Statistics.gov.scot improvement project: discovery user research report

Overall findings

The following tells the overall story of the user research by bringing together the various analyses and themes of each session.

Finding 1: A confusing user interface amplifies performance issues.

The unintuitive design of the site not only confuses users but also makes them more sensitive to the platform’s poor performance. When users struggle to navigate the site or complete tasks due to the complexity of the interface, the frustration caused by slow loading times and frequent errors is magnified. This is the case for those seeking data and those attempting to publish it. This results in a compounding negative experience, where each flaw reinforces the others, leading users to abandon the site in favour of alternatives, or deprioritise data publication on this site.

Finding 2: Complex language and outdated content dissuade effective use and undermine trust.

When users encounter technical language that they do not understand, outdated elements, and/or find data that hasn’t been updated, their confidence in the reliability of statistics.gov.scot diminishes. This erosion of trust is a significant barrier to the platform achieving its stated purpose of providing open access to official government statistics.

Finding 3: Ineffective search functionality pushes users to external sources.

The ineffective search tools on the site force users to spend more time and effort locating datasets, often without success. This frustration drives users to seek information from other sources, including popular search engines and freedom of information requests. Reliance on these approaches could be reduced with a more efficient search system. The inability to find data easily also contributes to the perception that the site is incomplete or unreliable, further decreasing user trust.

Finding 4: User frustration is a barrier to unlocking the platform’s potential.

Despite the challenges, most participants recognise the potential of the site as a source of official statistics and data. However, the current user experience prevents users from fully realising this potential. Frustration with the interface, performance, and basic functionality discourages users from exploring the site more widely, recommending it to others, or prioritising it as their primary resource. It is also important to recognise data publishers (and indeed the SG Open Data Team who provide vital administration and support) here, as users of the site who experience the same or very similar frustrations; if it isn’t relatively straightforward to publish and update data then encouraging use of the site will remain an uphill battle.

Concluding remarks

Throughout the research, two questions were repeatedly posed by participants: what is the site for and who is it for? For participants, even those with years of experience using the site, these were rhetorical; the implicit answers being ‘I don’t know’ and ‘not me’. While this programme of user research has given the project team a clearer sense of who and what the site should be useful for, it has also shown that the current site works best for a niche and largely absent user group. This is partly the result of quite reasonable and admirably ambitious projections about the widespread acceptance of open data and associated technologies, made in 2015/16, which have not been realised. A helpful illustration of this point is a Google search for SPARQL, which returns sites that are 5-15 years old, with statistics.gov.scot high on the first page of results.

Imagining who might really thrive on statistics.gov.scot lets us see just how narrow the site’s appeal has become since its inception. The ideal user then, is a statistician who is fluent in SPARQL, wants 5-star linked open data and appreciates N-Triples, but still finds a use for the on-site dimension-locking and datacube slicing features, and knows the exact datasets they need before coming to the site. We did speak to several statisticians, but none of them came close to fitting this description. The few participants who mentioned linked open data, interoperability, and/or reusability did so in passing, to position pursuing these as secondary to getting the basics right, e.g. publishing up-to-date, large datasets that are easily and reliably accessible on the web.

Of course, if the site met the very specific technical needs of the ideal user in addition to meeting the largely simpler needs of the much broader user base identified through this research, then this would not be a problem. Unfortunately, what should be primary, core user needs – to quickly and easily find, access, publish and update data – are supplanted by the ‘nice to have’ wants of a relative outlier.

statistics.gov.scot has the potential to become a trusted and valuable tool for a wide range of users, but this will not happen if the fundamental user needs discussed here are not met.