Was Glow derailed by a pointless procurement exercise? Jaye Richards-Hill reports
How did the world's first national schools intranet veer towards becoming the world's first Zombie intranet? Because that's what Glow, universally lauded, could become following the Scottish Government’s decision to extend the original version by 15 months at a cost of £5.5 million.
Scottish schools and education authorities, dismayed by more civil service bungling over Glow, may now opt for Fast-Track Apps from Google which pulled out of the procurement. What went wrong? And why did Glew, its popular potential replacement by Education Scotland end up in the Government trash box?
The announcement was made public on the Scottish Government’s website. RM, the company which produced the current version of Glow will now preside over a legacy system which will inevitably start to decay through lack of credibility and modernisation, and probably wither on the vine as user numbers decline.
The story goes back to early 2010 and discreet rumblings over the failure of Glow to live up to the expectations of the education community in Scotland. For the uninitiated, Glow is an intranet system comprising a suite of collaborative tools, email, and video-conferencing components.
Glow is based on Microsoft’s Sharepoint, for which substantial fees are paid
The core technology at the heart of Glow is Sharepoint, from Microsoft. RM, in partnership with Education Scotland, manages and operates the project for Scottish Government, and the development of the project has involved paying Microsoft large licensing and software fees. Microsoft blogs about Glow and appears to be very proud of its involvement. However, open source and free alternatives are readily available (as demonstrated by Glew).
Designed in the early part of the decade, it was bound to suffer from the speed of technological advance outpacing its development and uptake, and the hype generated by Learning and Teaching Scotland (now subsumed into Education Scotland). This masked deficiencies in project management and reliability which had resulted, by late 2009, in patchy adoption by schools.
Swathes of the country were unconvinced of the merits of fitting a national network system to such a diverse country. Having published the only piece of classroom-based research into its effectiveness, and writing in TESS (TES Scotland), I called for a moratorium on any further development until these issues had been addressed.
By mid-2011, the Scottish cabinet secretary for education and lifelong learning, Michael Russell, himself no stranger to technology and social networking, appeared to agree when he pulled the plug on the Glow Futures project, an exercise set up to procure the next version of the intranet. Having snuffed out the prospect of a second but similar iteration in an announcement on YouTube (above), Michael Russell convened a conference of the great, the good, and the interested from all sectors and areas of Scotland to decide if there was actually a need for a national intranet (let’s call it Glow2).
If there was, what form should Glow2 take and what tools should it include? This conference was preceded by a Wiki which hosted some lively, wide-ranging and well informed debate which influenced the face-to-face discussions that took place some weeks later.
So far then, so good. A government minister who, rather unusually for such incumbents, listened, consulted, and, more important, acted to engage the education community rather than alienate it. Enthusiastic and well informed classroom practitioners were eager to help develop government policy. Even the Glow team at Education Scotland appeared to take on a new lease of life, invigorated by this new spirit of togetherness and willingness to forget the mistakes and arguments of the past, finally led by someone who had an instinctive feel for what was needed to truly enhance learning and teaching.
Glow2 to be based on Google Apps or Office 365 – Google won popular vote
It was then, in this heady autumn of education 'love-ins', that the seeds of a future disaster were sown. Andrew Brown, the head of Glow at Education Scotland put together a small, top-quality team to turn the hopes and dreams from the conference and the wiki into reality. Team members included Ollie Bray (national adviser for emerging technology at Education Scotland), Stuart Campbell (Glow technical manager at Education Scotland who has an LA/corporate IT background) and Charlie Love (computing teacher in Aberdeen City and former LTS Consolarium development officer).
They looked at a range of possibilities including Google Apps and Microsoft's Office 365. As well as employing its own resident technical expertise, the team also consulted widely with other technical experts, software architects and industry consultants to make sure that all of their ideas were practical and feasible.
It was thought that any future Glow system would be largely based on the free-to-schools technology of Google or Microsoft and would require an authentication system 'owned' by the Scottish Government.
The potential for bringing across Glow user data safely to Google Apps or Office 365 was realised and the team set to work scoping out just how this might be possible using open source software.
Work progressed through the winter of 2011 and included a secure authentication system similar to the “Shibboleth” version used by Glow. This provides what is known as “single sign-on” so students and teachers don’t have to separately sign in to a range of different applications that they use.They also created a home portal based on open source Drupal software which could handle the apps of the individual user's choice. (Full technical details and specifications can be found on Charlie Love's blog.)
Now the result of this work, Glew, is a fantastic development and it’s open source, ie free to use. It’s simple, versatile, user-friendly and intuitive. And definitely not “clunky”, the description most often applied to the Glow platform it was designed to replace. In fact it was just what the doctor (and Mike Russell together with most of the education community in Scotland) ordered!
Encouraged by this progress, meetings were held in January 2012 between the Scottish Government, Google, and Microsoft. It was apparent at this stage that everything was ready to go with Google. Microsoft's Office 365 for Education was still testing and at the time still had costs associated with teachers using the system (Microsoft did not change its global pricing model until March 2012). And 365 was not fully compatible with all mobile devices (a poor fit with Michael Russell’s Bring Your Own Device – BYOD – dream for schools) and certainly not yet ready for a full national launch.
Google Apps already being used successfully in East Lothian schools
Google Apps, already used successfully in East Lothian schools (see "'Google it' takes on fresh meaning for Scottish schools"), had appeared, almost by default, to become the front-runner to form the basis of Glow2. This was where things started to go wrong. Innovation and government procurement are uneasy bedfellows.
Google Apps would certainly have picked up the majority vote from the ordinary teachers and children from all over Scotland who were given an opportunity to provide feedback on both Google Apps for Education and Microsoft 365 for Education. Indeed, the verdict from this user testing exercise was pretty conclusive. More than 90 per cent of the teachers and more than 70 per cent of the children preferred Google Apps to the Microsoft product. Or so we're led to believe by trusted insiders, because the actual stats are now buried, and only the Glow programme board, comprised of civil servants (with not a Glow end user in sight, by the way), know where they are!
A Freedom of Information request to the Scottish Government (see link below) attempting to elicit this information has been refused. This is strange given that a 'crowd-sourced' Glow is exactly what the education community said it wanted.
Andrew Brown's group made a recommendation to this very same Glow programme board, based on user testing, which was considered at its meeting on February 15 this year. This has been the subject of another (ignored) FOI request,
Many observers felt that attempts were then made behind the scenes to stack the procurement in favour of Microsoft rather than Google. And this was the reason for the secrecy. You wouldn’t have to be a conspiracy theorist to think this.
The switch in direction following the February 15 meeting is laid by some at the door of ISIS, the Scottish Government’s own IT unit. ISIS, led by chief technical officer Andy McLintoch, is considered to be extremely close to Microsoft, which is not surprising as it handles big schemes and big budgets and a Microsoft representative had desk space in its office for two days a week. Consequently, it is felt that the expertise within ISIS is focused on 'traditional' ICT, like Microsoft networking, rather than popular and easy-to-use technology like Google Apps.
At this point, development work favoured a Google-led solution. Google could deliver in terms of availability and scope. Google was also the preferred choice of the vast majority of potential users. Microsoft stood to lose out on the Scottish education Intranet, and the prestige and kudos that goes with it, particularly when it came to its profile with potential paying, commercial customers. Many industry observers note that the two organisations are engaged in a vicious battle for the 'cloud'; for Google to be able to hold up Scotland as an example of how effectively its apps suite can be used would have been a big feather in its virtual cap.
Was the commercial 'battle for the cloud' being pursued behind the scenes?
Could the intense cloud rivalry between Google and Microsoft lie behind what happened next? A decision was expected at the Glow board meeting in February because only seven months remained before Glow1 would be turned off and Glow2 turned on.
The plot thickened when, at that Glow board meeting, no decision was taken. Discussion rages as to whether any procurement exercise was even necessary, bearing in mind the requirement had always been for Glow2 to be a free product. The advice given on this by the government's lawyers isn't absolutely clear (and the subject of yet another unfulfilled FOI request) but insiders agree – no procurement exercise was necessary for a free product.
Then ISIS stepped in and offered to 'review' all the work on Glow2 so far. Why? Cynics might be forgiven for thinking that ISIS was worried about a non-Microsoft product taking centre stage. Was Microsoft using its Scottish Government connection to apply pressure? Were there discussions at ISIS when Microsoft picked up the vibe that Google was the front runner?
No one knows for certain, but speculation is rife. More than a month was wasted as the meetings and discussions with Google and Microsoft were all repeated. Did this delay prod the Scottish Government to open discussions with RM about extending the current Glow contract as the fallback option which eventually came to pass?
Three significant things happened next, in quick succession. First, the work done on the potential Glow successor, Glew, by Charlie Love fell under the ISIS spotlight. In a hastily convened meeting, all of this development work was subjected to an ISIS risk assessment review to assess its suitability, compatibility, and deliverability. In a highly technical "delivery confidence assessment review" Glew was given the thumbs down, despite the fact that its own, respected technical expert had vouchsafed all the work so far.
ISIS claimed that successful delivery of the work stream appeared to be unachievable; that there were major issues on definition, schedule, budget required, quality or benefits delivery. It did not consider the scheme manageable or resolvable. But it did leave the door ever so slightly ajar, stating that overall viability could be re-assessed, although this looked unlikely given its pretty conclusive sentence of execution on the project. Cynics suggested that this review was put in place in order to shove Glew out of the national picture – "set up to fail" as the expression goes.
Back to an unnecessary invitation to tender?
Second, from out of nowhere, and despite loud rumours of advice to the contrary, an Invitation To Tender (ITT) document was released for Glow2, together with a procurement timetable. The specification in this document was changed no less than three times during its short life span. It also included a reference to "offline functionality," something it was thought that Google Apps might not have satisfied, and a requirement to reveal the location of the data-holding centre, again something Google might not have been able to do (although it does undertake to hold data in compliance with data protection legislation). This was taken out of the ITT by its final version, possibly due to worries about the lack of interest from other providers, but too late for Google, The company withdrew from the process, and issued a stinging public statement roundly criticising the Scottish Government over its handling of the whole affair.
Third, representatives from all the Scottish education authorities gathered in Glasgow at a prearranged meeting, eagerly anticipating the unveiling of the new Glow2. Now these folk, having had their own teachers and students participating in the user trials back in January and February knew that the choice was going to be between Google Apps and Microsoft Office 365. Furthermore, they all knew that a recommendation based on these trials was supposed to have been made to the Glow programme board meeting of February 15. They also knew the popular choice – the resoundingly one-sided verdict of the participants, fully expected to be the preferred solution, was Google Apps.
Instead, they were forced to listen to several technical presentations on the requirements and "architecture" of the new solution. Even so, when Andrew Brown, the head of Glow at Education Scotland, stood up to speak, they still expected the anointing of Google. They were disappointed. Anyone who watches the available video recording of Andrew Brown's presentation could clearly see what his audience could see. A man under pressure. Andrew started by saying what a crushing disappointment it was to him personally that he was unable to reveal the new Glow2 solution.
He then outlined a brief version of the story so far and mentioned the subject of procurement. He talked about the legal advice sought and received by the programme board on the necessity (or not) of a procurement exercise for a free product, and even made a joke about it, inviting the audience to give their own views "on the back of a postcard please". Those who know Andrew Brown believed he was confirming that legal advice had confirmed that a procurement exercise was not necessary. A freedom of information request submitted to the Scottish Government asking for details of this legal advice, true to form so far, has also, been ignored.
Concern that Michael Russell's clear vision for Glow had been subverted
Since February, when I was tipped off by a local authority employee that all was not well in the kingdom of Glow2, I had been asking questions in private about all of the above, and particularly about the roles of ISIS, and the Glow programme board. I was extremely concerned about the way the board had appeared to take Michael Russell's very simple, straightforward, and popularly acclaimed vision for Glow2 and somehow turn it into an over-complicated disaster waiting to happen.
I also raised concerns, again in private, about the probity of the procurement exercise which, in my view, was unnecessary and skewed towards one particular solution provider, the same one which apparently had such a cosy relationship with ISIS where a Microsoft employee was based two days a week. I'd discussed these concerns with a few people to gauge opinion before voicing my concerns at a senior level within the Scottish Government. I continued to voice these concerns privately throughout April and May. Meanwhile, the work being done on the Glew authentication system by Charlie Love had slowed down until Charlie was called upon to defend Glew at an ISIS-run risk-analysis meeting where the project was declared too risky to go any further and was halted, at least officially.
Towards the middle of May a meeting took place between Michael Russell, Andy McLintock, Trudi Sharpe (from the Glow programme board), and Ken Muir (from Education Scotland). Following this meeting it seems that Education Scotland's team – Andrew Brown, Ollie Bray, and Charlie Love – were removed from any further work on Glow2 which then appeared to become an ISIS project.
Charlie Love's secondment to work on Glew/Glow, which had been arranged months earlier, was abruptly cancelled on the day he was due to leave his school and move to Education Scotland, and Ollie Bray has also returned to school. Sources say that Andrew Brown is now on sick leave. By now the procurement exercise was attracting substantial vocal criticism from education insiders across Scotland who felt that their hopes and aspirations for Glow had just been crudely undermined and shelved. The apparent cavalier treatment of popular educators added to the controversy.
Why did ISIS review work of colleagues with vastly more experience of education?
So, to summarise, Scotland had agreed a vision for the successor to Glow. A set of free tools with a secure authentication system had been created to transfer data from the original Glow before it was switched off. Google was the clear favourite among all those involved in the consultation and user testing. Work was well underway, carried out by a small, dedicated team which had worked hard to realise the shared vision. Staffing of the project was agreed with the secondment of a key member agreed and arranged with his employer. Legal advice was believed to have cleared the way for selecting the suite of apps without the need for a procurement exercise. And then ISIS stepped in and everything gradually came to a halt.
Why would ISIS want to review the work of colleagues with vastly more experience of education and an intuitive feel for learning and teaching? Several theories exist, but according to insiders the most probable could be these:
- ISIS and, indeed, the Glow Programme board don’t understand how education works and what its needs are, that teachers and other educationists want their ICT systems to fit their needs rather than the other way around – and grass-roots involvement is anathema to civil servants more used to administering systems and processes rather than innovating with others;
- That Microsoft applied behind-the-scenes pressure when it became obvious that Google would be chosen. This is not an unreasonable suggestion as Microsoft already had a close working relationship with ISIS:
- Taking over Glow has advantages for ISIS, the Scottish Government's IT unit – it would mean a bigger budget and more power and ISI has expertise with Microsoft services;
- The civil servants had messed up the project and were desperately scrabbling to cover up the mistakes, hence the unforthcoming FOI requests.
Whichever of these you put your money on, if not all of them, there's no doubt that education secretary Michael Russell had not been aware of the backroom shenanigans.The clamour for action, both in private and public via a torrent of tweets and blog posts fuelled by the refusal of government to comply with FOI requests, had become overwhelming. Michael Russell had to do something, and so he acted in the only way he could. By extending the original Glow contract with RM by another 15 months in order to buy time to sort out the mess created by the civil servants.
But at what cost? £5.5 million pounds to the very company associated with the first, uninspiring Glow, is a hugely expensive price to pay, not to mention the loss of world-class people from the project and the general air of mistrust that now surrounds it. Scottish taxpayers urgently need an independent review of the role of ISIS, particularly in relation to the procurement exercise and the possible waste of public money on this fiasco.
It's not too late to rescue Glow from the bureaucrats
Butt there's at least one ray of sunlight. Charlie Love has continued to work on Glew, and has made his work public so that teachers and schools can create an account and see just what we could have had in place by September to replace Glow, and at a fraction of the many millions of pounds we have alread paid for what is now a legacy system on which further development work is pointless. Glow in its present form will fade away as users desert it in droves, attracted by all the freely accessible tools out there on the web which actually work well. I've had a look around Glew and it's an amazing piece of work. You can see for yourself by visiting http://www.glew.org.uk/
So, where to go from here? Come on then Mr Russell; it's over to you. How about cutting through all the civil service failures, restoring Andrew Brown and his team to the Glow successor project, and getting them to carry out a review of just what has been achieved, and what's possible now there's a little more breathing space. Andrew Brown is one of the most respected figures in education today, and together with the international reputation of Ollie Bray and technical expertise of Charlie Love, he could manage the birth of a successor to Glow which, if not totally future proof (what is?), is at least what the people want, would work properly and meets the needs of learners and practitioners. It would also be well managed and provide better value for money than its predecessor.
Reclaiming Glow, in the same spirit of openness and consultation, would also put Scotland firmly back on the map for innovation in the use of technology for learning. Scotland's intention to "open source" the development of Glow, and sharing it, demonstrated the potential to change education and business models globally
Google and Charlie Love have just announced that Glew will be included in Google's current focus, its Fast Track for Scotland programme, based on Google Apps which Scottish schools may well gravitate to in the current, frustrating Glow hiatus. However, Charlie Love says he is still willing to work with the Scottish government to help complete the work started on Glow2 back in September last year..
What about a simple response to this sad tale of innovation and transparency turned to ineptitude, confusion and concealment? Some integrity and courage. There is nothing to stop Michael Russell cutting through the civil service red tape and picking up the phone...
Jaye Richards-Hill is a retired principal teacher and freelance consultant... who makes clocks from bottles
Artwork by Maia Terry
Video of Andrew Brown, Ollie Bray and Stuart Campball at March 2 Local Authority Event (Glow log-in required)
The Pamster – former Education Scotland emerging technologies development officer Pam Currie gives her take on events
Charlie Love Blog Post – #Glew - an open platform for learning with Single Sign-On
Jim Buchan (former Glow technical manager and now vertical architect at Cisco) reflects on Glow, Glow Futures and #Glew
Videos from Scotland's ICT in Education Summit (October 2011)
Glow2: Intranet or Ecosystem? – Theo Kuchel, researcher and IT education expert, on the need, or not, for another Glow
Freedom of Information requests