Insights continuous publication case study

Insights: the UKSG Journal switched to a continuous publication model in early 2018, following a decision by the Editorial Board in November 2017. The prospect of a switch had been discussed a number of times previously. In the event, the decision was quickly made and the transition largely smooth. Given UKSG’s position as a membership organization including many publishing members, it was considered that a case study of the process would be useful in helping members and other publishers to consider and execute a similar transition.


Background
Insights: the UKSG Journal began life in 1988 as Serials, a benefit of membership of UKSG (a community association for scholarly publishers, libraries and other individuals and organizations in the knowledge community). The journal, with three issues a year (March, July and November), was retitled in 2011, moved to an e-only format in 2012, and in 2014 became platinum open access (i.e. content is immediately available without subscription and without author payment -publishing costs are paid in their entirety by UKSG). On a number of occasions, the question of whether the journal should switch to continuous publication had been raised, and in November 2017 the decision was taken to go ahead with this change. This article is based on a series of conversations with the Insights Editorial Board.

What inspired the Editorial Board to consider the change to continuous publication?
The Insights Editorial Board (IEB) began to consider transitioning to continuous publication for a number of reasons. They wished to: • accelerate article publication, particularly for time-sensitive content Steve Sharp, Head of Library Resources at Sheffield Hallam University and Insights Co-Editor, explained that 'In the fast-paced, ever-changing environment of scholarly communications, the three-times-a-year format was becoming too restrictive. We were losing momentum, particularly in terms of engaging with authors. We had lost a sense of timeliness, particularly for topics emerging after September (the deadline for our November issue) -which might not then get published until the following March.' Graham Stone, Senior Research Manager, Jisc, and IEB member, added, 'We also used to publish a lot of papers that had come out of our annual conference and the period between that event and the copy deadline for the next issue could be very tight, depending on the conference timing. We wanted to reduce that pressure and risk.' • maximize the potential of the online medium Joanna Ball, Associate Director, University of Sussex Library, and UKSG Publications Officer, pointed out that 'Three issues per year was a hangover from the print world.' -32, 2019  • spread the workload of editing, proofreading and producing the journal throughout the year Stone said, 'We wanted to relieve the pressure on our staff and Co-Editors at bottleneck times,' and Ally Souster, Publications Associate, UKSG, agreed: 'The previous issue-based publication schedule resulted in spells of intense work, not least because articles were often commissioned in clusters following IEB meetings, leading to a flurry of work immediately afterwards. Being a part-time staff member gave me some flexibility to cope with this ebb and flow, but the work was largely reactive. UKSG is a very small organization, so there was no one to provide cover -meaning it wasn't really possible for me to take a holiday during the month or two preceding an issue publication date. I hoped that there might be a smoothing out of this, with suggestions made throughout the year to help spread the submission dates.

Insights
Lorraine Estelle, Director, COUNTER and Insights Co-Editor also found the pressure at certain times was far from ideal: 'I couldn't bear it when the big batches of PDFs came through [for editorial approval] -I never felt I could give enough time to each of 12 articles in one go.' • increase the visibility and impact of each individual article Estelle went on to explain: 'We didn't have time to promote things. I looked back at content we'd published, e.g. around the time of the Finch report [on Expanded Access to Published Research Findings] -we'd published some really key pieces and hadn't been able to give them the visibility they deserved.' Sharp added: 'Looking back, we could see we'd quite often been the first journal to publish on a topic, but we hadn't been able to establish a presence around that, and another paper in another journal would get all the limelight. We wanted to be able to give all our authors much better recognition.'

What benefits was the change designed to deliver?
The IEB hoped to: • improve the journal's reputation Sharp explained that this included meeting authors' needs for more timely publication, while being more selective about what we publish: 'For example, we're publishing fewer conference proceedings now, as there are better places people can communicate that material.' • develop a better service for, and sense of engagement with, authors Estelle said, 'We can work with authors much better on their chosen timeline,' and Sharp added: 'If they'd taken the time and effort to get something ready quickly, we could reciprocate by publishing quickly.' Ball summed up this new relationship: 'The whole thing feels much more of a partnership with the author -you get the article ready, and discuss when the best time is to publish it.' • build the journal's usage The Co-Editors explained that with continuous publication, there was 'better ability to promote articles, each in its own space' (Estelle) and 'improved timeliness in publishing articles on hot topics, and tying them in with events -UKSG's, or others'. (Sharp) • improve the quality of peer review 'A more managed workflow means we can work through articles as they come in, and give them the attention they deserve, so we're doing a better job. We are fresher, and see better what we can do to improve an article or bring it to life, e.g. choosing better "pull quotes". Both authors and readers benefit from the higher quality service and a "result" that's more meaningful in terms of conveying the authors' work.' (Sharp) 'we'd quite often been the first journal to publish on a topic, but we hadn't been able to establish a presence around that' 'The whole thing feels much more of a partnership with the author' • increase the currency, quality and structure of content for readers Sharp made an interesting comment about people's time management: 'When you give someone a fixed deadline that's three months away, they put it on the back burner. When you say you'll publish as soon as it's ready, they write it sooner.' This made planning difficult in the past, and Stone added 'We used to have some issues that were thinner.' Estelle explained: 'We were very schedule driven, and that had the potential to impact on the quality if we had to "fill up" an issue, or rush too many things through at once. Not having the artificial deadline means we can be more quality conscious, and have more time to go back to authors for clarifications.' Sharp noted, 'If we still want to bring out a set of articles together, because they're themed, we can still do that.' • improved ability to forecast and manage expenditure through the year Continuous publication would enable us to 'think more strategically now -not just one to two issues ahead, but planning future years' budgets, and whether we have the staff as well as the budget to expand copyflow.' (Stone)

What risks were considered?
The IEB considered the following risks:

Will we run out of content?
This was a relatively low-level risk, as the journal's platinum open access (OA) model, i.e. OA publishing without article processing charges (APCs) had made it an increasingly popular choice for authors in the field. It was anticipated that this, combined with increased marketing and continued commissioning where appropriate, would result in sufficient copyflow.

Will prospective authors be put off by the lack of a specific publication date, and associated timelines?
It was considered that this was a low-level risk. Insights had good time-to-publication prior to the transition, so authors would have no expectations of us being anything other than an efficient partner in helping them publish their work in a timely manner. It was also agreed that for any articles that authors might want to publish in a specific time frame (e.g. to align with a specific event), we would be happy to set a specific publication date and plan the publication process backwards from that. We had also always been happy for authors to publish articles as preprints in relevant repositories (e.g. LISSA), and could emphasize this option to any authors seeking more control over publication timelines.
No authors have indicated concerns along these lines since the transition was made.

Could we end up with too many submissions, requiring either that we exceed our calendar year budget for APCs, or that authors wait (potentially a long time) until the next calendar year's budget?
This was also considered to be a relatively low-level risk, as the journal had not historically had a problem with too many submissions. It was acknowledged that the attractions of continuous publication (e.g. quicker publication times) combined with the platinum OA model and increased marketing could result in increased submissions, but it was considered unlikely that these would create too great a backlog. The journal also had a reputation for efficient communication with authors, enabling the IEB to be confident of managing expectations. There was some discussion of whether authors seeking speedier publication of articles being held on the basis of budget could be offered an 'expedited publication' fee, i.e. covering the APCs themselves, but the bias of any such model against authors in the later part of the year meant this option was not pursued.

Would the loss of issues and the lack of continuous pagination make citation or cataloguing more difficult?
There was some discussion of this and again, it was decided that the risk level was low, particularly given that Insights would not be the first journal making this transition, and librarians and authors would presumably already have created precedents that our readers could follow. On the citation point, the Ubiquity Press platform on which Insights is published includes a 'how to cite' statement at the end of each article's abstract, providing authors with ample guidance.

Was the decision unanimous, or was there work involved in achieving consensus?
The first time the question of continuous publication was raised, it had not ultimately been approved. The journal had only recently transitioned to OA and moved to a new platform; continuous publication was seen as one step too many. 'Concerns were raised about whether we were adopting a relatively new idea too quickly.' (Stone). However, the second time the idea was raised for serious discussion among IEB members, the recommendation to switch to continuous publication was unanimously accepted.

What objectives/targets did you set or could you have set to enable measurement of the effectiveness of the process?
We explored targets including: • workflow efficiencies -would we be able to publish more or fewer articles with the same editorial/staff capacity?
• turnaround times (days from submission to publication) • usage (e.g. reads per day) • submissions • quality (of process/outputs) • author satisfaction • social media engagement • citations.

What steps were involved in implementing the transition?
The IEB recognized that such a change would result in additional work, over and above normal publication workflows, to ensure a smooth transition. The Publications Associate and Co-Editors put together a transition plan to ensure that articles would be published at regular intervals throughout the year, and that articles would be published on the most effective date. The steps taken were:

Communications to existing authors
Updated information was e-mailed to authors of articles already in the review stages to let them know that their article would now not need to wait until an issue date, but would be published as soon as it had completed the publication process. Authors were invited to share any preference about the timeline for the publication for their article (e.g. to coincide with a specific relevant event). The benefits of continuous publication were emphasized, and there was no negative response from authors.

Communications to potential authors
The general information provided to new authors via the website and e-mail was similarly updated.

Background analysis to influence plans
Existing data around turnaround times (from submission to publication) were re-analysed to provide accurate data for planning publication schedules. (It was determined that turnaround times had been artificially inflated by requirements related to issue-by-issue publication).

Schedule planning
Given the journal's budget, the IEB agreed to publish 45 articles a year, including three Editorials which would provide a periodic summary of the content being published. A publication schedule was defined (approximately one article to be published per week) to ensure that articles would be published in a sufficiently wellspaced fashion to enable the benefits around workflow and promotion to be realized.

Process review
Processes for tracking accepted and published articles in house were reworked to ensure all authors in the process would be clear on the timelines relating to their publication. Ubiquity Press had been used to blocking out production time for three batches of articles per year and would now be receiving individual articles on a more frequent basis, but this did not present a problem as it was 'closer to the model they prefer for most other clients'. (Souster)

What happened next?
The decision to move to continuous publication was made in November 2017, and it was intended that the transition would be made in spring 2018. The process was actually completed in a month, with the work involved in implementing the steps above estimated as having taken about a day.
Some unanticipated challenges included: • agreeing the publication phasing for articles already in the editorial process Authors had originally been given a publication date of early March, and were then updated about the transition and advised that articles could now be published any time from 1 January. The IEB had agreed to publish one article per week, to give 'space' around each article for promotion and author showcasing, so there was a need for some diplomacy and some additional correspondence and effort to manage authors' expectations and preferences in terms of the publication date for their articles. 'This was not entirely foreseen and took up a lot of additional time.' (Souster) • staying on track with the overall plan to publish articles on a weekly basis Authors', Co-Editors' and reviewers' varying workloads had a greater than anticipated effect on our ability to move articles smoothly through each stage of publication, and a surprising amount of additional administrative effort was required to ensure we had one article to publish each week. 'In the past, an article that was running late would just have been allocated to the next issue, or let go if timesensitive. This year has required an incredible amount of plate-spinning and constant recalibrating and checking throughout the year. In effect, I now have 45 deadlines a year instead of just three!' (Souster) • managing a more complex copyflow Insights is primarily managed via spreadsheet, and previously this had meant having one worksheet/tab for each issue -'once you got to the end of that tab, that was it done and dusted. Now it's a continuous spreadsheet and you're always adding and taking things off, and moving them from tab to tab. That's more labour intensive.' (Sharp) • wanting to break away from the standard workflow of our publishing platform This had tended to define continuous publication in terms of publishing whichever article had reached the end of the editorial process on whichever day was available in the platform's publication schedule. This meant we had to discontinue our process of inviting authors to suggest a preferred publication date. However, we were able to negotiate a workflow whereby a) one article would be published per week, with a few exceptions for special cases, and b) a list of preferred publication weeks for articles would be provided to our production editor about a month in advance to increase our control of the publication dates. 'It might have made life smoother if we had talked more to our colleagues at Ubiquity Press at the outset, to establish how we could best work with them to achieve what the Co-Editors wanted beforehand, so that our different workflows could have intertwined more effectively from the start, but this was just a teething problem and things now work well.' (Souster)

• less 'regrouping' time between issues
The peaks and troughs of our previous workflow, while creating intense periods (for example, during issue publication weeks), also allowed for 'breathing space' between issues. This particularly affects the Publications Associate, but also the Co-Editors.

Turnaround time
Prior to the transition, turnaround times (the time from submission to publication) varied from 60 to 180 days, depending on the timeliness of authors' submissions. The fact that a diligent author, submitting well before the issue's deadline, would then have a longer wait for publication was one of the reasons for transitioning to continuous publication. The average turnaround time for the November 2017 issue (the last issue before the transition) was 87 days (range: 62 to 128 days). The average turnaround time for articles published between September and November 2018 (i.e. a post-transition window comparable to the prior year's issue production window) was 88 days -however, the range (40 to 161 days) showed an improvement to our minimum turnaround time. The IEB's interpretation of this is that the journal is now capable of substantially faster turnaround times, but that turnaround times are primarily influenced by authors' capacity to process revisions and other steps in the process.

Workflow
There was a small increase in the time and costs required for the Publications Associate to manage the journal; however, these do vary from year to year and had been increased to a similar level in other years prior to the transition, so this change cannot be ascribed specifically to the continuous publication process -not least because the transition also coincided with the introduction of a) additional marketing efforts (e.g. wider communications to UKSG committees; social media promotion) and b) an author feedback survey.

Readership
Average reads per day (across all articles) increased from 2.3 in 2017 (pre-transition) to 7.6 in 2018 (post-transition) -it is assumed that this is a result of the additional promotional processes that continuous publication allows for, and the changing reading habits that this drives: 'people "grazing" over time rather than being presented with a banquet all at once'. (Sharp)

Submissions
The journal has seen an increase in submissions, particularly in terms of 'follow-up' articles (a further analysis of a hot topic), which can now be published more quickly.

Author satisfaction
The journal had not previously had a standard author feedback process in place, so it is not possible to gauge the trend in author satisfaction. Prior to the transition, authors had frequently shared positive feedback informally, and the feedback received via a more formal survey (introduced after the transition) continued to be very positive (for example: 'The editorial process was extremely helpful' and 'All in all, a joy to publish with Insights'.) 'the journal is now capable of substantially faster turnaround times' 'reads per day … increased from 2.3 in 2017 (pre-transition) to 7.6 in 2018 (post-transition)'

Next steps
'Now that we have this focus at the article level, there is so much more we can do to support authors.' (Estelle) The transition has 'taken the anxiety out of the process' (Estelle again), enabling the IEB, staff and Co-Editors to take a more strategic approach to the development of the journal -for example, exploring new partnerships, services and innovations. In February 2019 UKSG announced its partnership with Kudos, a system that helps maximize and centralize authors' own promotion of their articles in a way that UKSG can then build on and learn from. The IEB plans to select key themes (e.g. digital humanities) in order to identify past content that can now be promoted more effectively, potentially in conjunction with new articles. 'We have more time to build on our past efforts.' (Ball)

Conclusion
The IEB took the decision to transition the journal to a continuous publication model in November 2017, and implemented this change in early 2018. The project went better than anticipated. It had seemed a big change with a range of risks, but in practice the change was less complicated than expected and there were no problems. Turnaround times have improved, workflow 'crunches' have been reduced, and individual articles and authors are benefiting from increased marketing. There were some unanticipated challenges, including managing a more complex copyflow and the loss of 'regrouping' time between issues, but overall the IEB, staff, authors and readers are happy with the end result. Authors are particularly pleased with the increased level of promotion for their articles -one described this as 'better than any other publisher' -and the substantial increase in readership (from 2.3 reads per article per day to 7.6).