Telling a success story

Writing about web page http://go.warwick.ac.uk/repositories

WRAP reached its 2000th live item last week. We made a little news article about it for the library's home page and news feeds, including on Twitter and Facebook. We asked the University's central communications team to publicise it for us, as we did when we reached 1000 items. But whereas that was a big splash at a time of not many other news stories so we got a front page link on the University home page, this time we only got a mention on a news and events page that is buried inside the staff area of the Intranet.

We reached 1000 items in mid July when there weren't many staff around to read about it, which we thought was a shame. But now, even though there are plenty of people around, there is so much else competing for their attention that our 2000th item has passed with hardly a mention. Perhaps reaching 1000 over the summer was a blessing after all!

The 2000th item news went out on Thursday last week, and so far as I can tell the impact so far has been...

1) Two deposits from authors we've not previously been in touch with and a very slightly higher number of visitors to our deposit form - 24 in one day when the highest peak in previous weeks was 21. NB we still don't get anywhere near as many deposits as we do visitors to that form!

2) A handful more visitors to our main page about repositories at Warwick (linked). By which I mean a peak of 53 visits on Thursday when previous high peaks were 40/41.

3) Our approach towards 600 visits a day on WRAP itself was maintained last week. On Monday we reached 595 visits which was great. By Wednesday we'd slumped to 567 (still pretty high!) and on Thursday we were back at 588 visits.

4) The 2000th item which was linked to directly from our news stories attracted 15 visits, to date.

I shall continue to monitor visitor numbers to WRAP and to the 2000th item. Because I was busy last week, I didn't have time to write to the authors to explain to them, which I still intend to do. I also didn't find time to blog about our achievement here, which I've now done...

When we reached 1000 items we were also mentioned in a press release from the JISC, so that would no doubt have had more of an impact. Perhaps we will get more publicity from the JISC and from our University communications team when we reach 10,000!

There are many success stories to tell about a repository. The total number of items in it is a simple one, and reaching a first significant milestone is important.

The speed and consistency of growth is another type of success story. We took 11 months to reach our first 1000 items, and it took us just under 4 months to reach 2000. The story that tells is that it took us a relatively long time to sort out all our processes until they were optimised. I believe that it will take us longer than another 4 months to reach 3000 items (for various reasons but partly because of the need to process theses which take longer), but not significantly so. We can keep growing WRAP's collection in much the same way as we have been doing.

The quality of the items we have collected is a more difficult story to tell but I think that the most interesting number in this blog posting is that we are getting almost 600 visits daily. With only just 2000 items for people to be looking at. My feeling is that this is a significant part of WRAP's success story. People are reading the content and that is what gives us the motivation to go out there and get more!

Jenny Delasalle : 09 Nov 2009 17:04 |  Tags: Advocacy Google Analytics Statistics |  Comments (0) |  Close comments |  Report a problem

Reporting on statistics

Writing about web page http://writetoreply.org/actually/2009/10/28/thinking-about-user-tracking-on-writetoreply/

I just asked on the UKCoRR list about Google Analytics, after forwarding a link to Tony Hirst's blog, as recommended by Andy McGregor of the JISC.

The replies got me thinking about how we use the statistics that we get from GA. Some repo managers are writing regular monthly reports for managers, as blogged by the CADAIR team: http://welshrepositorynetwork.blogspot.com/2009/10/statistics.html

I look at the stats at least once a month, in order to write to our "top content" authors. I use that e-mail as a way of promoting WRAP to the authors, especially those who might not be aware of WRAP or that their article is in the repository. (Deposited by co-authors or administrators on their behalf.) It has resulted in raised awareness, some goodwill and conversations about WRAP but has never led directly to further deposits - yet. I have copied the heads of department in to some of these e-mails, when I know the author is already comfortable and happy with WRAP, although I've no idea whether they pay any attention to the e-mails!

What do I say to our top content authors? Here's a template, which I don't often have to vary much....

I'm writing to inform you that your paper in WRAP: (REFERENCE)

Is the most popular paper in WRAP in the last month. I'm keeping our highly read authors informed of what I can about the visitors to their content. I should point out that it is actually the record that is being visited/read, rather than the full text itself. There have been NUMBER pageviews of the record describing your paper from DATE to DATE. All visitors came from a search engine, the vast majority from Google. Most looked at the record and went away again, but some explored the subject area in WRAP.

There is a great variety of keywords that have led visitors to your paper, including the following: (LIST KEYWORDS)

Visitors came to your record from NUMBER different networks, so it is not all Warwick people looking at your work. Noticeably academic/educational networks that your visitors came from include:

The vast majority (NUMBER) of visits were from within the UK, but your paper's record had visits from....PLACES.

There have been no great peaks and troughs of activity: visits come every day and remain at or under NUMBER per day.

I did a quick Google search for PAPER'S TITLE and your paper's WRAP record is Xth in the results list.

Whilst looking at the stats I might spot something interesting, which I would usually blog about here and write to people in the library who I think ought to know: managers and subject librarians, or even our internal e-mail newsletter to all staff.

I know that our library management group are interested in big numbers, like how many pageviews there have been since we went live, from how many hundreds of countries/territories, etc. They want to illustrate the success story that we're gaining in visitors every week as we grow in content ever more rapidly! In compiling such a news piece, I might look at our growth chart on ROAR as well, or at the number of items we hold for a particular department, to provide further background information about the interesting pattern.

I also send out a "newsletter" once a term, by e-mail to people who are interested in hearing more about WRAP. I know that they're interested because I introduced an "I would like to hear more" tick box onto the deposit form and they ticked it!

Otherwise, statistics might make their way into my presentations to departments or articles that I write to raise awareness of WRAP, or onto our web pages about the repository. They are something to say when we talk about WRAP and it's important to be able to give the detail and context that they provide, to keep people interested in our work.

Jenny Delasalle : 30 Oct 2009 15:52 |  Tags: Advocacy Google Analytics Statistics |  Comments (2) |  Close comments |  Report a problem

Theses and early draft deposit in repositories: is that publication?

Does repository deposit of a work constitute publication and as such jeopardise the chances of publication by a more prestigious/established/profitable method and another agent?

It's not really a question that I can answer yet. I am certain that repository deposit ought not to cause any problems with regard to publication elsewhere, and I have not come across evidence to prove that it would cause a problem except in one particular instance that I investigated and which I describe below. But I'd like to gather more evidence on the topic because I can't prove that it isn't a problem either!

I do not usually like authors to deposit unpublished papers to WRAP. Part of my reason for that is that we want the highest quality content we can get: if the article has been accepted for publication, then that is some measure of quality. PhD theses are obviously of high quality and a separate case in their own right from this point of view: these are added to WRAP.

Quality issues aside, if an author were to write a paper with the intention of submitting it to a journal but wanted to make it available on OA as soon as possible through repository deposit (never happened yet although we've had some that have been accepted and are forthcoming), I would advise that author to look at the journal publisher's copyright agreement that s/he would be asked to sign. I know of at least one publisher who would consider repository deposit of the paper to constitute a prior publication, thus preventing the author from being able to sign the copyright form stating that it had not previously been published elsewhere: this was the British Psychological Society, who I investigated over a year ago.

Inability to sign the standard copyright form might mean that the work could never be published in that journal or by that publisher, but alternatively the form might be amended. I expect that the publisher's position would depend upon the precise circumstances. 

It occurs to me that the matter of an early version of a paper is probably different than that of a thesis from which a book or article might be published: after all, the content would have to be substantially re-written from a thesis, whilst different versions of papers might be very similar, so a publisher might be more concerned about repository deposit of papers but not as worried about thesis deposit.

Of course, our students can opt out of their thesis being made available in WRAP, even though they do have to submit it. So, if a student was hoping to be published and was unsure of the publisher's policy then s/he could always embargo the repository version from being made available anyway.

This is a big issue, and one that needs more thought and investigation, I believe. Because I would like to be able to advise students to allow repository availability of their theses, knowing more about how publishers would react.

Jenny Delasalle : 05 Oct 2009 13:44 |  Tags: Copyright Publishers Theses |  Comments (6) |  Close comments |  Report a problem

Visitor pattern over the summer

I'm just back from a lovely break in Australia, and I'm pleased to see that our visitor numbers are up again, as the Autumn term is about to start here and is no doubt underway elsewhere. Looking at the graph of visits to WRAP over the last year, there is a clear sag over the summer down to about 200 visitors on a week day, but we're back up to around 400 again now and I hope that we continue to grow in visitor numbers as the amount of content also grows.

We're way past the 1000 mark that we reached at the beginning of the summer break: over 1600 items now, and growing fast!

Jenny Delasalle : 29 Sep 2009 11:49 |  Tags: Google Analytics Statistics |  Comments (0) |  Close comments |  Report a problem

Theses from the past

Writing about web page http://ethos.bl.uk/

We have added over 200 theses to WRAP, including those digitised through the BL EThOS project, as well as those submitted electronically in the last year.

We chose not to contact former students in advance of allowing the BL to digitise theses, and since adding content to WRAP we have had three requests to take the content down, all of which we have dealt with within 24 hours of receiving the request. Two out of our three requesters were unaware of the EThOS copies and I believe that this is because WRAP content is indexed by Google and hence is easily findable, whilst EThOS content is not picked up by Google (at present).

Are we right not to contact the former students in advance? Well I think it is worth following the approach that we have done because we are trying to assist in scholarly research and to raise the authors' academic profiles through the availability (and longer term preservation) of their work, and we are not attempting to make a profit from their work.

I do have experience of trying to contact former students from a secondment I did to our Careers Service, where I ran one year's Destination of Leavers from Higher Education survey. It is an extremely time consuming and resource intensive process, involving letters, e-mails and telephone calls (some in foreign languages as we got through to family) to chase the former students, at both UG and PG levels, in order to achieve something close to an 80% response rate. That is just to get a response at all from either the student or someone related to the student, who knew their career progress, which of course is a different matter than attempting to get permission to digitise a thesis.

In comparison, our three requests represent only just over 1% of requested theses not available to the scholarly community, if I don't count those which we already refuse because there are embargoes against photocopying in place.

Note that these theses are not all recently written: one dates back to 1983. The chances of us being able to track down that author through either the details they left behind on our records or even through the former supervisor are very, very slim indeed.

I am very pleased that so much quality work is now available for the scholarly community, to assist in academic research. However, I will continue to monitor whether we have taken the right approach or not, in not contacting authors directly.

Jenny Delasalle : 02 Sep 2009 10:22 |  Comments (1) |  Close comments |  Report a problem

Busy with theses

Recent additions to WRAP have been nearly all theses, and today we have added our first multi-media item (http://wrap.warwick.ac.uk/1156/). This is partly because we get batches of theses through from the Graduate School, so our metadata librarian deals with them all in one go, and partly also because the E-repositories assistant who processes submissions through our deposit form has been on holiday for two weeks, and then again partly because, when I tried to process some items in her absence, we have been locked out of Sherpa Romeo and all things on the Nottingham Uni web domain and that has only been fixed for me today!

So, theses it has been: both the ones submitted to us in the last academic year, and also the ones digitised through the Ethos project. We could have waited for Ethos to send us them in a download, but we decided to download them individually for ourselves.

Hopefully we will be back to more of a balance with journal articles very soon, however, as our E-repositories assistant returns tomorrow...

Meanwhile, I am busy preparing for my first meeting of the UK Council of Research Repositories as their chair, which is on Friday.

Jenny Delasalle : 11 Aug 2009 17:10 |  Tags: Theses |  Comments (0) |  Close comments |  Report a problem

We made it to the front page :–)

Follow-up to A huge milestone: 1000 items! from WRAP repository blog

Yesterday we were on the front of the University's website. We still are, if you hover over the News and Events heading...

Jenny Delasalle : 17 Jul 2009 17:00 |  Comments (0) |  Close comments |  Report a problem

Open Access issues

Writing about web page http://www.friendsoffrontiers.org

Today, I got an e-mail from someone associated with this organisation. I looked into it and it's a mildly interesting story!

The organisation appears to be affiliated with an open access journal in the field of Neuroscience. This journal does charge a subscription fee and although they make content available on open access, they have to make something available for that fee (see their page at: http://www.frontiersin.org/subscribe/). So although the publishers are supporters of Open Access, they are also locking stuff down!

Meanwhile, this "Friends" organisation is another way in which the costs of open access publishing could be met. I like the website because it seems to me to exemplify the issues and concerns around open access publishing in general, within the context of a discipline. Perhaps the Neuroscience community will be very aware of the need for open access and the context of publishers' concerns...

Jenny Delasalle : 16 Jul 2009 17:18 |  Tags: Open Access Publishers |  Comments (0) |  Close comments |  Report a problem

A huge milestone: 1000 items!

Today is a day of celebration for WRAP, and I daresay I'll still be singing and dancing tomorrow. We've just reached 1000 live items, which is a huge achievement and now I need to capitalise on it as far as possible with my advocacy work. Here's a little list of approaches:

1) Tell Library Management!
2) Get a news story onto the Library News page, with its RSS feed, and ask for a link from Library home page to that story.
3) Get a Library Twitter account tweet about it.
4) Update the Library's Facebook profile with news.
5) Tell all of Library staff by e-mail, and ask them to also sing it on the mountains!
6) Blog here.
7) Start editing WRAP web pages to tell the glad news.
8) Tell University Communications office (who I already warned) that we have now reached our milestone and ask them to help spread the word.

I'm up to point 6!

Jenny Delasalle : 14 Jul 2009 17:21 |  Comments (0) |  Close comments |  Trackbacks (1) |  Report a problem

All new visitors

I had another little play with Google Analytics this morning, just because I happened to be signed in to Google (looking at the Australian repository managers Google group). Today I explored the "Advanced Segments", available in Beta.

I compared "New visitors" to "All visits" in the last month, and the trend is remarkably the same pattern, just slightly lower. Very few people come back to WRAP after having visited once: it is not a site that people wish to visit and explore. They dip in for the content that they want, and they don't necessarily come back. 

Our "Bounce rate" is also just over 71% and this also confirms the pattern of visitors going in and then back out again. Their landing page is therefore the one they want, and they follow our link to the published item (which is what we want). Or else they were looking for something else and they go back to their search results. They might have read our abstract and learnt enough about the paper to know that they want something else. Or they might not have wanted an academic paper at all. We can't actually tell where visitors go when they leave WRAP.

Looking at which pages are visited most often in WRAP tells me a little bit more. Just under 5% of all visits were to the WRAP home page. The "non-bounce" rate for the home page closely follows the total number of visits to that page, so the home page is not putting visitors off! Our second most popular page is our advanced search form. Which is interesting because our home page links to the simple search form, so I'm not sure how the advanced form became more popular than the simple one: must be the Google effect! The bounce rate for this page is just over 57%, which is quite high. I don't particularly like the advanced search form in terms of user-friendliness, so this doesn't surprise me.

Our highest bounce rate amongst the top content pages is our Information page (80%), which is also to be expected because the main way we use that page is to link to content on the Library's website, which is far more easy for us to edit and keep up to date. The next highest is for one of the papers, which is 72%, and so far as I can tell at a glance, this is typical of the bounce rates of our papers. Bounce rate brings in an interesting different dynamic for when I write to our top content authors... one with a high number of visitors who also remain in WRAP (about 54%) is also one which for we do not make full text available. I looked at the Navigation summary for this page, to try to get a feel for where the visitors to that paper were coming from and going to. 25% had come from a previous page in WRAP, and 29% went on to another page in WRAP. Mostly, people seem to have been clicking to see the pdf. But the locks are all working well, so people must be just testing them. The paper will be released from embargo at the end of this month. Other visitors did go on to view another paper by the same author, and some of them were clicking on our subject heading. I rather suspect that this anomaly is the result of my using this author's RSS feed on my example page, to show people how easy it is to incorporate WRAP results into a Warwick web page! I chose this author's content because it was already popular, but I will keep an eye on top contents' bounce rates in future.

The general trend of visitors dipping in and out of WRAP rather confirms my impression that most people will find repository content when searching elsewhere, rather than searching in the repository itself. This may change as our body of content grows. But for now, even though we're pleased to be approaching 1000 items, our content is pretty disparate across the subjects and as such, not the most useful of collections to researchers.

Referral traffic on WRAP is slightly higher than direct traffic, and both are bouncing around that zero line but again, it is search traffic that follows the overall pattern of visitors. Perhaps academics will get used to sharing WRAP URLs in the future, and we will get more referrals.

The pattern of visitors to WRAP pretty much confirms our dependency on Google and its ilk to bring us visitors. Which is OK, but a slide that I saw yesterday from Tom Abbott, who co-ordinates all Warwick's iTunesU output demonstrated the difference that being a destination site (or promoted on one) can make. The usage graph he showed, for a file on Warwick's own website, which was then published on iTunesU showed a vast increase. Visitor numbers for the podcast on Shakespeare's portrait seem to have been astronomical as a result of Apple promoting it amongst their collection (millions of visits rather than hundreds, as with WRAP!). iTunesU is a destination site. Repositories like WRAP are not, but perhaps repository cross-searching sites could be. That's a part of the reason for making our metadata such high quality and harvestable.

Looking at the highest referring sites to WRAP gives some clues as to who might become more important. The Index to Theses (http://www.theses.com/) is likely to be a destination site for some. They link to us from: http://www.theses.com/idx/registered_users/etd/96.aspbut as yet don't link to our metadata records from theirs. Fair enough, 'cos we've been a bit slow off the mark with the theses. But we're adding them now, and as we come to add more theses, this might be a source of more visitors: they do have an example of a Cranfield thesis with a link to the full text in the repository from their record. Although I didn't think that the link was all that obvious: it's in the left hand margin, rather than in the text of the record describing the item. So many academic resources seem so un-user-friendly!

Other referrers to WRAP include bing.com, which is actually a UK specialist search engine (and has recently become the second highest search engine source of visitors to WRAP), google itself (in various guises), Warwick's own research information system (whose records link to ours) and the NHS Evidence Health Information Resources website. I'm not sure why Bing and the Google guises reported in this section are not included in the search engine hits... but at any rate, I consider them search engines!

OpenDoar and ROAR brought us three visits each... and I think that at least one of those visits would have been me, checking our record with them! Repositories and repository cross-searching tools are not destination sites in their own right yet (or at least WRAP isn't one: perhaps larger repositories see a different pattern). Repository cross-searching sites are not bringing us visitors. Search engines are...

I don't know of anyone who does research by visiting a repository cross-searching tool. I don't believe that they're even promoted by academic librarians in their liaison with departments. OpenDOAR and ROAR seem to me to be more tools for those in the repository and information management sector. My personal favourite cross-searching tool is OAIster, but I haven't investigated them all thoroughly, and I think it's a side of repositories that deserves further investigation. Now that we've got content in them, with high quality metadata, how do we ensure that people will find it in ways convenient to them? Google is great but it only goes so far...

Once we have tools that we might recommend people to use, to find repository content, that's a whole other advocacy/information literacy journey we would need to go on, persuading researchers of how to find and use repository records. Our Psychology librarian has made a start: her tutorial for undergraduates, which she created in collaboration with tutors in that department, recommends that students might like to look at their tutors' papers in WRAP, so that they can learn about their areas of expertise and interest and get ideas for their projects. But there is a long way to go!

Jenny Delasalle : 09 Jul 2009 15:03 |  Tags: Google Analytics Statistics |  Comments (0) |  Close comments |  Report a problem