19 Ludmila Polshikova of the Russian Justice Initiative also mentioned that many of the photos she receives are large in size: 4 to 9MB per photo. To store and work with them, she or members of her team, have to compress the photos and compile them into one document, which can be a timeconsuming process. Keeping track of where external data storage sources are kept was also mentioned as a problem. Multiple people mentioned that they kept encrypted backups of sensitive data on external hard drives in places outside of their office, for security reasons. Though having multiple copies seems like a good idea, knowing at any one time where all of those copies are is important - in particular as it was mentioned that sometimes those copies can go astray, especially as the number of external data sources increases. MANAGING EXPECTATIONS For human rights defenders who are engaging with a new tool, managing their expectations was identified as an area of particular concern. Groups who decide to develop their own tool often underestimate the timescale – for example, Rick Bahague from the Computer Professionals’ Union (CPU) based in the Philippines recalled a case when in 2004 CPU was approached by a human rights organisation to help them create a system for their fact sheet. Developing that tool took nearly three years. Relatedly, knowing what the data that is collected through the tool can (and can’t) be used for is important; for example, SafeCity, a crowdsourced data initiative from India, are aware that the data that they collect can’t be used in legal cases because of the fact it is crowdsourced. Patrick Ball identified the importance of knowing what your data can be used for in terms of drawing conclusions or statistical analyses, and when that data is unsuitable for the kind of analyses that would be more useful. INTERNAL RELUCTANCE Multiple intermediaries and HRDs highlighted the challenges associated with getting a new tool to be used; as Kody Leonard of The ISC Project mentioned, “people like to stick with what they know”. Without clear incentives for engaging with a new tool, people identified that it can be difficult to get a new tool to be used. Mourad Dhina admitted that a “good amount of lobbying had to be done internally” to convince people within the organisation that they needed better tools, and Janvier Hakizimana mentioned that it can often be hard to get buy-in from management to spend time and resources on getting used to a new tool. One strategy for combatting this, cited by Indira Cornelio was focusing reasoning for the new tool to be protecting the safety of the people they are trying to help, rather than defenders’ own safety. COMMUNICATION When developing and choosing the tool, effective communication between those building the tool and those using the tool seems to make a huge difference to its success. When people with very high levels of tech capacity but low understandings of the realities of human rights defenders work are brought in to advise on their tech use, interviewees who worked with HRDs cited occasions of a “culture clash” between people with very high levels of technical literacy, and activists or HRDs with lower levels. One interviewee who works between technologists and HRDs mentioned that people coming from hackerspaces who are very dedicated to security tools often recommend switching to open source tools which often less userfriendly than proprietary equivalents. This approach can backfire, as outlined below in the Digital Security section. To help address this issue, both intermediaries and developers we spoke to emphasised the need to prioritise understanding the issues involved on a human level prior to tech development. Understanding the sensitivity of the issue at hand can seen in everything from the tools implementation, training styles, and the design choices made in the tool itself. For example,, Mourad Dhina, Executive Director of Swiss organisation Alkarama, stated that his organisation had changed the language initially used to register a violation in a database from “create victim” to the more appropriate “register victim”. Interestingly, there was some overlap between information management and communication tools, and some cases where tools were used for purposes other than the one for which they were designed. For example, an activist collective who supports parents whose children are missing in Mexico, used Gmail to archive information by sharing a common Gmail account.