ level 7), many respondents seemed unable to identify their highest NVQ qualification level. In future specific qualifications should be named eg. ‘O’ level/GCSE/HND ’A’level, undergraduate degree etc.
8. The use of Smartsurvey was extremely effective, and efficient for data collection. The use of that and AI for analysis and report writing has saved a considerable amount of staff time. Given the higher than anticipated volumes, there would not have been the staff resources to cope with a survey of this size, without using these technologies. However, the required manual checking and additional report writing requirements were far more significant than anticipated. They would need to be factored in, for  any future use. The lesson learnt report provides further information.

Acknowledgements

9. Welsh Government thanks all those that helped design and pilot the survey and all those who responded. 

Next steps 

10. The next report will look at subsections of the data e.g.looking separately at those in the paid workforce and those who undertake voluntary work. Also, cross-tabulations, for example by age, gender etc, where the number of responses allow.  That may include taking a particular question and analysing it by e.g.  type of work (e.g. full-time/part-time) or against demographic information such as age or geographic location. The analysis will be seeking to identify commonality/difference in views or experiences by different demographics  and characteristics. The intention is to inform the next steps for nature workforce planning. 

The number in brackets indicates the number who responded to that question. Respondents were ‘routed’ based on the answer to previous questions e.g only those in paid roles answered questions about salary. 

1. Role in Nature Recovery (1,180)
   Respondents were asked about their involvement in nature recovery. Of the 1,180 people who responded, 39% reported undertaking both paid and voluntary roles. A further 27% said they only volunteered, while 24% indicated they only held paid roles.
2. Paid Work in Nature Recovery (780) 
   780 people reported they were in paid work (with or without doing voluntary work as well) . When asked about the nature of their paid work, 43% stated they worked full-time and another 14% part-time, but in each case, all of their time was on nature recovery. Another 27% worked full-time and 11% part-time, but they only spent part of their paid work on nature recovery
3. Funding Duration for Paid Roles (752) 
   Respondents in paid roles were asked about the duration of their funding. Nearly half (49%) reported that their roles had no time limit on funding. But 21% were in posts with under two years funding. Another 17% described their funding as so complex it did not fit any category.
4. Paid Roles, Salary Composition (739)
   In terms of how their salaries were funded, it is a complex picture for paid workers. Again, about half, 48% of respondents said they had one paid role, funded by a single source. Another quarter, some 27% reported they had one paid role, but it was supported by multiple funding sources. But 12% had multiple paid roles, with multiple sources of funding. A further 7% have a paid role, none of the categories explained the funding.
5. Primary Employer of Paid Roles (740)
   Of those in paid work, respondents were asked to identify their primary employer. The largest group over 260 people (36%) worked for nature-focused charities or not-for-profit organisations, while  over 100 (15%) were employed in regulatory functions such as Natural Resources Wales (NRW) or the Health and Safety Executive (HSE). 45 people (6%) responded from National Parks. Over 80 people (11%) responded either from nature based private sector employers or were self-employed.
6. Unpaid work (1,068)
   1,068 people responded (including those also doing paid roles). Of those, 47% did unpaid work or volunteering with a nature-focused charity and 5% with National Parks.
7. Time Spent on Nature Recovery (1089) 
   All respondents were asked, taken together (both paid and unpaid role(s)), how much time do you estimate that you contribute to restoring or enhancing nature in Wales? 28% said they contribute up to one working day per week, 54% between one and five days, and some 17% work more than five days per week.
8. All Roles, Geographic Scope (1092)
   In terms of geographic focus, 41% of respondents worked within a single local authority area, 21% said that ‘Most of my time I work in one local authority area in Wales, but I also undertake multi-area/ pan-Wales/ outside Wales work’ whilst, 34% worked across multiple areas or pan-Wales and only 2% only worked outside Wales.
9. All Roles, Local Authority Area of Work (680)
   Of the 680 who chose a specific geographic area where they worked; the spread was uneven. The most frequently cited areas of work were Pembrokeshire 120 people (18%), Powys 80 respondents (12%), and Gwynedd 66 or 10%. 58 people  (9%) responded  from Swansea responded, compared with just 33 from Cardiff (5%),  despite having the largest population and a number of head offices in the city.
10. All Roles, Training Undertaken (1,086) 
    Half of all respondents (50%) both paid and unpaid reported receiving training relevant to their role in nature in the past 12 months. However, 15% took training relevant to their role more than five years ago and 16% reported that they had not undertaken any training relevant to their role in nature.
11. All Roles, Training Hours in Last 12 Months (1,089) 
    When asked about the amount of training received in the last 12 months, 28% said they had none, 29% had up to one working day, and 17% had up to two working days of training.
12. All Roles, Current Studies or Training (1,089) 
    When asked their current training or study; 31% of respondents were engaged in work-based learning, 6% in Further education and just five people reported that they were on an apprenticeship.
13. All Roles, Highest Qualification Level (1,174)
    Respondents, both paid and unpaid, were asked about their highest qualification, 1,174 answered. One in five, some 20% held an NVQ level 7 qualification. However, 40% either selected “None of the above”, or “Other”. It is not known what ‘other qualification they actually have. All qualifications have an equivalent NVQ level, so it  may actually be because they didn’t know what NVQ level their qualification was.  A look up table was provided, to for example, convert and ‘A level’ to and NVQ level, but it is possible that participants didn’t use this or didn’t understand it. In any future survey how qualifications data is collected, including using NVQ levels rather than the actual name of the qualification (degree, O level/GCSE, HNC, A level  will be reviewed).
14. All Roles, Knowledge and Experience (1,171)
    Respondents, paid and unpaid were asked to choose three areas of work that best described their knowledge and experience. The most reported areas of expertise (paid and unpaid) were surveys and fieldwork (27%), practical land-based work (26%), and ecology (23%).
15. All Roles, Barriers Facing the Workforce (1,158)
    Respondents were invited to pick the three most significant barriers facing the nature workforce in Wales. Of all the responses, the most cited were lack of long-term funding 660 people (58%), followed by insufficient project funding 350 people (30%), low prioritisation of nature 350 (29%), and loss of experienced staff, over 220 people (19%).
16. All Roles, Home Address (1,173)
    Reflecting the place of work; the most common home locations were Pembrokeshire (14%), Gwynedd (9%), and Powys (9%).
17. All Roles, Age (1,148)
    Taken together, paid workers and volunteers; the age range of respondents was from 20 to over 80 years old. 41% of respondents were between 50 and 69 years old. 13% were over 70 whilst 10 % were under 30 years old. Further analysis will look at paid workers and volunteers separately, to identify if there are differences between the age profile of those in work and those in voluntary work.
18. All Roles, Gender Identity (1,162) 
    54% of respondents identified as female, and 43% as male.
19. All Roles, Health Conditions (1,157) 
    22% of respondents reported having a long-term physical or mental health condition lasting or expecting to last longer than 12 months
20. All Roles, Welsh Language Ability (1,164)
    12% of respondents said they were fluent in Welsh.  11% could speak a ‘fair amount’ of Welsh.  66% of respondents said they could either say a few words or could speak a little. 11% said they can’t speak any Welsh.  
21. All Roles, Ability to Work in Welsh (1,164)
    In line with responses about ability to speak Welsh, 24% of respondents said they could perform their role through the medium of Welsh.
22. All Roles, Ethnicity (1,152)
    1,152 people chose to respond and 28 people chose not to respond to this question. Of those that responded, 98% identified as White. 16 people specifically identified as one of the following groups: Asian, Black, Arab or Mixed/Multiple Ethnic groups.  A further six people identified as any other ethnic group.
23. Response Rate and Data Quality
    Due to the open nature of the survey, no formal response rate can be calculated. However, the volume of responses (1,180) provides a solid basis for analysis. The first question has 1,180 responses and the lowest response was the local authority based  location of paid work (680 responses)

That said, approximately 649 responses were incomplete or empty. Given the person did not formally ‘submit’, they did not provide consent for their data to be used. As a result they are not included in this report. It is not known why the responses were not completed. 

Survey approach and design

Introduction

1. The survey was voluntary and self selecting. No definition of the "nature workforce" was applied, due to the diversity of roles and lack of comprehensive data. The survey was designed to be inclusive and self-selecting, including anyone who believed they contributed to restoring nature. The approach acknowledged the complexity of the sector, including full-time, part-time, and project-based roles, as well as paid unpaid contributions. The survey was open to individuals living or working in Wales, including those in cross-border or remote roles. It also included those in policy, regulation, enforcement, and operational delivery roles.
2. The survey was developed electronically using Smartsurvey, with input from stakeholders and tested through two small pilots. It was bilingual (Welsh and English), with respondents choosing their preferred language. Respondents were invited to join a stakeholder list for future engagement. Participation was voluntary, and data were stored separately to maintain confidentiality.

Information on those targeted; the nature workforce

3. The survey intentionally avoided any rigid definition of the nature workforce, the inclusive, self-selecting approach reflected the complexity and diversity of roles across sectors. The survey was completed by anyone who thought they were contributing to restoring nature. It was a self-selection process. There was no attempt to define the nature workforce. There were known to be a myriad of roles, with some working full time on nature, others part-time, funded by a range of projects. It was a very complex picture and as a result it was unlikely that all organisations/businesses could be identified. It was considered that larger organisations would find it difficult to accurately report on the nature workforce. Respondents were asked to complete the survey only once, regardless of how many roles they held to avoid duplication.
   
   i.    Scope The survey tried to reach practitioners doing nature recovery on the ground, but also those in policy, regulation or enforcement. Anyone who lives, or works in Wales, acknowledging the realities of cross-border working and remote roles. This flexible definition was appropriate and helped ensure that contributors to Welsh nature recovery were not excluded.
   ii.    Unpaid/volunteering and paid roles, the survey design allowed individuals to report on both paid and unpaid contributions, recognising the importance of volunteering in the sector.
   iii.    Job roles: survey includes full-time, part-time, and project-based roles. The survey was intended to reach a wide range of contributors, including land based roles, practitioners and volunteer management. But also decision makers (e.g. elected members, CEOs), those who contributed to delivering nature projects; people working in communications, volunteer management, finance, fundraising, project management. This breadth reflected the real-world complexity of nature recovery work. Although the survey was not aimed at internal Welsh Government or Senedd staff, the self-selecting nature meant that some may have participated.
   iv.    Volunteering; The data collection included unpaid workers. Volunteering forms a key aspect of delivery for nature. But each person only filled it in once, irrespective of the number of roles. So, an individual who worked for a private contractor doing peat bog restoration, volunteered for their local nature group and was also a community councillor overseeing nature projects locally, filled it in once.
4. Workforce Information collected: 21 questions were developed to cover a range of areas of interest to future workforce development and planning:
   
   i.    Demographics; age, gender, race, disability, Welsh language ability.
   ii.    Workforce information; full/parttime, funding source, tenure of post, category of employer, location, qualifications and training.
   iii.    Nature knowledge; broad role, areas of expertise. Highest qualification, recent training etc.
   iv.    Perception of current workforce planning, issues and priorities.
5. Stakeholder list. Respondents were also invited to join a stakeholder list for future engagement, to build a direct communication channel with the workforce and reduce the reliance on indirect distribution methods. Participation was voluntary. Recipients were asked to agree to be on the stakeholder list and agree that their email address be held for 5 years. The intention was to create a contact list to contact the nature workforce directly e.g more detailed surveys on for example qualification and skills. The data were held in a separate survey and on a separate database. 

Survey design process

6. Smartsurvey: Not least due to resource constraints, but mainly to try and reach as many people as possible Smartsurvey was used. It is an electronic survey tool used by Welsh Government. It was relatively easy to use, to create the survey. The routing function ensured respondents were only asked questions that were relevant to them. The entirely electronic survey was developed with stakeholder input and two small pilots were undertaken to test methodology and questions.
7. On-line only: The survey was available to complete online, including on a suitable mobile phone. It was sent out via email and crucially, people could forward it on to their contacts, extending the reach. It was assumed that anyone in paid work had access to ICT. It was possible that some volunteers did not have independent access to ICT, but they would volunteer for an organisation with ICT. Online was a significant benefit, but also had possible limitations due to ICT coverage and literacy.
8. Bilingual: The survey was bilingual. Each person chose the language of response. However, for some questions, if the user required further information, they were directed away from the survey, to the .gov website for more information. The .gov website was in English only in compliance with Welsh Government guidance.

Method of distribution

9. Timeline: the survey was launched on 18 March. Most responses were received in the first 8 weeks. Following the final communication about the survey closing,  it was formally closed on 18th August.
10. Distribution: The survey was distributed via email to known contacts with a request to cascade it further i.e those with an email address. This was intended to reach a wide audience of individuals. This was a known trade-off of the distribution method. This snowball sampling approach relied on indirect distribution and was intended to reach individuals beyond the Welsh Government’s direct contact list. However, self-selection meant that some employees or volunteers did not complete it.
11. The contacts included ENGOS, umbrella organisations e.g WCVA etc local government and statutory organisations e.g NRW. These organisations included those engaged with delivery, grant recipients, those engaged in policy development etc. Recipients were then asked to forward it to anyone they think might consider themselves part of the nature workforce. This was particularly important for contractors working on nature projects, for whom contact details were not known. The methodology was reliant on those who procure services to send the survey to the contractors they use.
12. A survey of employers was not used due to the dispersed nature of the workforce and the potential burden on HR departments, especially in large organisations (who might not have had information about funding routes, or latest qualifications and training). Also because of the potential burden of completing the survey, again for large organisations with many employees.
13. Duplication: The method acknowledged the risk of duplication, because individuals may have received the survey multiple times. No complaints were received.
14. Publicity; staff in CCRA publicised the survey at stakeholder meetings, conferences and included the link in the email signature. Deputy First Minister issued a press notice announcing the data collection. Given the true population was not known it was unclear how successful it was, save to say, over 1,000 received and completed it.

Sample/response rate

15. Response rate. The response rate could not be predicted or calculated after the survey due to a lack of knowledge about the number of people in paid or voluntary work for nature. The final dataset included 1,180 responses.
16. Participant drop-out: (i.e. attrition) was not known. It was not known if overall response and data quality would have been better with an electronic, or a manual survey, or using HR Departments. The nature workforce was a highly connected one and there were a number of groups and forums which meant it could be circulated by email, rather than knowing a contact address or organization. But the electronic survey relied on access to ICT via computer or mobile phone. This is a weakness and potential limitation, although for the paid workforce it is unlikely that many would be in employment and be without access to ICT.
17. Manual surveys would have needed to be distributed to an unknown and highly dispersed group (without contact details) and then returned by post. Both provides a significant hurdle to participation. HR Departments may not have known who was in the nature workforce and/or may not have had up to date information.  The methodology appears to have been successful with over a 1,000 responses..
18. Not everyone answered every question. There were 1,180 respondents. But for example only  780 respondents indicated they were in paid work. Whilst all of them answered the question about full-time or part-time status. But, only 740 of them answered the question about their employer.
19. Valid Responses/Uncompleted returns:  Only responses deemed valid by Smartsurvey were included; 649 incomplete responses were excluded, primarily because respondents did not reach the end of the survey and therefore did not consent to data use. Of the uncompleted returns it was not known how many of those people ‘restarted’  and went on to complete a return. It is also possible that some opened the survey and realised it was ‘Wales only’ and that it was not applicable to them. Others may have not been able to use the technology or other reasons.

Limitations of the approach used

20. To summarise the above, several limitations were identified:
21. Integrity and Trust: Some engaged with the survey design and pilot expressed skepticism about respondents’ confidence in providing information to the Welsh Government and data confidentiality. The survey underwent extensive survey control processes and review and was approved by Welsh Government Knowledge and Analytical Services (KAS). Over 1,000 people did respond.
22. Low or Patchy Response: The self-selecting nature and indirect distribution would seem to have led to uneven participation by geographic area e.g Pembrokeshire.
23. Use of an on-line survey and Smartsurvey: an on-line survey may have reduced participation. A small pilot was conducted to test the survey and technology. It had been used by other departments, but possibly not using the snowball methodology. Communications colleagues were involved to ensure the survey met guidelines. It was a survey of workforce, and it would be expected that respondents had access to computers and mobile phones. But the survey’s digital format might have excluded individuals without internet access, or possibly more likely, digital literacy. It might have increased the likelihood of partial completions. But it was not known how responses rates or quality would have been affected by being collated manually and/or or though HR departments.
24. Duplication: Multiple survey invitations could not be prevented, and respondents were asked to complete the survey only once. The survey was intended to be circulated as widely as possible, to capture as many people as possible. In the introduction to the survey, the issue was highlighted and respondents asked to only complete it once.
25. Snowball Sampling: This method introduced some uncertainty about who received the survey and potential bias in responses.
26. Data checking: Smartsurvey results were checked manually. In addition, all information produced by Co-Pilot was double checked. 

Analysis approaches (including AI tools and quality checks)

27. The analysis was conducted using Smartsurvey. Data were analysed and reported using that and Microsoft Co-Pilot. The results were only about those who responded, nothing was recorded about those that did not respond. 

Quality assurance process

28. The simple question by question analysis was initially produced using Smartsurvey. This was a simple process, with automatically generated tables. The report was not very user friendly. So, Co-Pilot was used to create a very simple summary, identifying totals for each question.
29. Quality assurance was a key aspect of the analysis process.
    1. Figures were  rounded to the nearest whole number to avoid a false perception of precision.
    2. It was noted that in questions where the respondent could only provide one response, answers might not add to 100% (i.e. might be slightly more or less)
    3. A download of the base data was generated by Smartsurvey and put into Excel for sense checking. As expected, the number of rows was 1,180. Sense checking, including using estimation were very important tools. The data produced could not just be assumed to be correct, spotting any errors and anomalies was important.
    4. Anomalies were investigated and confirmed to reflect actual responses. The responses were  not ‘incorrect’ they reflect what respondents said, but they were of note in terms of  interpretation and use of the data.
       1. Geographic distribution, by local authority did not follow population in Wales.
       2. Qualifications data; 40% reported an NVQ level 7 but most other respondents did not choose a specific qualification level. It is likely that NVQs levels were not understood.
    5. AI-generated outputs from Co-Pilot were cross checked with Smartsurvey outputs, for accuracy.
    6. Respondents could skip questions, and routing logic, meant the base number was different for most questions. Data were checked to ensure they reflected those that responded as shown in the table below;

Total number of respondents, by question

KAS assurance process 

30. Once the report was written. KAS QA process required that the percentages produced by Smartsurvey be checked manually. The Excel original spreadsheet was then modified to:
    a. Count the number of respondents who answered a particular question and cross reference that with the number reported by Smartsurvey 
    b. Excel was used to re calculate the percentages, based on the number of responses.

Manual checking of question 18: gender

• Check 1; Smartsurvey created the report shown. It reported that 1,162 responded and 18 skipped. 495 selected male and 629 selected female, 38 chose ‘other’ or ‘prefer not to say’.
• The total number of responses was checked; 495+629+38 = 1,162.
• It was clear that the M/F split was fairly similar, with slightly more F and a very small ‘O/PNTS’ response; the reported percentages 42%/54%/3% reflected that.
• A detailed calculation was made for each; 495/1162, 629/1162 and 38/1162. These calculations provide the same figures as Smartsurvey; 42%/54%/3%.

Manual checking of question 8: location of unpaid work

• Check 1; Smartsurvey created the report shown. It reported that 1,092 responded and 88 skipped.
• The total number of responses was checked; 448+234+373+24+13 = 1,092
• It was clear that the ‘answer 1’ was about twice that answer 2  and answer 3 would be a mid point between answers one and two. 400 out of 1000, is 40%, so 448/1092 is about 40% and so on.
• An exact calculation was made for each; 448/1092, 234/1092 and 373/1092 etc. These calculations provide the same figures as Smartsurvey.

31. This checking was replicated across all 22 questions. Statements created by Co-Pilot were also checked. 

Report writing

32. A summary report of the data was written, as noted above some was generated by AI. A methodology and lesson learn section were included. 

Data publication

33. The survey was anonymous. Confidentiality was always a concern to recipients and the stakeholder list information was be kept separately from the survey responses. The survey itself and privacy notice was discussed at length with KAS colleagues. Figures were rounded to the nearest whole number, because including decimal places might provide a false sense of accuracy or precision, the data were only of those who responded. Whilst very few people from black and ethnic minority communities had responded it was however crucial to begin to tackle the issues of inclusion to be able identify this issue, albeit with small numbers.
34. The report, this detailed methodology and lesson learnt would published, bilingually on the internet. All those that provide a contact email address would be sent the report.
35. This initial report is intended to be the first of several reports on the data , more would follow on different aspects of the survey; such as cross tabulation, reports the paid   and unpaid workforce etc. each will be made public.

Early reflections on the methodology, risks and lessons learnt

1. Direct-to-Individual Approach
   The decision to bypass traditional employer-based/organisation based surveys and engage individuals instead directly was a success.
2. There was a strong response of 1,180participants, paid and unpaid; a far higher response than anticipated.
   1. for the first time data reflects individuals. Each person was  only counted once, even if they worked/volunteer for several organisations.
   2. HR departments did not have to compiledata.
   3. Contact details from participants have been collated.
3. However, some reflections are;
   1. a further 647 responses classified as invalid. They require further analysis; some were entirely empty/bank others have one or two questions answered.
   2. response was uneven. Some geographic areas had very few responses, while others—particularly Pembrokeshire—seemingly overrepresented. Further detailed analysis may reveal greater insight.
   3. Participation was voluntary and some employers seem low as a whole; e.g despite Natural Resources Wales (NRW) employing approximately 2,000 staff and having a focus on nature, very few responses were received from this organisation.
   4. Some individuals received the survey multiple times, while others—particularly contractors and those not on known mailing lists—may not have received it at all. Still, the strategy relied heavily on goodwill and informal networks. As a result, some key groups were missed, and response levels varied widely between organisations and regions.
4. That said, overall, this confirmed the approach was effective in reaching a wide range of respondents. But it didn’t reach everyone (with gaps in reach to some  organisations, employees and geographic areas).

Survey design 

5. The bilingual electronic survey, developed with stakeholder input and distributed via email with a request to cascade further, enabled broad participation. The open invitation to anyone who considered themselves part of the nature recovery effort in Wales allowed for inclusivity across roles paid and unpaid.
6. But around 647 responses were incomplete/empty. This suggests that while the design was accessible, some respondents may have encountered barriers to completion—it is not known if they realised they were not eligible or due to survey length, clarity, or other factors.
7. 100% response may never be achieved, but potentially future surveys may benefit from more targeted outreach and possibly improved survey design,  for example for the qualifications data, explaining NVQ levels to encourage more balanced and complete participation.
8. Responses to the question on NVQ level appear to indicate the question was not well understood. The geographic distribution does appear to not reflect population.

Distribution and duplicate emails 

9. The method of allowing the survey to be forwarded and asking individuals to complete it only once proved to be a resource-efficient way to gather data. The simplicity of the approach likely contributed to the strong response rate.
   But the lack of control over who received the survey may have contributed to overrepresentation in some areas and underrepresentation in others.
10. While it was acknowledged that some individuals might receive the survey link multiple times, this was anticipated and addressed in the survey introduction. There is no indication that this caused confusion or deterred participation. 

Response, inclusivity and reach

11. This risk of low response did not materialise. The volume of responses exceeded expectations for a voluntary, self-selecting survey. While the data is not representative of the entire workforce, it provides a robust baseline and valuable insights into the nature recovery sector.
12. The open, self-selecting nature of the survey encouraged participation from a wide range of individuals, including those in non-traditional or multiple roles. This was a strength of the approach. It is not known if some perspectives were underrepresented. This limits the ability to generalise findings across the entire nature workforce. However, the low responses from some areas and incomplete responses reduced the usable dataset. Clearer guidance and/or improved survey design in future iterations to reduce drop-off may be needed.
13. The survey successfully gathered data on demographics, employment, volunteering, qualifications, training, and workforce perceptions. The inclusion of an agreement mechanism for future contact was also effective, laying the groundwork for ongoing engagement.
14. The volume and quality of completed responses provide a solid foundation for analysis; the data allow for basic tabulation and, in many cases, cross-tabulation by role, geography, demographics etc. But the patchy coverage means that some analytical opportunities were constrained. These limitations will be clearly acknowledged in any interpretation or presentation of the findings.

Use of technology 

15. The survey was a new initiative for the Climate Change and Rural Affairs (CCRA) portfolio. Despite initial uncertainty, the response rate was strong, with 1,180 individuals participating from across Wales. This suggests that the approach was effective in reaching a broad audience and generating meaningful engagement.
16. No significant issues were reported regarding the survey platform (Smartsurvey) or question clarity. The pilot testing and involvement of communications colleagues appear to have been effective in ensuring accessibility and usability, but the number of incomplete/unusable responses warrants further investigation.

Integrity and trust

17. There were concerns that individuals might be sceptical about sharing information with Welsh Government. However, the high number of responses and the breadth of participation across sectors and roles indicate that these concerns were largely mitigated. The approval by survey control and proactive communication by staff likely helped build trust.

Conclusion

18. The methodology adopted for the Nature Workforce Wales Survey 2025 was broadly successful. It overcame many of the challenges posed by a dispersed and complex workforce and delivered a valuable dataset. However, the uneven response distribution, underrepresentation of key organisations, and the volume of incomplete responses highlight areas for improvement. Future surveys may benefit from targeted outreach, improved survey design, and additional support to ensure more balanced and complete participation.
19. Overall, the risks identified prior to the survey were well-managed, and most did not materialise to a degree that undermined the survey’s success. The approach proved to be a resource-efficient and inclusive method for gathering baseline data on the nature workforce in Wales. The results now provide a foundation for future workforce planning and policy development.