There are some basics that you need before you start:
Connection with the users (we had an insider among digital humanists);
Digital collections and metadata collected and shareable (we had an open data portal);
Support and resources to develop (we had moral and financial support from the management).
To build value out of the large digital collections in our library, we aim to build a virtual lab space to give access and guidance to journalists, academics and researchers.
1.
Find analogues & good examples
Who do you want to follow (GLAM Labs communities have done this in other libraries)?
Gather ideas (interfaces and outcomes in other labs).
Tips
Ask for recommendations: colleagues in other libraries who tried similar things are happy to share their successes and failures to learned from. We’re in this together!
Talk to the creators: we talked to the creators and managers in other labs. Learned that practically other labs often kept an experimentalrole in their institutions;
Summarise: we created a document that contained different examples, screenshots and a vision forward.
2.
Service design
Figure out who, what, why want it;
Who are the user groups (e.g. students, scholars, journalists, creatives);
Learn their expectations (nice visuals or just the data or something in addition to data);
Build the network in the process (get people interested).
Focus on one group: we focused on researchers because they have some skills of data analysis and already want to use the data. This is good because they already know what to do.We already hadgood contacts with the research community of digital humanities, so it was easy to start;
Listen: we did interviews with 13 key users and organisers who we thought could use our data. For example, to access the data, many said that no complex user interface is needed. They can do the work.
Build a network: we also asked for recommendations on who else we should talk to. You can snowball out to find new communities;
Keep contact: in discussing with the key users, we also made new contacts. Some were positively surprised about what we were planning to share and are ready and eager to work on them.
3.
Legal analysis
Learn what material you can actually put out there for researchers and what conditions you must meet;
What are the limitations set by intellectual property and personal data protection law?
Is there a specific procedure that you or the researcher must go through to give and to get access to data?
Gather an overview of what you have learned – on the one hand, what your target group needs (service design) and, on the other hand, what you actually can do (legal analysis);
Make a plan and stick to it;
On the way there are many details to solve (e.g. how a button should work on your website), make sure it does not distract you and slow down work on the main goals.
Tips
Be realistic about the resources you have – both in terms of finances and personnel;
Map the stakeholders within your organisations – who could have an interest in what you’re doing and who should know about what you’re doing;
Cooperate with marketing from the beginning. They need to understand what and why you are doing. They will later help to reach your audience!
5.
Develop access point & web platform
For all data sets you have, tools you can give for researchers to use;
For examples of how data has already been used and can be used;
To keep contact with the target group – develop and maintain the network (news, events relevant to digital humanists);
Not distract you and slow down work on the main goals.
We planned step 6 in parallel with this step
Tips
This is the most time-consuming step – lots of work must be done in a limited time period;
Have flexibility in the budget – plan extra finances for possible (or rather probable) changes that could not have been foreseen in the beginning;
Need to be well prepared; details matter – figure out what you need from the website, especially what you need the website to do technically. What is a must-have and what is nice-to-have;
Be prepared for diverse input (important to not lose focus) - visuals vs functions;
Divide the development into small sections – every section of the page or functionality include the following steps: plan, design, develop, review, re-develop, test [repeat the last two until you’re happy];
Plan and design keeping in mind the user journey;
While you work in sections, you still need to keep in mind the general picture – one section must support the other;
Check-check-check – everything and all the time. Do not presume that the task is done or that it is done the way you expect or was agreed on.
6.
Develop & prepare data & tool
The data should be easily findable and accessible (see FAIR principles)
Preparing the data is a lot of work (need data science skills to do this)
Figure out the endpoint (what formats do you want by the end, how will you update)
We planned step 5 in parallel with this step
Tips
Pick one or two datasets to begin with and prepare them for publishing fully. You will learn on the way how to prepare other datasets;
Study the data and tools in other organizations, there are some models that could be easily followed for your data while giving fresh insights for your users.
7.
Present the results
Present to the board and to the internal audience of your organisation;
Show the final product and what it can do or what can be done with it.
Tips
Keep in mind the goal. You want to both show what is done as well as point outpossible future steps;
Show the value of what was created to the makers behind the datasets. The digital resources that you can share were made by diligent work by your long-time staff. Give the spotlight to them and help them enjoy the fruits of their labor.
8.
Launch & gather audience
When the work is done, we need to find the community around it. Show the audience that this is a product for them, listen and discuss.
Tips
Take time to prepare the launch (event, invitations);
We would recommend physical event – it allows people to network;
Launch is an introduction and a first date. Contacts and trust take time to build up.
9.
Take care of audience & start again
The lab is built in small steps (don’t expect it to ever be finished);
Keep the users involved (personal contacts matter, users have many other places to go);
Think of the next steps;
Be ready for the users when they come.
Tips
Think of the next steps when finishing the project. What do you need for it, how can you bring the current project and the community to help;
New concepts can be difficult to take up, don’t expect the users to arrive straight away. Many of them also work on a project basis and need to plan their steps ahead a year or two.
