Six Recommendations For Glasgow
In making recommendations for Glasgow I should preface my comments by stating up front that these proposals should be considered to contained within the recommendations of the Logan Report — not as a counter set, or to replace the broad framework that Logan has laid out.
This document contains lessons to be learnt for, and proposals to shape the implementation of Logan in Glasgow.
These proposals should also be seen as widening the scope of the Scottish tech sector. Glasgow has its own Harlem Globetrotter problem with Edinburgh — competing with Auld Reekie on the capital’s terms is not wise. I propose a third road to global companies, a widening of Scotland’s capabilities.
Glasgow must reuse a lot of the infrastructure so painfully built up in Edinburgh, EIE, Turing, the network of funders, angels, VC, the legal and accounting firms aligned with the tech sector. In the fullness of time, after sustained success, that might want to change, but not now.
The implementation of Logan in Glasgow must be built on top of her existing communities, organisations and networks.
Edinburgh has CodeClan which widens the funnel for software engineers, Glasgow has CodeYourFuture — a similar, but volunteer, organisation that teaches refugees and other marginalised groups coding and software development. This remarkable and refugee-established and -led body has grown to chapters in Birmingham/Midlands, Manchester/NW, London, Rome, Medellin and Cape Town. (Full disclosure: I am a volunteer career mentor at Glasgow CYF).
Glasgow-based Rookie Oven has flown the flag for the Glasgow startup scene for over 10 years now.
The EU’s General Data Protection Regulation (GDPR) has been one of the first examples of the EU flexing its regulatory muscles and its impact has been tremendous — bringing visibility to the murky practices of selling customer data and tracking people’s behaviour on the web, for private companies, and behind them the states that regulate them.
GDPR is a great example of EU softpower with many American web properties forced to comply to an extra-territorial regulator. WordPress sites consists of about 1/3 of the internet, and WordPress platform compliance with GDPR was led by a Glasgow-resident, and unpaid volunteer, Heather Burns.
Giants, unsung, walk in the dear green place, they need to be at the heart of Logan in Glasgow. Building the next generation Glasgow tech infrastructure needs to build on these foundations.
Recommendation 1 — Culture Kick-Starter
In the context of Logan’s recommendation No 1 — a National Tech-Scaler backbone — Glasgow should learn the lessons of Edinburgh.
1.i — Culture First
In Edinburgh CodeBase achieved its success by decanting a thriving cultural tech and startup scene and dense network of relationships into a couple of quite unsuitable buildings.
The Glasgow Tech-Scalar (which I would guess will be the Tontine) needs to take active steps to bring the existing Glasgow scene into a structured environment rather than try and bootstrap a culture from scratch.
The key elements are (that I am familiar with):
- Glasgow TechMeeup — the regular tech meetup
- CodeYourFuture — the code school turning refugees and disadvantaged groups into software engineers
- Rookie Oven — the regular startup meetup where entrepreneurs meet, learn and collaborate
1.ii — Timetabled
A timetable of tech events for Glasgow needs to be curated to avoid clashes with each other as much as possible.
The key role of beer/soft-drinks’n’pizza should be recognised. A catering grant budget should be established upfront and new and current meetups should be offered 6 months free food.
Meetup sponsorship is a very effective crude measure of viability. Tech-Scalar should help meetups start but then encourage them to get their own sponsorship.
1.iii — Double Heartbeat
The meetup ecosystem should be organised around the principle of heartbeat and orbiting meetings, with two heartbeats, tech and product.
The Tech-Scalar should reach out to Glasgow University’s Adam Smith Business School and the Strathclyde Business School and try and get them to co-locate some of their meetings, perhaps to take on the organisational burden of the product heartbeat.
1.iv Speaker Fund
The national Tech-Scalar network should establish a pre-funded grant for international speakers travel:
- a small number of US
- a few more European
- a lot of English
- inter-Scotland
The process of applying for support should be simple and delegated down: a 1 page application resulting in an immediate up/down decision.
This fund should be available to all meetups and events in Glasgow. An initial yearly fund of $15,000 for Glasgow would go a tremendous way.
Recommendation 2 — The Glasgow Road To Global
The strategic narrative that Glasgow should focus on the is the COSS road to global. The challenge is to develop processes and institutions to uncollapse the funnel. Brexit and looming recurrent Trumpism are real shaping calamities — and the COSS model allows us to sidestep some of their consequences.
On top of this single strategic direction I make a number of recommendations that layer onto it, all drawing on the lessons of Edinburgh and all reusing existing infrastructure and ecosystem resources.
The goal is to build in Glasgow a technical and startup community and ecosystem around the early stages of the COSS pipeline and from within that community support the creation and growth of high-value COSS companies.
COSS companies have some different business models based on their internal technical structure. Simple libraries of software are composed and organised in different ways. On the one side we have tools to support the development of new bespoke software — frameworks and ecosystems. On the other side is the deployment of infrastructure on which new bespoke software runs, infrastructure like databases and storage and full platforms.
Both of these will be increasingly critical.
COSS has its own barriers to success. One of them is the dual challenge of building a commercial company at the same time as building a community of software developers. COSS companies only become entrenched, with defensive moats and network effects to protect their income and margin if they manage both. Many companies fail by being only successful on one side. Some communities grow to global scale but the associated commercial entity remains stillborn. Other companies are simply software companies whose offering happens to be open source, but have no actual protective developer or user communities around them.
When we look at the successful US COSS companies they have some common themes. They emerge fully formed from deep research communities or big tech firms. The have operational hygiene because they reuse existing governance structures. This provides certainty and predictability of governance, of licensing agreements, of roadmaps, of codes of conduct and other critical but unloved and unappreciated emotional labour of open source communities.
They get books and training materials produced and available early, the stamp of O’Reilly that provides the needed testimonial and social proof that they are serious and here to stay.
The world produces open sources software projects in huge numbers. Most never get any traction or developers. Developers and companies are reluctant to commit time and resources to projects that might never take off. Social proof, evidence that an open source project will have a long shelf life.
The primary success of open source projects is ideologically touted to be due to their technical virtuosity but actually derives from the quality of their social production: social proof, communication skills, governance and roadmaps, community building and onboarding.
I propose that we mitigate the barriers to success for COSS by sequencing these two tasks:
We do this buy creating new tech event based on the Turing Festival Fringe — that is to say a festival of tech events dedicated to OSS/COSS and using that to create communities and networks around and within OSS and existing functioning Scottish firms that are dependent on OSS to make COSS startups from them.
As well as being collapsed the OSS funnel is also grossly mismatched. Individuals working for companies that rely on OSS contribute as individuals to projects — not as employees. Their companies struggle with ‘getting a product’. Most small firms using OSS are consultancy or people-based (ie not intrinsically scalable nor venture fundable). Many have aspirations to move into scalable, and hence fundable, product companies.
These firms have the basic management operational hygiene capabilities (legal, HR, IP) but lack business models that can scale globally.
These companies attempt to transition to scalable startups by funding product development with client money. This has not historically been enormously successful — although some, like Kotican, which sold services to the rapidly growing FanDuel and Skyscanner have been successful, but not at global scale.
Instead of switching from service-to-product we should explore switching these firms from unscalable service models to scalable-COSS-service models.
This reorganisation of OSS/COSS community development is just a Scotch Pivot — reducing the cost of experimenting with new software communities and then match those communities with existing companies that already have cash flow and primitive commercial acumen (invoicing, profitability, the notion of cash reserves and sustainable work patterns) to launch small scale COSS companies.
Typically open source products mature slowly and take many years to get to having their own conferences or meetups. (with the partial exception of O’Reilly’s OSSCon which is deeply embedded in, and shaped by their business, and its connection in Silicon Valley).
The proposed event would be a festival of tech conferences and critically has the same risk distribution structure as the Edinburgh Festival Fringe which is financially stable even if individual events within lose money, or worse, go into bankruptcy. The Fringe model is an enabler and not a profit centre in its own right — the goal is not to create a single commercial company but to build networks and communities. As a public good, it should be considered for state support — otherwise like the main Turing it will be forced to walk up the value chain to £500 tickets.
It is important to note that the actual events we will try and grow under the Fringe will need to adapt to the new world, breaking out their ticketing to include per-session tickets as well as half-day and full-day ones. The venues will need to be densely packed and walkable.
Unlike the Edinburgh Festival Fringe though — the individual components of the festival are intrinsically financially viable. £2k to £5k sponsorship is very achievable for quite small events if they have the right focus — a half day with 30 people on a niche topic. The key is matching the event with the sponsor.
The event must be seen in the wider context. Glasgow (and actually Scotland as a whole) has a strong CS research tradition and the academic context needs to be shaped to act more like the Stanford model as the ovary of a COSS company development oviduct.
Unlike the US, we are second wave in the 4th Industrial Revolution — a critical part of our innovation is not pure or deep tech but in commercialisation, embedding, productionising. Our business schools need to be a key player in developing new business models, providing actionable research into the COSS sector, lessons learned globally about success and failure and so on and so forth.
The Fringe must locate itself like the Edinburgh Festival Fringe within a universe of feeder institutions. The Edinburgh Festival Fringe now draws on cultural and artistic organisations around the world to produce the events that are showcased in Edinburgh.
So too this new ecosystem we are trying to build must consciously build its feeder networks: initially heavily from Scotland, then rUK and then Europe.
The map of the future is largely shaped by direct flight availability from Glasgow Airport, cities in Europe with a direct flight to Glasgow or Edinburgh and a University with a CS department.
2.i — Commission The Turing Festival Team
The new festival should be organised by the Turing team.
Full disclosure: I was a founder at Turing and have an emotional investment in its success, but not commercial relationship with it.
2.ii — Timing
It should take place in the Easter University holidays at the other end of the year from the main Turing
2.iii — Three Year Horizon
There should be a three year initial horizon
- year 1–80% Scottish
- year 2–10% EU, 50% rUK, 40% Scottish
- year 3–10% US, 20% EU, 40% rUK, 30% Scottish
2.iv — Co-location
The local organising team should be co-located with the Glasgow Tech-Scalar (like the Turing team in CodeBase)
2.v — Outreach
There should be an outreach programme contacting and bringing in individuals, project and OSS companies as well as student organisations. It should map its activities to the European transport connections of Glasgow/Edinburgh.
Attempts should be made to spread the TechMeetUp brand to other University cities in Europe and this should be funded.
There should be an outbound travel fund to send Scottish speakers on the road to build these networks.
2.vi — Superspreader Programme
Technical Authors are networking super-spreaders and technical soft-power projectors. There should be a sabbatical fund that pays a stipend of £1,500 a month for 6 months that lets Scottish-resident techies write tech books. It should be conditional on having a publisher and steps should be taken to encourage an indigenous tech publisher scene to emerge.
Major Scottish tech firms should be approached to see if technical authorship can be incorporated into their career development programmes and funded directly by them.
2.vii — Integrated Business Schools
The various Business Schools should be approached to perform analysis on the COSS sector (full disclosure: much of the insight in this report is anecdotal and observational — a more rigorous review should certainly be conducted) and to communicate that to the wider community.
2.viii — Integrated Investment (3 years out)
In the fullness of time new Investment Events (organised by or in co-operation with Edinburgh’s EIE team) should be established.
2.ix — Integrated Externalised Excellence (3 years out)
In parallel to this investment event a new international golf’n’whisky signature event should be set up (Clubhouse style, exclusive Silicon Valley VC etc, etc).
At the risk of sounding ridiculous I would say a Scottish Burning Man, but my inner Calvinist revolts. The key is it should be a repeating calendar event where key constituencies meet in social circumstances — something that international funders and tech players look forward to all year.
This on paper would be entirely independent of both the investment event and the Turing Fringe Festival but closely scheduled against them (immediately before or after) to give senior investors an excuse to come to Glasgow for them — just as Steve Wozniac dropped in on Turing whilst he enjoyed the Edinburgh Festival.
Recommendation 3— Build Out A Post-Covid Meetup Infrastructure
As companies switch to more and more remote working, they will more and more start behaving like existing remote working companies.
A characteristic feature of them is regular, yearly or twice yearly all-hands, or departmental meetups where people are brought together to work intensively and build the internal relationships that companies need to thrive.
Lots of companies will be wrestling with this problem. Lots of cities will start producing packs (hotels, attractions, leisure packages) to lure them in.
Glasgow should focus on tech companies, particularly those that use or have open source infrastructure and endeavour to build a long lasting relationship with them, getting them to hold all hands, building bonds and connections that can be converted into sponsorship opportunities, activities within the Festival of Conferences described in Recommendation 3.
Recommendation 4— Build Out The Funnel-Widening Infrastructure
There is room for more than one model for widening the funnel. CodeClan — a fully staffed model is one — CodeYourFuture — a volunteer-led organisation is another.
The advantage of being fully-funded is dedicated training staff and the focus that paid employment brings. The advantage of the volunteer model is that the learner is embedded in a deep network — the mentors and teachers are volunteers and embedded in their professional lives.
The difference is not as stark as it appears — there is room for paid support — this is the model of CodeYourFuture in Birmingham with two full-time staff members paid for by local government.
CodeYourFuture Glasgow is also piloting the build out of an Open Source Maintainers Group — acting as both a paid work and networking bridge from graduation to permanent employment.
There is an emerging network of sponsorship for people to work on Open Source emerging around Github’s sponsorship infrastructure. CYF is trying to grow an income stream to help fund CFY graduates on the journey from graduation to full time employment.
Any funded support for CYF should help the development of the Open Source Maintainers programme to a wider community. The aim should be to nurture a community of OSMs in Glasgow and also a try and establish the OSM programme as a global programme that seeks sponsorship from global companies and takes in graduates for the many global organisations that are teaching marginalised and economically excluded groups to code.
This would enable Glasgow to start creating work networks that would act as a feeder to our global COSS ambitions.
An evaluation of CodeYourFuture and its potential should be undertaken with a view to providing a measure of paid support.
Full disclosure: I am a career mentor volunteer with CYF and piloting the Open Source Maintainers programme.
Recommendation 5 — Summer Experience For Students
Logan recommends a National, pan-university Tranzfuser-style summer-school. University students have 3 summers. There is room for more summer activities.
There is money left lying on the table. Google Summer Of Code pays for students to work on 200 open source software projects over the summer, with fixed mentoring and help provided by volunteers.
The stipend in Glasgow is £5.4k for the summer.
In addition there are other programmes, like Outreachy, which provide similar sums for interns in Open Source from backgrounds that have significant barriers to participation.
There should be a post established co-located with the National Tech Scalar in Glasgow to identify additional programmes like these and encourage and manage applications to them.
For Google Summer Of Code this should aim to get 50 to 100 students placed, with other programmes appropriate numbers.
For these programmes you can only apply as individuals not groups but that doesn’t preclude putting a collective experience around it.
A very lightweight wrap should be implemented: a slack channel, regular meetups, a summer finale show-and-tell in the presence of industry people. The aim should be to maximise money left on the table and build out networks and experience — both in the Glasgow scene and wider.
As we develop a COSS strategy getting core projects into GSOC to build out the community and developers outwith of Scotland should also be pursued.
Recommendation 6— The Regulatory Road
The best way to shape the future is to discover it. What has been going on with tech cannot continue. If Scotland wishes to prosper in the future world, then it behooves us to understand and shape the regulatory environment that will have to be established.
The Universities in Glasgow need to seize the opportunity and establish an institute or school dedicated to studying the international tech industries and how the international community might want to regulate them in a fashion that transcends local markets and local powers.
The quicker we start debating, discussing and synthesising the possible approaches the better prepared we will be for when it comes to pass.
The situation we find ourselves in closely resembles the fraught period leading up to German rearmament in the late 1940s and early 1950s. The French didn’t trust the Germans, the German’s didn’t trust the French.
Slowly, collectively, an organisational architecture of mistrust was established. The strategic war industries of coal and steel where first removed from national control and internationalised. The nuclear industry was likewise subsumed into Euratom.
The skeleton of the European Union — the Parliament, the Council Of Ministers, the European Commission were all at first instruments of industrial regulation: they came from the European Coal And Steel Community
The same atmosphere of distrust exists now. Only a fool would trust the big American firms with the data they have and the power it gives them. But only a fool would trust their own government, let alone someone-else’s, with the same power.
A single example will suffice — Cryptography. It has long been know that the process that produces the cryptography that the modern internet needs to function has been penetrated by the NSA. The working assumption was that that is a benign situation. In the age of Trump that looks like a disastrous error.
The only multinational body familiar with an architecture of mutual distrust is the European Union. It is the only defender of international liberal order. It is the only place that could launch and support an independent international cryptography institute.
We find ourselves physically but not mentally outside the EU, but we should prepare for the new world and try and shape it as best we can.
Scotland needs to have the sort of intellectual firepower that can image such a regulatory future — one which is both capitalist and liberal, a free market under the state — based on the architecture of mutual distrust that has transformed Europe and which is the blueprint for the modern world.
If we do then we can shape the world, and in doing so find a place in it for Scotland, for Scottish firms, our place in the future.
This is the longest, the hardest road. That makes it all the more important.
