How to increase online trust and user experience for legal services

This research looks at online trust, and what features affect trust, both perceived and actual. A set of heuristics were drawn from a systematic review of 160 papers on trust. The heuristics were then used for an audit of 40 UK legal websites. Five mock legal websites were created using features found on real legal sites. These mock websites will be used to measure the impact of the different designing features on trust. The sites will then be tested by both the solicitors and end-users.


INTRODUCTION
Online businesses success is determined in part by the consumers' ability and willingness to trust the supplier. There is a perception that there is a risk involved with information disclosure. Therefore it is essential that the user feels that the entity collecting personal data can be trusted (Beldad et al., 2012). Trust is an important component during a transaction when there is any uncertainty, anonymity or if the user feels any lack of control, such as sharing sensitive personal or financial information (Bhattacherjee, 2002).
Trust is especially relevant to legal services, only lawyers can foster this trust, and by delivering a trustworthy service, they are building for the client, not just trust in their firm but trust in the legal system (Mah, 2005). People who need help seek a good lawyer and the most efficient way for a lawyer to gain new business is to signal trustworthiness. Lin Wu et al. (2016) stated that further research is needed where heuristics can be used to evaluate trust in real-world situations and effective risk communication (Lin Wu et al., 2016). Heuristics are simply short cuts to ease the cognitive load of decision making (Lin Wu et al., 2016).
In the 1990's Jakob Nielson and Rolf Molich developed a heuristic evaluation this method was designed to be an informal method of usability analysis and is still widely used by software developers and UX Designers, he found that without heuristics individuals would only find 20-51% of the usability problems (Nielsen and Molich, 1990). A systematic literature review was completed, thereby identifying evaluating and interpreting results relevant to a particular research problem (Kitchenham and Brereton, 2013). The research question applied to the systemic review was "what features facilitated online trust" so the descriptors adopted were "trust" and "online trust" in Boolean search in titles, abstracts and keywords. For publications starting in 1967-2020, using The University of Sunderland's, Library services, limiter was "full text", and source type was "academic journals". With a subject of "internet", "information technology"," trust". This identified 1,451 articles. The inclusions criteria were (a) articles with a clear description of features that affect online users trust, (b) articles that explained what trust is. After reviewing the paper 160, were used for the full study, by reviewing the literature, eight heuristics where compiled, which has been shown to affect online trust.

TRUST HEURISTICS
• Image content should contain people, they should not contain any negative images (Riaz et al., 2018), and there should be no missing images (Cyr et al., 2009;Clare AM Sutherland et al., 2015;Joyner et al., 2018;Schultz et al., 2018) and images should be of a good quality (Lian and Yu, 2019) • Trust transfer is an effective mechanism in brick and clicks cases (Chun-Chia Wang and Chang-Cheng Sie, 2012; Xiao, Zhang and Fu, 2019), by being able to demonstrate that the firm is a real company by providing an office telephone number (Beatty et al., 2011), a real address, directions of how to find the office, display images of real staff members and premises, will increase trust and create synergies between the physical and online part of the business (Nicholls, 2004;Doong, Wang and Foxall, 2011;Lee, Chung and Lee, 2011;Mesch, 2012;Toufaily, Souiden and Ladhari, 2013) • Sites should include visible trust features, also known as trust cues, that reduce the risk for the user, for example, 3 rd Party recommendations, accreditations, awards, strong brand and reviews ( (Grimes and Marquardson, 2019), furthermore, there is also a strong link between satisfaction and loyalty (Faisal et al., 2017). Sites should offer a multi-channel approach to improve online visibility (Lian and Yu, 2019) and have good SEO so that they can be found easily • The firm should show benevolence, as is one of the three key dimensions of trust (Murphy and Blessinger, 2003;Yousafzai, Pallister and Foxall, 2003;Hung, Cheng and Chen, 2012;Oliveira et al., 2017)this could be done by offering free advice (Brengman and Karimov, 2012)

APPLYING THE METHODOLOGY
A total of 40 real legal sites were tested with the heuristics, which was moderated by a score sheet. Each Heuristics was awarded a score from 0-5 which was then added to a table so that areas that had common faults could be understood. Each site took approximately 2 hours, 7 minutes to evaluate.

The trust features found in Legal Sites
The best sites had images of people interacting, the words and images wherein harmony and care had been taken in presenting staff images in a coordinated way. All the images were Next-Gen formats JPEG2000, JPEG XR or WebP. The images were the correct size and optimised for the web.
The branch was able to show brick and clickability by giving a real address along with contact information, displaying information about car parking or bus numbers. They showed opening hours and which team members worked at the branch alongside a google map, offering directions.
Visible trust features such as Lexcel awards were evident and that the Solicitors Regulation Authority regulated many sites. Firms showed third-party accreditations and had a strong branding strategy.
It was apparent which services they offered, and there was a good user flow. The menu was nicely designed and did not use legal terms. Modern sites had excellent staff profiles which showed what awards or accreditation each staff member with a comprehensive case history and cv and linked to their linked in profile.
Modern sites were Search Engine Optimised (SEO), with outstanding security and had correctly installed the SSL certificate to all pages, all links where to HTTPS pages. The JavaScript was free from any vulnerabilities.
Benevolent sites displayed mentoring schemes and information about fundraising for chosen charities. They also offered free advice and had video links to advice as well as free PDF downloads or had useful information blogs. Some offered helpful tools; others offered fixed-fee pricing structures or payment options. Most of the high scoring benevolent sites had caring mission statements.

The worst errors found on legal sites.
The worst legal sites had a disconnection between the text and the carousels images. Some sites had negative images, such as parents fighting. Research has shown that emotionally negative images will cause people to skip webpages. One site had an image of physical abuse. A common problem found was missing personnel images, the generic blank person image should be avoided not being able to see a person's face, affects users trust and the solicitor with the missing image is not likely to be contacted. Instead of using a blank figure, use the company logo or image of the office, this looks less intimidating. Some sites had blurry images.
Some sites had used legal terms on the navigation menus, such as "Private Client" a user would not understand the meaning, making the site unfathomable. Another site had 1076 links, by adding too many links to a site or too many buttons and menus, the site will be confusing to the user.
The worst sites had no staff profile at all, so it was impossible to gauge if they were capable solicitors.
Some of the older legal sites felt bland and textheavy compared to the more modern interactive sites; new sites need to offer the user a joyful experience. Some sites had low scores for search engine optimisation, which would make them hard to find. A number of the sites had colour palettes where questionable.
The basic security of legal sites is worrying, over 1/8 of the tested sites where insecure, some had no SSL certification, others linked to insecure sites. A few had known JavaScript vulnerabilities. This should be investigated further in future research.
Firms that were not benevolent did not show any charity work or any pro-bono and were vague about pricing. Benevolence is an essential part of the user's trust; users must feel that the company will act from them rather than just maximise profit for the company.

CONCLUSION
The Image scores were the poor across legal sites for all categories.
Most firms were good at showing the firm's thirdparty accreditations. Brick and click's scores were also excellent, with most firms having a real business address. However, many legal firms did not invest much effort in showing their staff's ability, which was a surprising result given that for legal firms, the quality of their solicitors is paramount, as they are a service. Security was also an issue and likewise, benevolence. The modern score included SEO, which is vital to ensure that the legal site is found via search engines. Future work to be done later this year involves user testing of the trust heuristics to explore the different impacts on the users.