This post summarises the last three plenary sessions from day 2 of IWMW 2013. It follows on from the summary of day 2′s first three plenaries.
Day 2 Plenary Sessions (part 2)
P8: Paul Walk: Working with developers
Paul Walk talked about how learning to work well with developers can help institutional web managers get the most out of their software/systems. He said that managers need to get good at estimating the time projects will take, and learn some of the developers’ jargon (though they shouldn’t pretend to know more than they do), specifically understanding what key concepts like ‘the cloud’ and ‘software as a service’ really mean.
Over the years we’ve had lots of different words for developers: developer, programmer, engineer, analyst, data-wrangler… even “software authors”. Different words have different connotations; the words we choose can tell us a lot about how we think about the work of development. We also have different words for the process: we talk about “writing” software, software “engineering”, software as a “craft”, as an “art”, and so on. “The last time people constructed a vocabulary for software development was in the 1960s, when they coined the phrase software engineering, both as a wish and a direction for the future.” The different words reflect different development paradigms.
Walk gave an overview of the ‘waterfall’ methodology and then moved on to ‘agile’ development and how it differs. The ‘agile’ attitude can be summed up as “rough consensus, running code” and its articles of faith are that innovation happens in a local context and local developers empower the organisation to innovate. ‘Local’ in this sense means ‘close to the users’ — we have to recognise that the users are part of the process, and that building software is about building on existing relationships. There are lots of advantages to having local developers, e.g. better understanding of local context, and better availability/responsiveness during projects.
It’s not just about developers doing development work, though; Walk argued that Universities need more technical people in their higher echelons: “bring engineers into your company at all levels, including the top” (as Eric Schmidt puts it). He also identified a gap in the career path for developers, the difficulty of letting developers keep developing but also contribute to management, strategy, and governance, and claimed that universities can learn from industry in this area.
P9: Ranjit Sidhu: 9am, 16th August, 2012: What the fcuk just happened then?
‘What happened then’ was that the A-level results came out, and universities got a shock as they realised that the game had changed. 17 of the Russell Group universities went into clearing in 2012: applications were down by 7% across the board. But we don’t need to wait for A-level results to find out how bad things are: we can get the data and work out who we need to target and how.
Sidhu talked about how ‘data science’ was becoming more prevalent (in fact, apparently, “the sexiest job in the 21st century“) and how we need to work out what we should be analysing: the web is uniquely placed to analyse statistics and help universities to target the area it needs to. And this targeting is big business: Sidhu estimated an £8,500,000 spend by universities in the week after clearing; a keyword ad in the top spot on Google during clearing costs £22.12 for just one click. Online marketing is perfect for targeting our spending more effectively, and we need to get more students in — Sidhu claims that tuition fees are the only way that HEIs can make money now.
However Sidhu warned of the dangers of thinking ‘big data’ will solve everything (“If you didn’t do anything meaningful with the small data you had, why will you do anything with the big data?“) and reminded us that databases are easy, but you can’t outsource thinking. He was particularly sceptical about the value of UniStats (apparently fewer than 1 visitor per day actually clicks the UniStats widget).
After lots of hard figures Sidhu finished with some more philosophical questions: whether free content is destroying diversity; whether the free pricing model unfairly benefits the rich (should people pay more if they can afford it?); and whether early adopters of tracking/analytics technology (what we used to call ‘spyware’) have a responsibility to make sure it’s used ethically.
P10: Paul Boag: Institutional culture is crippling your web strategy!
Paul Boag didn’t pull any punches in his talk; highlights included telling us that we were a bunch of whingers and that when it came to our organisational structures we were all, not to put too fine a point on it, completely screwed. It’s just his way of showing how much he cares…
He identified the problem as our “pre-web organisations” being mired in bureaucracy, with slow, committee-based decision-making, lots of internal politics, a “project mentality” and lots of departmental silos, and a misunderstanding of what the web is (a marketing tool, an IT service). The “project mentality” is unhelpful because the web requires cross-departmental working: organisational structures are a barrier to this. Departmental silos occur for the same reason: websites and services are organised around organisational structure, departments, faculties etc. This slow-moving bureaucracy results in “big bang redesigns” (rather than incremental change) leading to yo-yoing expenditure and long periods of ineffectiveness.
Boag argued that fundamental changes are needed – we need to:
- Become user-focused
- Have a strong centralised web team
- Create policies and procedures
- Create a consistent user experience (including establishing a tone of voice)
- Work incrementally (be agile)
- Create a way to talk to each other
Policies can help us say “no”, e.g. if we can’t remove redundant content simply because it’s irrelevant, we can establish policies to help facilitate this, such as saying that if a page isn’t updated within a certain amount of time, or falls below a threshold of visits based on analytics, then it gets marked as archived, removed from search, and so on.
Boag suggested some other techniques to help us improve our sites in the face of organisational obstructions: “user cards” (and insisting on them before introducing new content/functionality) can help us build a user-focused experience; testing and monitoring everything can help us justify our decisions. We also need to communicate: we need to share best practice, blog our experiences, find ways to talk to each other. We also need to talk to senior management, tell them what we think, market ourselves within the organisation.
His diatribe against institutional culture ended with telling us that if we want to enjoy our jobs, we need to fix our organisations or look for another job. I’m not sure the audience were entirely convinced by his insistence that losing/leaving your job is nothing to be afraid of, but I think most of us recognised the relevance of at least some of the ranting!