Tag Archives: PrinceXML

2012–2013 analytics analysis

Arthur was away for a bit, and so I took his place in a Smart Monkey Interview with Russel Southwood:

Part of our focus on university prescriptions meant ensuring that prescribed English and Classics setworks that were already available on Paperight were reformatted. I drew up a list and progress chart for the reformatting of these texts, and Philippa began the process of scrubbing the HTML for those titles. In the meantime, I continued to hack away at the CSS for the automated HTML-to-PDF conversion, to fix compatibility issues that had arisen when we switched over from Prince XML to DocRaptor. And, I continued testing the PDF-to-Paperight-PDF document conversion software on dev. Oscar continued to work on the university prescriptions list, completing the data and making sure that it was updated for 2014.

On our other content-generation fronts: Andi began an internship with us, working on the grade 11 and 12 question bank. Our conversations with Susanne Collins’ agent were going well, and our next step was to secure funding for the project.

I tried to set up a meeting with OUP to discuss reading tree books and expanding their catalogue, met with Hyreath from Juta, and sent her our contract and additional information. We also had further contact with Elsevier, WHO and Jacana about contracts and content, and set up a CoreSource channel to deal with RHS doc transfers.

Near the end of December, I did a comparison of analytics from 2012 to 2013. This included an analysis of traffic to paperight.com, taking into account both the total visits (including recurring users) and unique visits. As the current version of the Paperight site only went live in May 2012 (marked in the data by a *), all data for August 2011–April 2012 is for the earlier version of the site.

1

Fig 1.1 Overview of total visits from 2011–2013.

In comparing data from 2012 and 2013 it is clear that:

  • There is an overall increase in volume of traffic from 2012 to 2013
  • There are two definite dips, and two peaks in site traffic each year
  • Dips occur in June/July, and December each year
  • In 2012, the peaks for visits occurred in February and March, and again in October and November.
  • In 2013, the peaks for visits occurred in May and October, with two smaller peaks also occurring in February and August.
2011 2012 2013
Jan NA 369 1979
Feb NA 1282 2499
Mar NA 1613 2141
Apr NA 1078 2812
May NA 1211* 3999
Jun NA 950 3041
Jul NA 1349 2467
Aug 12 1398 3054
Sep 185 1264 2494
Oct 334 2319 4363
Nov 215 2460 2779
Dec 203 727 672

Fig 1.2 Figures for total visits to www.paperight.com from 2011–2013.

2011 2012 2013
Jan NA 302 1454
Feb NA 1120 2071
Mar NA 1420 1663
Apr NA 871 2076
May NA 871* 3110
Jun NA 567 2289
Jul NA 875 1793
Aug 10 810 2455
Sep 166 964 1993
Oct 239 1765 3622
Nov 157 1857 2207
Dec 125 309 469

Fig 1.3 Overview of unique visits to www.paperight.com from 2011–2013.

In analysing the license sales for the period from May 2012 to December 2013, it was necessary to show the number of copies sold per completed transaction each month. Transactions marked in red are the result of inhouse testing, those in green are actual sales.

2

Fig 2.1 Overview of actual vs. test license sales from 2011–2013.

In comparing data from 2012 and 2013 it is clear that:

  • There are peaks in May, July (Pelikan Park and Silverstream) and September (Kwamatua High and Tambalethu) of 2013, both as a result of sponsorship deals.
  • While in 2012 we had fairly high organic sales in August and October, we did not see these happening in 2013. The majority of these sales came from Silulo branches.
  • Graph also shows that actual sales in June and July 2012 were roughly the same. Spike was as a result of inhouse testing.

A comparison of license sales data to site traffic:

3

Fig 2.2 Overview of site traffic vs. license sales from 2011–2013.

  • The most interesting thing to note here is that high traffic volumes do not result in high sales. In fact, both years, almost every peak in sales is in a month where traffic is lower.
2011 2012 2013
Jan NA NA 16
Feb NA NA 25
Mar NA NA 170
Apr NA NA 44
May NA 22 398
Jun NA 76 72
Jul NA 121 861
Aug NA 228 84
Sep NA 82 1410
Oct NA 186 44
Nov NA 173 25
Dec NA 20 10

Fig 2.3 Overview of license sales on www.paperight.com from 2012–2013.

For interest sake: here is a breakdown of sales per province.

4

Fig 2.4 Overview of license sales per province from 2012–2013.

Publisher registrations

  • Niz Publications (2/12/2013)
  • Vij Books India Pvt Ltd (14/12/2012)

Updates to the privacy policy, and move over to DocRaptor

By August we had a large backlog of documents to be uploaded. We began the process of prepping and uploading these. We also started scrubbing HTML for setworks (including Shakespeare plays), in order to reformat these as A5 booklets, with no advertising, or line-marks. The clean HTML would be put through the converter to create cleanly re-formatted documents.

I briefed Nick on posters advertising exam packs and study guides, and Dez and I delivered these to schools in the southern suburbs (this involved my contacting schools’ academic heads and arranging a time for drop off etc). We also updated our privacy policy , and informed publishers of all agreement changes.

Near the end of the month Arthur asked Shaine to look into using DocRaptor instead of Prince XML for document conversions. This would save us a massive ongoing expense, but required an update to all CSS due to compatibility issues.

Publisher registrations

  • Palanga Publishing (7/8/2013)
  • Jermaine M Charles (12/8/2013)
  • Xhosa Fundis (19/8/2013)
  • Berg+Bach (30/8/2013)

Automated conversion implemented on Dev

A couple of important milestones were reached in June. First of all, PDF-to-Paperight-PDF and HTML-to-Paperight-PDF conversions were implemented on dev, and we set about testing and bug reporting. Specifically, we had to tweak issues regarding page positioning, page numbering. and margins. In the case of HTML-to-Paperight-PDF conversion, this required CSS hacking. We created CSS that would work using Prince XML to create a well-designed, workable PDF output.

Other new functionality had to be tested as well. Our bug testing of the reminder emails for publisher payments led to emails being sent out in error. We had to send all relevant parties an apology and explanations. We linked to this post in the emails, explaining what had gone wrong.

I took over from Philippa as the ‘handler’ for Rondebosch Colour and Copy when she left for her Washington Internship. When I’d been away, each team member had taken on an adopted copyshop, to see whether this added level of support affected sales and competency using the service.

One of my main accomplishments this month was in compiling and submitting a full security audit for Pearson. Their evaluations lead to us scoring 8/10. They were very impressed with our watermarking functionality (we only fell short due to not having been in operation for long enough to do an annual review).

Held meetings with Jeremy Boraine (Jonathan Ball), and Francois van Schalkwyk (African Minds), and approaching of new publishers: The Answer Series, Nali’Bali Sunday Times section, Unisa Press. Pan Macmillan document prepping, and organising of PR material with Nick.  We submitted our application to receive funding for Frankfurt Book Fair  via PASA. We also did the final proofreading of PYWA, and subsequently released it on Paperight.

Publisher registrations

  • Lule Publishers (5/6/2013)
  • Lost Plot Press (9/6/2013)
  • Tawqeer’s Tutoring Service (11/6/2013)
  • African Minds (14/6/2013)
  • Jurassic London (20/6/2013)
  • Ediciones Microtemas (29/6/2013)